Cadence Design Systems Stock

Good afternoon. My name is Jade, and I will be your conference operator today. At this time, I would like to welcome everyone to the Cadence Third Quarter 2026 Earnings Conference Call. [Operator Instructions]

I will now turn the call over to Richard Gu, Vice President of Investor Relations for Cadence. Please go ahead.

Thank you, operator. I'd like to welcome everyone to our third quarter of 2026 earnings conference call. I'm joined today by Anirudh Devgan, President and Chief Executive Officer; and John Wall, Senior Vice President and Chief Financial Officer. The webcast of this call and a copy of today's prepared remarks will be available on our website, cadence.com. Today's discussion will contain forward-looking statements, including our outlook on future business and operating results. Due to risks and uncertainties, actual results may differ materially from those projected or implied in today's discussion. For information on factors that could cause actual results to differ, please refer to our SEC filings, including our most recent Forms 10-K and 10-Q, CFO commentary in today's earnings release. All forward-looking statements during this call are based on estimates and information available to us as of today. and we disclaim any obligation to update them.

In addition, all financial measures discussed on this call are non-GAAP, unless otherwise specified. The non-GAAP measures should not be considered in isolation from or as a substitute for GAAP results. Reconciliations of GAAP to non-GAAP measures are included in today's earnings release. For the Q&A session today, I would ask that you observe a limit of one question only. If time permits, you can re-queue with additional questions.

Now I'll turn the call over to Anirudh.

Thank you, Richard. Good afternoon, everyone, and thank you for joining us today. I'm pleased to report that Cadence delivered another outstanding quarter with all key metrics coming in well above our guidance. Backlog reached a record of $8.2 billion, reflecting strong bookings driven by growing customer demand. Given our strong business momentum, expanding technology leadership and compelling AI growth opportunities, we are raising our revenue outlook for the year to 21% growth, and our EPS outlook by $0.30 to $8.40. John will provide more details on both our Q3 results and the updated financial outlook.

Agentic AI represents a transformational market opportunity, and Cadence is exceptionally well positioned to capitalize on it. Our differentiated three-layer framework is anchored by our physically accurate simulation and optimization engines in the critical middle layer, tightly integrated with AI agents and orchestration at the top and accelerated compute and proprietary data at the base. Agentic AI continues to evolve at a remarkable pace. With our AI super agents gaining strong traction, we are now addressing the long-standing customer demand for domain specialized models. We believe the future of Agentic-AI-based design is increasingly domain specialized models combined with intelligent model routing for performance, and reinforcement learning loops with underlying tools for recursive self-improvement.

Today, I'm excited to introduce my stroke, a portfolio of domain specialized AI models, purpose build for chip and system design. By controlling all three layers, our AI super agents, Maestro models and proven EDA and SDA tools, we are uniquely positioned to deliver optimal performance, efficiency and throughput results for our customers. Initial results are very compelling, demonstrating up to 30% higher quality and up to 5x greater token efficiency versus general purpose foundation model. Moreover, the Maestro models can be further post trained on customers' proprietary data to deliver even better outcomes. Our model router chooses among our models and other leading frontier and open wide models as needed for optimal results. Customers can opt to deploy Maestro models in the cloud or in their on-prem environment. Maestro is gaining strong early customer tractions with several deployments underway.

Momentum of our AI super agents continues to accelerate with compelling early results and more than 100 active customer engagements. We added a new agent to chip stack for RTL generation and early PP optimization. That created RTL delivering an average of 24% area and 18% power reduction versus pure foundation model generated code. Importantly, this demonstrates the power of combining our AI super agents with AI models and our underlying EDA engines to achieve outcomes that foundation models alone cannot deliver.

Honda is using the RTL generation agent and AI-powered automation to improve productivity from specification to RTL development. In Q3, MediaTek deployed chip stack into its production digital verification flow, achieving 50% to 70% improvement in verification coverage and 1.4 to 2x reduction in engineering efforts with deployment now scaling across design teams. In analog and custom design. ViraStack saw 4x increase in customer engagement with deployments in various production designs at multiple customers. The combination of agentive workflow, embedded skills and our Virtuoso platform is unlocking significant new productivity opportunities in a traditionally difficult to automate domain. We are seeing strong customer interest in InnoStack, our digital AI super agent ad customers push to improve PPA, accelerated design closure and boost productivity.

In Q3, the Renesas MCU team successfully deployed InnoStack across a wide range of engineering costs, achieving a 34x productivity improvement with plans to expand its usage to layout automation. Also in Q3, Samsung Foundry successfully used InnoStack to dramatically improve engineering efficiency and accelerate PPA exploration, reshaping its approach to technology migration and BTCO optimization.

Socionext is deploying InnoStack in its production design environment. And I expect its combination with our advanced physical design tools to deliver up to a 5x overall productivity gains while maintaining design quality. In PCB and 3DIC design, the early response to our AuraStack AI Super Agent has been overwhelmingly positive. Nokia is deploying AuraStack on our Millennium supercomputer to advance electronic system design at scale. By unifying Agentic AI, our physics-based solvers and accelerated compute in a secure on-prem platform, Nokia is achieving a 5x acceleration in key design and analysis workflows.

Moving on to our businesses. We are delighted that all product groups delivered outstanding results with year-over-year revenue growth of 20% in core EDA, 24% in IP and 30% in SD&A. Accelerating design activity continues to drive strong demand for our core EDA, SDA and IP products. At the same time, we are seeing amazing customer momentum for our Agentic AI portfolio, which is also fueling increased demand for our underlying products. We are pleased that recurring revenue sustained its strong momentum, growing 18% year-over-year. I'm particularly proud of the significant wide-ranging agreement we have signed with top customers. We furthered our long-standing partnership with a global marquee systems customer through a significant wide-ranging expansion of our core EDA, hardware and SDA solutions.

In Q3, Cadence and a top telco deepened their long-term strategic partnership with a broad expansion across our AI offering, EDA software, hardware and SDA solutions, establishing Cadence as their primary design partner. Core EDA revenue growth was driven by broad-based strength across the portfolio. Our digital and sign-up business continue to expand, achieving 16 new full flow wins during the quarter. As agentic flows increasingly require full vertical workflows, there is growing demand for our synthesis and sign-off solutions. Our sign-up product saw growing adoption and our verification software product had strong growth in big digital and memory designs with increasing usage at major hyperscaler and memory companies. As AI and HPC designs continue to grow in scale and complexity, hardware-assisted verification has become an increasingly strategic capability for our customers. This is driving sustained unprecedented demand for our Palladium and Protium emulation platforms, which delivered yet another record quarter. We added several new logos and saw a meaningful expansion with marquee customers across a wide range of verticals.

Our hardware cloud offering showed strong growth with a leading semiconductor company adopting more hardware cloud services to accelerate new product development. In analog and custom design, Spectre FX continues to gain momentum, highlighted by major competitive displacement at a marquee infrastructure silicon provider with multiple new 2-nanometer projects adopting our platform. Our IP business delivered another outstanding quarter, driven by deepening engagements with leading hyperscalers and semi customers. We continue to benefit from strong AI-driven tailwinds, fueling accelerating AI infrastructure investments, growing adoption of custom silicon and increasing demand for high-performance IP. Along with the transition to advanced nodes and chiplet-based architectures, these AI-driven trends are creating a strong full for our differentiated STAR IP portfolio across interface, memory and foundation IP. We are well positioned for sustained growth as we continue to deliver superior power performance and area across our expanding advanced IP portfolio.

A leading space technology company, selected our IP portfolio for their custom chip program, paving the way for a broader services and silicon engagement. Tensilica delivered a record revenue quarter and we announced a joint development with analog devices across next-generation audio applications.

Our system design and analysis business delivered another strong quarter, driven by growing adoption of our integrated solutions and proliferation of our expanded electronics and physical systems intelligence portfolio. Allegro momentum continued, including a competitive displacement at a leading EV customer that chose Cadence's integrated PCB design and multiphysics analysis platform as its primary solution. Our Millennium platform had strong adoption driven by deployment of our GPU-accelerated electronic and mechanical systems suite, along with our reality digital twin solution.

During the quarter, Switch, a leader in sustainable high-density AI infrastructure selected our Millennium and reality digital twin solutions to advance the design and simulation of its next-generation AI factories. The MSE business is the integral part of our physical AI strategy, and its integration is progressing well. MSC Adam our industry-leading multibody dynamic solution helped secure a competitive displacement at a leading power company, along with adoption of other cadence products.

In summary, Cadence delivered exceptional Q3, and I'm delighted with the accelerating momentum of our business. We are executing extremely well in a vibrant AI market, have our strongest competitive position ever and are extending our leadership in AI. Our highly recurring and ratable business model, combined with growing customer R&D investments is positioning us very well for continued strength in this dynamic AI environment.

With that, I will turn it over to John to provide more details on our Q3 results and our updated 2026 outlook.

Thanks, Anirudh, and good afternoon, everyone. Cadence delivered strong third quarter results driven by robust demand across our businesses and growing momentum from AI-driven designs. Growth is accelerating and broadening Q3 revenue grew 22% year-over-year with strong double-digit growth across all products. We exceeded our guidance on every key financial metric and ended the quarter with record backlog of $8.2 billion.

Here are some of the financial highlights from the third quarter, starting with the P&L. Total revenue was $1.633 billion. GAAP operating margin was 26.2%. Non-GAAP operating margin was 44.1%. GAAP EPS was $1.05. Non-GAAP EPS was $2.11.

Next, turning to the balance sheet and cash flow. Our cash balance was $1.452 billion, while the principal value of debt outstanding was $2.650 billion. Operating cash flow was $424 million. DSOs were 66 days, and we used $400 million to repurchase Cadence shares during the quarter, reflecting our confidence in the business and our commitment to returning capital to shareholders. Underscoring that commitment, our Board has authorized an additional $2.5 billion for our share repurchase program.

Before I provide our updated outlook, I'd like to highlight that it contains the usual assumption that export control regulations that exist today remains substantially similar for the remainder of the year. For our updated outlook for 2026, we now expect revenue in the range of $6.387 billion to $6.417 million. GAAP operating margin in the range of $28.2 to 29.2%. Non-GAAP operating margin in the range of 44.2% to 45.2%. GAAP EPS in the range of $4.84 to $4.90. Non-GAAP EPS in the range of $8.37 to $8.43. Operating cash flow of approximately $2 billion, we expect to use more than 50% of our free cash flow to repurchase Cadence shares in 2026.

For Q4, we expect revenue in the range of $1.695 billion to $1.725 billion. GAAP operating margin in the range of 30% to 31%. Non-GAAP operating margin in the range of 44% to 45%. GAAP EPS in the range of $1.23 to $1.29, and and non-GAAP EPS in the range of $2.19 to $2.25. And as usual, we published a CFO commentary document on our Investor Relations website which includes our outlook for additional items as well as further analysis and GAAP to non-GAAP reconciliations.

Overall, I'm pleased with our Q3 results, especially our recurring revenue growth of 18% year-over-year. At the midpoint of our updated 2026 outlook, we now expect revenue growth of 21%, operating margin of 44.7%, and EPS of $8.40. As always, I'd like to close by thanking our customers, partners and our employees for their continued support.

And with that, operator, we will now take questions.

[Operator Instructions] Your first question comes from the line of Jason Celino from KeyBanc Capital Markets.

Anirudh, really nice to see the Maestro in announcement. You know I'm not an EDA engineer, I'm not an AI scientist, so maybe can you just explain what this is a little bit better. It sounds like it's part model router, but also maybe domain-specific models. So is Cadence developing these main specific models? Maybe just elaborate, and then how the monetization of this new product will work?

Yes. Hi, Jason, and thank you for the question. Yes, there's a lot in that release. So let me try to unpack it a little bit. First of all, we are very pleased with the momentum of our agent stack and all the four super agents. And you can see there are like more than 100 engagements now. So as we work with customers over the last 6 to 9 months, several trends became pretty clear, okay? And to me, we are launching Maestro as what I would call the next generation of agents. And there are three big factors in the next generation of agents. So one is, like what has happened in the last one year is that, of course, we work with all the Frontier labs from the very beginning, whether they're doing chip design themselves, or we are using their LLMs. But over the last one year, there is this emergence of very, very good open rate models, okay? And I always wanted to do like domain specialized models, but now it's finally possible with open weight models. And it's actually very cost effective to post train these models, to deliver like domains intelligence, which is remarkable. We're getting up to 30% better quality by using our tools and data to post train these models. And also, you don't need like -- it's a portfolio of model. So all the way to like small decisions LLMs to small and medium LLM all the way up to fronter class open wage models post trained on our data and intelligence. So that's the first part of Maestro. So it gives really customization for chip and system design and gives much, much better results. It's actually quite remarkable how much these models can be improved by post training, okay. Now the second part of that is in our agents and when they do these tasks have a variety of tasks, right? So we don't need only one size or type of model to do the task. So the second part is model routing. Just like you don't need to use an airplane to go to your local grocery store, right? So similarly with agents, you have a variety of tasks that we need to do from simple task to like deep thinking and reasoning tasks. So the moderator will efficiently knowing what the task is from our super agents use the right model. And that can easily give like up to 5x improvement in token efficiency. So either much lower cost or much more work for the same cost. And then the third thing is which I always talked about, this reinforcement learning loop with the underlying tools. And we have great examples we talked about our new RTL agent that one of the customers publicly endorsed Honda gave 24% area and 18% power because the integration with the underlying tools, and how we call them, and continuous self-improvement whether it's skills or models gives us a unique advantage. So if you combine these three things together, domain specialized models -- domain specialized family of models, plus model routing plus deep integration with our tools and data, it gives a unique differentiation that nobody else can do. And it also allows us to control all three layers of the cake and innovate in all three layers of the cake. So I'm actually super excited. And I think this is what's going to happen in all industries as these LLMs mature, this domain specialized intelligence model routing and integration with underlying engines.

Your next question comes from the line of Jim Schneider from Goldman Sachs.

I was wondering if you could maybe give us an update on the monetization of the Agentic AI solutions you're seeing. I believe last quarter, you talked about recurring revenue growth of 20%. This quarter was 18%. Maybe talk about what the difference there and the recurring was? And maybe underlying that, whether there was any acceleration in the Agentic part of that recurring revenue.

Yes. Thanks for the question. I'll start, and then John can comment. The -- I'm very proud of the recurring growth. And I mean, this quarter, there were some specific things. Like if you compare it to Q3 of last year, there were some specific things with China revenue moving around, which is not there this year. But in general, recurring revenue is doing phenomenally well. And if you step back and compare last year, recurring revenue was 11%, and it's 18% for this quarter, and it was very good last quarter. So I expect recurring revenue to do very well in 2026. And this is a result of, of course, our Agentic AI solutions, but also Agentic AI solutions driving more usage of our base tools, which, as we always said, is going to happen. So that's what we see. A lot of engagement is [Audio Agentic and a lot of usage of our base tools. So that provides a lot of growth for recurring and then you add that with our ongoing strength in IP and hardware, and you get a very, very good set of results.

Yes, I'd agree with that, Anirudh. I mean it basically reflects strong demand across our software portfolio and healthy customer expansions and continued strength in core EDA. Yes, I think that's the way to look at it in terms of strong high teen growth in recurring revenue. We're very pleased with this.

Your next question comes from the line of Vivek Arya from Bank of America Securities.

Anirudhn, we recently heard three things from your competitors that they can grow sales kind of at a mid-teens pace next year and then compound it that they can achieve 50% operating margins? And then also, interestingly, there is a royalty-based model that can help add nearly $1 billion to their top line over the next several years. So I realize, right, this is not the call to give a multiyear growth outlook, but I was just hoping to get your perspective on how you envision sort of this growth and royalty component of Cadence, and how well is Cadence is positioned to maintain or expand your market share and your leading operating margins over time?

Yes. Thanks for the question, Vivek. So overall, I'm very pleased with our growth and not just growth, but profitability and margin. And we always talk about like a rule of 40 metric or revenue growth plus margin. So I think we will end the year, if you look at our guide, about 21% revenue growth and slightly less than 45% margin. So we will end up north of 65% in 2026, okay? That's the important thing to point out Okay. And we still have one more quarter to go. So we'll see -- it's an important quarter for us. Fourth quarter is always important. So we'll see how that finishes, and we can update you like we usually do in February for the next year. But overall, the environment, I think, is very positive for us. As you know, the AI build-out continues the amount of design activity in all aspects of AI and even the traditional companies and analog mixed-signal companies, digital company, memory companies, system companies, so the amount of design activity is the strongest that I have seen. And our -- we're gaining share in almost all segments of our business, whether it's IP or hardware or EDA, Agentic AI. So our product portfolio, I think, is the strongest it has ever been. And then we have this huge opportunity with Agentic AI. And I think we are redoing how agent will be defined for all industries with our own domain specialized models and the flexibility to continue to work with all the frontier models. So if you look at all these three things, I feel Cadence is in a very, very strong position. Now in terms of your specific question on royalty, we are always looking at different kind of business models. So in IT, we always had royalty for a while in Tensilica, and now we have royalty in our design IP business as well. But to me, that's just one vector. We're always looking at how to best serve -- give value to our customers and then get, of course, value and return.

Your next question comes from the line of Siti Panigrahi from Mizuho.

I just want to ask on hardware I see the inventory again grew this quarter, I'm wondering hardware is a strong quarter. How sustainable hardware growth for heading into next year 2027. Any color would be helpful there. And a quick clarification, John. There was a spike in G&A expense in Q3. Was there anything onetime there?

Yes. Sure, Siti. Great. Look, I'll take the hardware question first. The hardwood continues to perform exceptionally well. I mean we delivered another record quarter, supported by strong demand from AI and HPC customers. And we're also seeing growing interest in cloud-based hardware access. The hardware-assisted verification is becoming increasingly important because customers are designing systems of unprecedented complexity. They need the capacity to verify those designs before committing to silicon. We have a strong pipeline entering Q4. We continue to expect another record year for hardware as always, harder could be lumpy from quarter to quarter depending on product deliveries and customer stellate schedules. But the underlying demand environment remains very strong, very healthy, and we're very, very encouraged by the pipeline of customer engagement. In terms of inventory, the increase in hardware inventories to support customer orders that we have a significant backlog that we're trying to work our way through, but there are no capacity or supply chain constraints in terms of addressing that. We're just building inventory to take those numbers. I think you asked about that's in part of the quarterly movement this quarter reflects the timing of annual cherishable contributions, which occurred in Q3 this year compared to Q4 last year. That would extrapolate that timing difference to the underlying G&A run rate. So there's a slight impact on G&A in Q3 because of that -- those cherishable contributions.

Yes. And Siti, just to add on the hardware piece. I mean Palladium and Protium performing exceptionally well. Not just for AI. And of course, we are working with all the major AI companies, but also any other kind of chip design right now, hardware became a indispensable, almost indispensable way to design these things. Now one thing is always like will this trend continue? And we'll see, but there's one megatrend, okay? So hardware is -- you probably know this already, but just to highlight it even further. Hardware is normally sold or consumed based on design size. So if the design size is bigger, then you need more and more hardware. So actually, I was recently just an example of what is going to happen to design, I mean the design sizes are already big, but what's going to happen to design size in the future. So just to give you an example, I was that IMEG recently, I gave a talk at IMEG, which is a European research consortium. And TSMC will also gave a talk same session. So TSMC showed their process road map and said in the next 5 years, the chip size will go up by 48x. Now typically, this would happen to be like 5 to 10x in the old days, but I think it's going to be 48x because you have to add regular technology scaling plus 3D IC. So all these 3D IC packaging wafer scale integration. So that's going to even further create more and more demand for hardware and die-to-die IP and things like that. But -- so now our customers are not only verifying 1 chip or 1 chiplet, they're verifying those chiplet working with other chiplets and the size of the chips is going up and then a amount of chiplits are going up. So I think this trend, 1 criticality of hardware-based design and verification combined with the growth in size is going to continue for some time.

Your next question comes from the line of Joe Quatrochi from Wells Fargo.

I was curious if you could just kind of talk about the change in the '26 outlook, particularly from an operating profit perspective because I think when you flow through just kind of the incremental margin on the upside of revenue, it's quite significant relative to maybe your kind of targeted over 50%. So maybe just any help there in terms of understanding what's going on there.

Sure. Yes. Look, we continue to be very pleased with the underlying profitability of the business, despite integrating Hexagon's DD business and making targeted investments in our strategic growth opportunities, we're now expecting approximately 44.7% non-GAAP operating margin for 2026, which is slightly above last year. But if you look at the revenue growth this year over last year, it's about $1.1 billion. Non-GAAP operating profit is growing about $500 million. And would you consider the Hexagon's DEV being part of that. And that's not really contributing much to the way of operating profit at all this year. But -- and I think we said at the start of the year, we're expecting Hexagon to be about $160 million for the year, is doing slightly better than that, but it's pretty much on target. But I mean, if you back that out, we'd be where you typically expect us to be, which is kind of mid-teens for revenue growth of high teens for EPS growth in our usual kind of 54%, 55% incremental margin rate. So the Hexagon DD is kind of impacting the incremental margin for this year. But we'd expect that to help us next year because the integration go really well. And we're taking steps now to improve things for next year in the same way as we did with paid the last couple of years.

Your next question comes from the line of Charles Shi from Needham & Company.

Glad to hear about the Metro model portfolio just announced today. In light of your competitor who announced the GPG synopsis, can you tell us what your thoughts are regarding collaborating maybe with the Frontier labs for a specialized model, given that Maestro based on what you just described, it sounds like it's a collection of open models post train by cadence. And I think [indiscernible] customers also have choices between your model portfolio, and let's say GPT synopsis, what do you tell your customers when they have both choices and why they should take Cadence?

Yes. Thanks for the question, Charles. So first of all, I just want to point out that we have worked with open AI from the very beginning of when they got started. So in terms of chip design, and we have all kinds of collaborations with open AI. And same thing with Entropic and Gemini, the three leading frontier model. So we always discuss all kinds of collaborations with them. Now in terms of what we want to make sure as -- in Maestro, the key thing is Maestro is not just our own models, but it still has maximum flexibility to use the right model for the right task. So let's say, if certain for a certain percentage of task, a certain frontier model is better, we will, of course, use that to give the best performance to our customers. But the key thing in our solution is that we want to control the three layers of the cake, and we want to control the interaction with our customers. We want to be the primary interface to our customers, which is what our customers want. And in return, our agents will provide them a lot of value, and we'll use the right LLM for the task, okay? But the critical thing is as we became the primary interface, and we guarantee the confidentiality of our data and their data. So the other thing we will allow with Maestro is if a customer wants to post train Maestro further, and this is another request from our customers, we will allow that. If the customers want to have it on-prem, we will allow that. If the customer wants to host on the cloud, we will allow that, okay? So it gives maximum flexibility and maximum performance by combining all 3 layers and only Cadence can provide this. And that will give without any doubt, the best results for our customers.

Your next question comes from the line of Harlan Sur from JP Morgan.

Last quarter, you mentioned the pickup in the more traditional analog power microcontroller segments of the market, right, targeting industrial, automotive, consumer IoT, that's the positive cyclical dynamics we're seeing now as excess customer inventories have bottomed in these end markets. Are you continuing to see the positive demand trends in this segment? And do these same companies are now very focused on embedding next-generation physical AI-based capabilities into their chip designs, right your MCU customers, for example, are adding NPU capabilities to their trip designs, our analog customers are designing more sensor-related capabilities into their chips more ADD/DDA converters on top DSP processing. Is this also translating into a better book of business for the Cadence team?

Yes, Harlan. I guess a great point. So I think the momentum is very strong in analog and mixed signal, and you can look at their results as well, driven by physical AI and overall recovery of the market. And of course, memory, which we work with all the big memory companies is having, of course, amazing improvement from a year or two ago. And I think what it shows up is, first of all, we are very, very strong in mix-signal companies. So because we have the most complete portfolio with both analog and digital, so these are designs like you mentioned, are more and more mixed signals. So that plays to Cadence's trend. And also ViraStack. I'm really pleased by the momentum in ViraStack because as analog customers, whether they are mixed signal companies or these big AI companies also have huge team. I don't know if you realized that some of these big AI companies designing their own IP and all have huge analyst teams. So there is a massive demand for analog productivity, which we could not provide earlier. I mean we always wanted to. But with Agentic AI, we can do a lot more in analog productivity. And you can see that in some of the customer codes we have seen and the amount of engagements we have with ViraStack. So overall, I'm pretty pleased not just with the environment of analog mixed signal, but also the new capabilities we are launching with ViraStack.

Your next question comes from the line of Joe Vruwink from Baird.

Joe Vruwink, a reminder to unmute your device. Your line is open.

Your next question comes from the line of Lee Simpson from Morgan Stanley.

Anirudh, it does seem as though you and your peer are pursuing very distinct routes to monetizing agent in chip design. I mean, I thought you answered one of the previous questions quite well. It really does bring out that you're looking to have your own control on the modeling your deployment flexibility comes through. So you can see that adoption, as you see it versus a frontier model-centric partnership is something you think will gain quite good adoption. But is there something in the performance that we need to better appreciate or even in the delivery costs that really led you to go down this route in strategy? And maybe just a follow-up question. You did mention that a leading space company was adopting some of your IP. Just wonder if there's anything specific to call out there. Are they taking just typical IP blocks, or does it come with any broader IP services attached?

Yes. Let me answer the second part first, I come back to the first one. I mean, in general, like we have mentioned that now 45% of our business is coming from the system companies, the nontraditional versus the traditional semi, so these are new customers. And what is also encouraging to see is that it's not just big AI, hyperscalers whether doing design. But we are seeing a lot of other system companies like the one we highlighted like a space company even companies in finance doing their own chips. Okay, this is -- as we make chip design easier and easier, this is always the promise that more and more companies we'll do that. And when these new companies do design, we naturally have IP opportunities but also hardware and also naturally design services because they're getting started. And we have one of the best design services capabilities as we are the best at using our own products. So I think this is some space we have to watch, but I'm pretty encouraged by more system companies doing chip design. Now coming back to the to the AI solution, we want to have control and we want to give flexibility to our customers. So this is the right way to look at it. So -- and there is a lot of value in the model itself, and it is actually pretty cost economical to post train a frontier class model. Okay, this could not have happened like 2 years ago. But now it's actually pretty manageable cost to -- and we get pretty remarkable results, like I mentioned, by very incremental cost to post train a frontier model. And of course, that's an important thing to have in the bag along with a commercially available frontier model. It's just using the best solution of the two. And then in our own genic development, right? We have our own IP team, a lot of people using our -- of course, using agent stack and all the solutions. So the benefits of costs apply to us as well. So even though there is some incremental expense to train them, there's also incremental saving in our own deployment of AI. So overall, I think it should not affect any financial model of cadence it only gets stronger as we add like a new kind of revenue capability and all the margin progression and all should though as we have planned for a while. But what Maestro does is this flexibility, but also gives customization ability, it gives control and then it gives cost savings. So this is best of all words.

Yes, Lee, if I could add there. I think that that's a very, very important point that Andrew makes in terms of the flexibility. We really want customers to have choice and control over the models they use while benefiting from cadences, debate expertise and deep integration with our design tools. I mean our strategy remains centered on delivering high-value engineering software rather than building a large general purpose AI infrastructure business. We're focused on improving the efficiency and the value of agentive workflows and maintaining attractive software economics as customer adoption scales, but we think the flexibility works in our model. Our model allows customers to use their old LLM infrastructure bear their associated bottle compute cost if they want to, a Cadence, the underlying design and orogenic orchestrate capabilities. The economics of particular deployment options could vary, of course, but the principle is that greater customer usage should create greater value for Cadence. So I think that's where we differentiate ourselves.

Your next question comes from the line of Joe Vruwink from Baird.

Can you hear me?

Yes.

Yes, Joe, please go ahead.

I want to go back to ViraStack and InnoStack. I think it's that the analog and back-end areas haven't seen the AI proliferation yet like front-end design has. Are you seeing this change is specialized models and post training, something that helps better address these domains? And if all of this is starting to happen, is there a commensurately higher revenue implication for Cadence relative to even what has been seen to date?

Joe, I mean, first of all, it there is we started with chip stack, but then we have launched like 3 other major super agents, as you know, ViraStack for analog, InnoStack for back end, and AuraStack, the latest one, for package and PCB. And I have to say the demand is actually phenomenal in mind is phenomenal, and especially because these chip stack is doing great, but also these other agents are even more integrated with our base tools. So InnoStack, if you look at our endorsement, typically, when we work with customers to get a public endorsement is only happens like 1 out of 10 customers because not everybody wants to talk about what they're doing, but we have maybe 10 endorsements this quarter. So that tells you the extent of the deployment of all these agents. And if you look at InnoStack, Renesas talking about like 34x productivity improvement, Samsung Foundry, see, there are other use cases that can happen, not just physical design. One big use case is design technology co-optimization with the foundry. So the InnoStack a critical agent for that. Socionext talked about 5x productivity improvement. [ Rapido ] talked about 2x improvement. So there's a multiple x improvement happening with InnoStack. ViraStack take already said that demand is off the charts. And AuraStack, this the latest area where innovation hasn't happened. So Nokia showed 5x productivity improvement with AuraStack. So ChipStack is probably a year now in deployment, and these are maybe 6 to 9 months, but there is acceleration deployment of these -- the newer, but they are even more tightly integrated with the base tools. And then Maestro adds further differentiation on top of these capabilities. So I feel very, very confident. And what's going to happen is we have a lot of R&D investment in our traditional tools in the middle layer, but now we are putting significant differentiation and investment in the top layer. We are defining, I believe, what will be the identic flow for the future in all industries. So as we make the Agentic, the top layer, and more and more complicated, more and more integrated, more and more differentiated, is very difficult to compete with Cadence in the Agentic AI, whether you're a start-up or if you are trying to write -- a customer tried to write some agents, but the capabilities with motor routing with domain-specific models with tight integration with agent and the tools. This is something that our customers can't do themselves. So that also bodes very well for all of these customers realizing that and working more and more closely not just in the middle layer, but in the top layer with us. So I feel very good in terms of how the speed and differentiation of our Agentic offerings.

Your next question comes from the line of Jay Vleeschhouwer from Griffin Securities.

Anirudh in your prepared remarks, you made some important references to the components of your core EDA, particularly in digital. It's demonstrable that over the last number of years, you've gained considerable share in core EDA. But I'd like to delve into what the components of that share gain have been. In other words, when we think about the various large categories that comprise core EDA, especially digital, would it be fair to say that the single largest difference in terms of the dollar and share gain you've had over the last number of years has been with Innovus in implementation and perhaps, to a lesser degree, in areas like synthesis, RTL simulation, analysis and so forth. Or perhaps you can speak about the breadth of where that incremental revenue and shares come from in core EDA to the extent that it's not really largely concentrated in banks and route.

Yes, Jay, great question. So yes, I mean, over the -- if you look at over 5, 7 years, we have gained so much in digital implementation, which I was becoming basically the platform of choice for advanced nodes design. And then what is particularly encouraging in the last 6 to 12 months is not just Innovus, but all the tools around Innovus. And we always -- I have talked about for 10 years now. Of course, there's integration at lower nodes. So that was going to happen anyway as we go to 3-nanometer, 2-nanometer, it's difficult to separate sign-off from implementation or synthesis from implementation. But then with this Agentic AI, there's a further kind of need for vertical flows, which are well integrated. So what I -- and I mentioned this last quarter also, I'm particularly pleased to see -- I mean, not only our tools and synthesis and sign-off are much, much better now versus a few years ago. that we are getting this real full flow deployment at not just small or medium customers, but some very, very large customers are deploying our full digital flow, and this is a great trend for Cadence. At the same time, we maintain our undisputed leadership in analog, okay? And then because of the strength of Palladium, strength of Jasper, strength of Verasium and Agentic Flow ChipStack, it is also helping our verification business and logic simulation. So overall, that's what I was saying, I feel that all the three big parts of EDA, we are in a very, very strong position and gaining momentum, not just individual tools being better and proven by large companies, but also this integration that advanced integration with the agentic solution. So we'll see how it is. But overall, I feel pretty good about it.

Your next question comes from the line of Gary Mobley from Stone X.

Your bookings in the somewhere in the ballpark of, I don't know, mid-20%, call it, 26% year-over-year. If I recall correctly, this was supposed to be a low renewal period 2026 was supposed to be. So my question is, what's driving the strong backlog, the strong bookings? Is it utilizing traditional EDA tools as a result of AI agent workflows. Is it a reflection of the strength of the semiconductor industry? Any color that you can add there would be helpful.

Sure. Thanks, Gary. I mean, we're very pleased with the both the bookings performance and the growth in backlog, particularly given that 2026, as you mentioned, is it comparatively light new year for us. The strength reflects the underlying desire activity, expanded customer engagement and strong add-on opportunities across our portfolio. And what we're seeing is that Cadence is becoming increasingly important to customers as they look to accelerate their engineering road maps and address more complex design challenges, backlog itself provides good visibility into that. But it's important to revert that backlog and bookings number reflects the timing of revenue recognition that depends on the structure and delivery of a digital ratio stuff like that. But -- so there's always a blade type add value in those. But strong bookings in a lighter renew here are particularly encouraging because they reflect expanding customer engagements adoptively the renewal calendar. But we're very, very pleased with that booking strength.

So Gary, needless to say, I mean, we've done three quarters. There's still one more quarter to go, which is traditionally [indiscernible]. So we'll see where we end up.

Your next question comes from the line of Lachlan Brown from Rothschild.

Are you seeing much evidence that new AI design tools like ChipStack lowering the barriers to intrigue for custom silicon for nontraditional designers. And I appreciate that land expand naturally takes time. But are there signs that new logos could become more of a meaningful growth driver for that TAM expansion going forward?

Yes, absolutely. I mean like I've talked about this for years that as we make chip design easier to do, there will be more and more people doing chip design. And if you step back, we made chip design maybe over the last 20 years, 100x easier to do because maybe 20 years ago, all these IBM Intel would do like CPU design with 500 people in like 5 years. Now you can do a CPU design 6 months with like 50 people. So that's like 100x better. And that's what led to more and more designs. And with AI, we can have and add another 5 to 10x. So it will be like 1,000x easier than probably 20, 25 years ago. And this was invariably lead to more and more design activity. And you can see that even with the hyperscalers, if you go back 1 or 2 years ago, it was not clear that the hyperscalers will do or we always get questions will they continue to do, but now with Google and Meta and Microsoft, Amazon and Tesla, everybody is doing more and more designs. Now at the same time, the traditional semiconductor companies are also doing well, of course, the big semi companies. And this is only going to accelerate, in my opinion, with -- like we talked about space companies, finance companies, physical car companies, the amount of design activity we are seeing in automotive is actually quite remarkable and robotics and things like that. And then these things, robotics also require a lot of different chips, right? The amount of chips that are there in one hand, there are like 25 chips in one hand of a humanoid robot to control all the individual motors. So I think we are -- I believe this for years, and I think you will see more and more of this. And our job is to make these things easier to use, have more -- give more flexibility to our customers, like we did with our agent solution in Maestro, so they can do more themselves.

Your final question comes from the line of Kelsey Chia Citi.

Anirudh and John, congrats on great results. My question is on EDA. EDA growth has been very strong for the last three quarters, 18%, 19% year-on-year growth well above our historical growth rate. And it sounds like a majority of that growth is primarily driven by hardware and agent attribution is still quite small. Is that a fair characterization? Or could you give us a sense of customer adoption levels for identic workflow today? And as much of the sort of revenue acceleration from Agentic AI as to ahead of us?

Yes, thanks for the question. Yes, of course, we are very proud of the hardware growth. But as you know, we will break out recurring and nonrecurring growth. So like I was mentioning earlier, and John mentioned also, our recurring growth has improved from 11% last year to 18%. So that is not hardware. That's ex hardware because hardware will typically show up in nonrecurring growth. So that's a major improvement of our recurring growth from last year to this year. And that is a combination of Agentic AI plus the growth of underlying tools and design starts on the software side.

Yes. Kelsey, I think based on lists to answer Anirudh's question there a few minutes ago that essentially, we believe that Agentic AI is a demand accelerator for EDA. I mean customers have always built their own workflows around Cadence tools. And we expect them to continue to develop their own agents as well. Our strategy accommodates that. What's exciting because that agents could perform much more designed exploration than an individual engineer, calling our underlying engines repeatedly, and that creates a potential for significantly more consumption of our core EDA tools. And I think you're starting to see that reflected in our EDA growth this year.

At this time, there are no further questions. I will now turn it back to Anirudh Devgan for closing remarks.

Thank you all for joining us this afternoon. It's an exciting time for Cadence as we finish our Q3 with AI-driven product leadership and strong business momentum. The Wall Street Journal ranked Cadence's Board #2 among all S&P 500 companies in its 2026 Best Board list. And on behalf of our employees and of our Board of Directors, we thank our customers, partners and investors for their continued trust and confidence in Cadence.

Thank you for participating in today's Cadence Third Quarter 2026 Earnings Conference Call. This concludes today's call. You may now disconnect.

Good morning, everybody. Welcome to the Goldman Sachs Communacopia Technology Conference. I'm Jim Schneider, the semiconductor analyst here at Goldman Sachs. It's my pleasure to welcome Cadence and CEO, Anirudh Devgan, to the stage today. Welcome, Anirudh. Thanks for...

Thank you. Thank you. Great to be here.

I've been asked to read a safe harbor to begin. Today's discussion will contain forward-looking statements, including Cadence's outlook on future business and operating results. Due to risks and uncertainties, actual results may differ materially from those projected or implied in today's discussion.

With that out of the way, let's get rolling.

So first question for you, maybe high level. I mean, I think Cadence has had a very strong first half of the year with double-digit growth across pretty much every product group, record backlog, two increases to full year guidance. Before we get into individual businesses, how would you characterize what's changed in customer behavior over, say, the last 12 months?

Yes. Thank you for the question. I mean the customer environment is probably the strongest I have seen. Because last few years, of course, some companies were doing phenomenally well, the big AI companies or the hyperscalers, but some of them were not. But if you look at it in '26, universally, the industry is doing great. The semi companies are doing great and then all the system companies, hyperscalers, the commitment to silicon is the strongest that I have seen because sometimes we used to get questions like a few years ago, well, will all these hyperscalers really do chips or not? But you can see now the success of. So I think what I would say at the highest level is the environment is good. And we always try to check like how long this party going to last, but it looks like party is only getting started. I have talked to all the people. They are very confident in next few years, that's number one.

Number two, I think our products are performing great. Of course, we are in the tech business. So best product always wins. And our comparative position is very strong, that's the second. And third, we have this new TAM opportunity, new expansion of agentic AI on top of our, kind of, traditional offerings. So that's all new TAM for us. So if you put it all together, these three things are what is driving this growth that you're seeing, yes.

Great. Now you framed Cadence's differentiation as a three-layer cake, especially tempting as we get closer to lunch here. But anyway; a base, that's accelerated computing data; middle layer, the physics-based simulation and optimization; top layer of AI agents. Why is that particularly relevant for EDA versus other kinds of software that's in the market today?

Yes. And I've been only saying this, like for 5 years now, I think the cake. And people say, like, what -- first of all, all things have to have three things. Answer to life is, e, right, the universal constant. It's what my adviser used to say, it's 2.7 because if it's less than 3, it's like too little. And if it's more than 3, nobody remembers anything.

A pi of more than 3.

Yes. Slightly more, 3.1. So whether it's 2.7 or 3.1, you can choose your favorite universal constant. So I think 3 -- and the reason I call it a cake, you can call it a stack, if you want or -- the reason I call it a cake is because if you eat a cake, unless you're a 2-year-old, you eat all the layers together. And you have to bake all of them together, means they interact with each other. So that's the reason to call it a cake. And the reason I put AI at the top and compute at the bottom, we can put it in because -- so first of all, the middle layer is super critical. And this is going to happen in all -- by the way, it's going to happen in all markets, not just EDA or not just chip design. You have to ground the AI with physics, especially in these kind of complicated engineering software or engineering workflows. Now in some cases, the middle layer may not exist or is maybe simple, but definitely, in our business, you need to ground the AI, the physics and then, of course, run it on compute and data. So all three are critical. And the real value will accrue to the vertical application, not the horizontals because in the beginning, it's always horizontal. In the end, it is always vertical. Like Waymo is a vertical application, for example. So if they have a AI model, do you know what model it is, doesn't matter, right? Can you get from point A to point B? They, of course, have control theory navigation, and then they have the silicon. Just to give you an example. And same thing will happen in chip design. And the reason I put agents in the top is because agents are very good at directional kind of orchestration. Like if you want to go from here to Palo Alto, that's directional thing. But actual navigation and detail, they are not as good. But they're great for orchestration, planning, optimization. So that's why the top layer calls the middle layer that sits on the compute.

Yes. Okay. Now the bear case that investors often raise with me relative to the EDA industry is that if you have a sufficiently frontier -- sufficiently capable frontier model, you could basically automate chip design from prompt and basically bypass the commercial EDA software flows. Why do you believe that's wrong? Specifically, why do you think deterministic physics-based engines and proprietary data are kind of essential, especially for leading-edge designs?

I think they're all going to be important. One thing that AI is, like, people who graduated a few years ago, they think, well, I will make a model of anything, okay? Whatever I need to know, like, what is the -- and then people who graduated 30 years ago is, it's all curve fitting. What you need to know is reality and how things work, right? Whether it's physics or mathematics or economics or whatever it is. The reality is you need both. There's no need to take a side in that. You need both. And to have a successful thing. AI itself, it's a nonlinear curve fit, right? That's what these LLMs do. If you give it input, output, it fits a nonlinear model to it. It used to be a transformer architecture. But fundamentally, they cannot do nonlinear differential equation state where -- this is mathematically not possible to do it. But together, they can provide a good combo. I do believe AI like we have seen, can provide more scenarios to optimize and then that can be optimized in the physics-based layer. So mathematically, it's not possible to do the middle layer. But we want to innovate in all three layers. We just don't want to innovate in the middle layer, which is classical physics-based. It's the combination of the three layers that will [indiscernible], and you'll see that more and more in all industries, yes.

Yes. And then why -- I mean, I guess the other question follow-on is...

And nobody is trying to do that, by the way. All the LLM companies, all the hyperscalers, they're all using our tools to design chips. Just to be clear. There's no chips being designed without using our tools, yes.

Just to push that back for a second, like, why do you need all three layers together, why can't we have somebody else's solution for the top or a bottom layer and yours for the middle?

Yes, that could happen. Yes, you could have an agent and some customers are writing some agents that call our tools and not use our agents. That could happen. What you have to remember is the top layer is a brand-new TAM opportunity for us because what the top layer used to happen, these AI agents was basically done by humans in the past, okay? So what agents are doing is they're not replacing the middle layer, they're replacing what humans used to do, okay? Now in some scenarios, agent could call our tools, but it is not that efficient. So because we wrote the middle layer, we wrote the top layer. So a lot of times, we have access to the internal that is not exposed to the user, but it will be natural for some users, especially in the beginning to write their own agent. But in the end, they realize, okay, it's more efficient for Cadence to do it. And we have like these four super agents, which are more aligned with functions. So like tools, we will have -- the middle layer, we'll have like 30, 40 products. The top layer, we have 4 super agents like front-end design, physical design, analog design and PCB and packaging. So they are integrated closely with our middle layer, and we have unique advantages. We have like 10,000 people in R&D. So they are writing both the top and middle. But even in the top layer, we don't need to get 100% of that market. Even if some of it is written by our users, or they could have like 10 agents, but the 4 big ones are by ours and 6 could be there more domain-specific. That's all fine. Even in the traditional flows, they -- a lot of customers do customization on top of our tools.

Yes. So then on agentic, how do you think about the monetization of agentic? Specifically, where do you expect to sort of drive incremental revenue above and beyond what you're already doing? Is that the new agentic workflow product themselves? And how do you think about the opportunity for higher consumption of your existing [indiscernible]?

Yes, yes, yes. It will be a combination of -- like we have a new business model for the top layer, which is consumption plus subscription. And then, of course, our existing business model for the middle layer, okay? And a good example of that is because one worry always is if something is like, let's say, 5x more efficient, then you will use 1/5 of the middle layer. This is also some perception in the market. And this is not new, even actually in, like, 2006, I launched a simulator, and it was like 10x faster. And then my marketing team was worried that people will buy like 10x less, but that never happens. That's the history of EDA. And the reason for that, there is a fundamental reason, which is different than almost all other software markets. So that's why EDA is so exciting. And sometimes we get lumped in general software. Those guys never thought we were software, and we never thought they were software. Because our software is so mathematically complex that accessing a website or a database is we don't consider that. That's just one small part of what we do, and they thought we have semiconductors or something like that, but it doesn't matter. I think what happens in this kind of application, EDA or chip design, the workload is exponential. Workload is exponential. So if you look at TSMC road map, next 5 years, they think that -- they said that chips complexity or size will go up by 48x. This is not happening in any other software market. So I talked to some customers or big hyperscalers. They think every year, they want to -- if they continue like this, they need to hire 2x more engineers. It's not sustainable. So if the workload is exponential, the requirements of headcount is exponential, you need this 5, 10x automation. If the chip size is going to be 50x bigger, there's no way they're going to hire 50x more engineers. So you need this 5 to 10x improvement to even sustain the growth. I think the customer's head count will grow, but with automation, with AI will be less than -- and this is the history. Like if you look at late '90s or early 2000s, we would design -- our customers would design a CPU. It would take them 5 years and 500 people. This is not uncommon in all these IBM, Intel, DEC, all these companies. Now you can design a CPU with 30, 40 people within 6 months. So that's 100x faster than 20 years ago. And the amount of silicon is only going up -- an amount of design activity only going up because exponentially, the size is exponential, also the applications are. So this is going to continue. If you look at the road map from Imec and all that, this kind of exponential is still projected to go until 2042, which is still, how many, 16 years at least. And by then, they will have some other technology. So this is not going to slow down, which is very unique to any other software market. So we are always looking. We are always looking at improving the efficiency of our solution, and it gets absorbed even faster. You get all the road maps around NVIDIA or Google or Apple. I mean they are -- and they are doing even more and more with that. So this is something not to be afraid of. This is something to embrace that the productivity will actually help us sell more, right?

Yes. And can you say something about sort of, like, what is -- so if we think about the monetization of it in terms of revenue terms, what is different about the agentic flow that's actually driving accelerating recurring revenue growth today versus the past things like Cerebrus, other AI features, which were maybe in your core offering, where we didn't see that kind of like acceleration in revenue?

Yes, that's a very good question. And of course, we always did a lot of good work, but what is new with this agentic AI. And we always wanted to do it. Just going back decades. We wanted to automate -- more automate the running of our tools. Our tools are fairly complex. And typically, what happens is they run for a few days, this is not like it, it doesn't run for 5 minutes, right? If you're doing some blog, it will run for a few days to do all kinds of optimization. But what the customers are doing is they run it one time and design is naturally iterative. So they have a RTL. They would change it and then they will run it again and they change it and they run it again, okay? And typically, a user would do like 3 or 4 experiments at a time because that's what typically a human would do. But if an agent is running it, first of all, agentic is much more meaningful to us than GenAI because some people said, well, GenAI has been around for 4 years, why they did not have a big impact on chip design? Because GenAI helps, like, improve the IO of the tool, you can talk to the -- look up documentation or whatever. But that's not -- okay, that's useful, but that's not shattering, okay? What is interesting in agentic AI is that you can define a workflow for a graph or, like, you do A, you do B, you do C, if you get stuck, you do -- and this is all relatively new with Claude Code and all, like, about a year ago or a little more than a year ago. So this kind of workflow, combined with our base tools, can give a lot more productivity. And then when the agent runs it, it runs, like, 100 experiments. It's not running 3 or 4 experiments. But this kind of workflow is the new thing. So that's why I'm so confident that our agentic solutions will have a real impact versus GenAI a few years ago. And Cerebrus and all were good, but now with agentic and the base, but it calls more of the base than less of the base. And then this kind of productivity, this 5, 10x productivity or at least several x is possible, will help meet the exponential demand of our customers. And this is -- the demand for all these -- we have engaged with all the top companies with all our agentic solutions. And of course, the usage of the base tools is also going up, like, we see in our results.

Yes. So -- if you think about that acceleration, sort of where do you think Cadence is getting most competitive traction today? And sort of what are the product areas represent, like, the most remaining market share opportunity for the company over the next few years?

I mean, right now, we are doing well in almost all of our products, which is great. Normally -- you always want to see that, but it doesn't happen that often. But right now, I think we are hitting in all cylinders. And we are not dependent on one critical area, but right now, all of them are firing. EDA, anyway, we have the broadest portfolio for chip design. I don't know how familiar -- we not only do digital design, we do analog, memory, mixed signal, packaging, PCB. So Cadence has always had the most complete portfolio, and it's -- and then we work closely with TSMC for a long time, with ARM for a long time and now with Intel and Samsung. So core EDA is strong as it has ever been. And then we put all the agentic on top of that, right? And I think we are definitely leading in agentic. And then hardware, which is, like, hardware acceleration, which can run things, like, 1,000x faster. We are the only company that designs our own chip actually at TSMC. If you take a look at our hardware systems, these are as complex as the latest GPU or XPU systems. So these are liquid cool, fully optically connected rack and then we have, like, a 10-, 15-year lead in designing our own. So that's hardware. And the demand for hardware is going up because, first of all, more people are designing chips, but hardware is used in proportion to the size of the chip. So if the size is going to go up by 48x in the next 5 years, so that's a systematic improvement. And then IP was the weak point of cadence historically, and I didn't invest as much in IP because it's not as profitable as EDA. But now I think, especially with AI and 3D IC, there is more opportunities in IP. So if you look at IP, our business is up 30% this year. Was up, I think, 30% last year, probably. So last 3 years, it has grown much, much higher than the market. And I feel that IP can still continue to grow well with all the Intel and Samsung and of course, TSMC. So I feel all these 3 major areas, 3 or 4 and system business is going pretty well. So we are in a good position. And the main thing is our customers are growing. So if the customers are growing, they want to do more and more innovation.

Yes. I want to get back to IP, but first to just close the loop on hardware for a second. I mean you've talked about demand being supply-constrained, I think. What's structurally driving that demand for hardware? I mean, is it the scale of the designs, which you mentioned, or is it also kind of like your customers shifting towards emulation to more of a strategic capability rather than sort of a project level thing?

Yes. I mean, one thing -- I don't know how familiar with this is like give me a few minutes to explain what the hardware systems do. I mean we call it hardware, but it's hardware plus software. People would call it like full stack basically. But basically what happens is that at this point, you cannot design any complicated chip without these systems. It's not possible. And there are multiple reasons for it. What these systems will do is even before, let's say, you're designing a chip for like 9 months or 12 months, whatever it is, 6 to 12 months typically is the design time. You want to verify the chip in your environment, whether it's a software environment, it's like Windows or CUDA or iOS or whatever. So we can have a chip behave like a chip, RTL, we can make it behave like a chip even before it comes back from TSMC or any foundry. And that is used to not only develop software, but also verify the functionality of the chip. Because if you can boot some OS on top of your chip and run your application correctly, then, of course, you know the chip is correct. And that's only possible with this kind of Palladium kind of systems. So then they become like irreplaceable. Otherwise, what will happen is you would do the design and then you would check and then if you redo the design and take few iteration, which is the old way of doing it, and only a few companies are doing -- most of them have moved to hardware-assisted design process. And the second reason they are popular is not only you can verify the chip, you can write your software because you can emulate the chip, and these are custom chips that emulate the chip like 1,000x faster than CPUs. I mean they're still slower than real life but much, much faster than anything else. So you can develop all your software. So all these system companies, the hyperscalers, they are developing chips, of course, they have software to develop. So for those two reasons, it became irreplaceable. And then the amount of hardware you buy is proportional to the size of the chip, which is going up. So one, it became irreplaceable; two, there are more chip design; three, their size of the chip going up. So it has been a record year for, I don't know, last 6 years. I don't think that's going to slow down.

Yes. Very good. IP, let's come back to that one for a second. With your market -- in terms of your market position there, you've got a very wide product breadth across a bunch of areas, including DDR, SerDes, PCI, even process or course to some extent. Maybe talk about sort of the diversity of the IP offerings and like what are the specific areas where you feel like you have most competitive advantage?

I think IP, the interesting part is, of course, we focus on lower nodes and HPC IP, which is exactly what is, of course, growing the most. Because we didn't want to do all parts of IP because it's not as profitable. And also we want to do, of course, where the [ work ] is growing. And we focus on like 5 or 6 critical pieces of IP. Some of it we developed, some of them we acquired. So like this is like the SerDes IP, the PCIe, UCIe, which is chip-to-chip HBM connection to memory DDR. So these are -- in terms of design IP, these are the critical IPs that a lot of customers want. And then the other key thing that happened is our team is much better than before. I mean, in the end, right, these are standard-based IP, so the customer will buy if the PPA is good. In the end, it's not just having the IP, just like in anything is how good is your IP. So our team is -- we -- anyway, I personally believe all the leaders should be highly technical and engineering background. So that's true for all my GMs. And I think we have -- this is one thing that has changed in the last few years. Our EDA teams is always world class, okay? Hardware teams, world class. Now our IP team is world class in terms of design capabilities. And they can also use AI to further accelerate their own. So then the output of the IPs are very competitive at TSMC and other foundries. And then the third thing that happened is these other foundries also want to get in, but we need to develop IPs for them. So whether it's Samsung, Intel, Rapidus, along with TSMC. So I think these three things, our focus is correct in terms of the market segment. Our team is much better and PPA is much better. PPA is power performance area of IPs. And then the market is naturally growing with newer foundries.

Got it. Okay. I want to move on to your last segment, system design analysis. My personal interest is like, I think, is the most interesting segment you have in terms of the evolution. You've talked about SDA-enabling companies like aerospace and defense OEMs to simulate the whole system. How different is the product strategy when you're selling to somebody like a Boeing relative to somebody like NVIDIA or AMD? I mean, do they want the same physics models, or do you have to take a fundamentally different approach to R&D for that?

No, it's similar. That's why I did it. By the way, I don't know if you know the history. I'm the one who started it in 2017 and people thought this was preposterous, like why would EDA and SDA be together because they were not together, okay? And there were multiple reasons for it. I don't know how much time I have to explain the reasons. But at the highest level, first of all, the math is very simple. R&D is very similar. And SDA is easier than EDA. Of course, the SDA guys don't like it when I say that. But EDA algorithms are much more complex than SDA algorithm. Electromagnetic is much simpler than circuit simulation. But they're in the same direction. They are also mathematical software. So all companies want to expand, but you want to expand in your core strength. So what is -- because they asked me like, okay, Anirudh, you're going to be CEO. I became president, you're going to be CEO. So what is your strategy okay? So the strategy is that go amplify your core trend. What is our core strength in Cadence or my background or all the EDA is numerical analysis, computational software. This is not -- like I said in the beginning, this is not like some database software or look up a website. This is mathematically deep like as deep as you can get. Of course, everybody thinks what they do is hard. But you can look at what we do. It's like the most difficult CS plus math plus physics. So that could be applied to two systems. And then the question is, why do you apply to systems. Because if you look at the market, so that's our core strength, mathematical software. If you look at the market, I always thought the market will evolve into these three concentric circles. Again, this is obvious now, but the silicon is in the middle than system and then data, okay? A perfect example is like a car, right, or self-driving. You have all the navigation data, then you have the car, which is mechanical plus electrical, hardware plus software and silicon that drives the car. And this is going to happen in all markets. So if you take those three concentric circles and overlay the strength of ours, which is computational software. Of course, computational software applied to silicon is EDA, chip design, EDA and IP. And that was always our core, always wanted to make sure that we are #1 in EDA because the other mistake people move is they expand into other markets but lose focus on the core market. So our always focus from the beginning is EDA should be #1. That's why over invested in EDA versus IP, even though IP is interesting now. But in EDA, we have the broadest portfolio. We are clearly the company to work with. But then if you apply computational software to systems, that's SDA. And we want to do things which are synergistic to chip design, so which is like thermal and electromagnetic analysis, which are 3D IC, which are closer. And then computational software applied to data is, of course, AI. By the way, the AI is even simpler than SDA, okay? AI people don't like that either, okay. It's just linear algebra, okay? That's like -- I took 7 courses in algebra in under grad, okay? So don't forget even grad...

You just forgot about the...

Yes, yes. So AI is just -- the algorithms in AIs are even simpler than -- but it's a good -- I mean it has a lot of application, but it is same kind of computational software applied to chip design, which is the most complex and, of course, growing exponentially, then systems and then data.

Great. Just a minute or two left, but I wanted to quickly ask you about physical AI. So how should we think about sort of physical AI being a long-term opportunity for Cadence and sort of where are companies seeing practical value in those applications today -- your capabilities today? And how should we kind of think about physical AI being in terms of magnitude of revenue contribution over time for Cadence?

Yes. I'm super excited about physical AI and have been for some time. And note that we are, of course, excited about the current trends of data center. I mean, those are huge but also physical AI will be a very big application. I mean, if you talk about the cake in the beginning, the three-layer cake, also for 5 years, talked about three slices of the cake. These are vertical slices. Because in the end, of course, the value will be vertical, right, not horizontal. So the big slice right now is data center and infrastructure. And I think we are very well positioned. We are working with all the Mag 7 like we discussed, you can see it in our results. But the other thing in strategic direction is you want to make sure you don't miss any of the other big things. One thing is you have to grow in your core strength, number one. So I'd explain like computational software. Number two, you have to grow with the market, so then chip companies are becoming system companies and AI companies, which is obvious now look what NVIDIA is doing or Broadcom and Google and Apple. And number three, you don't want to miss any big trends, okay? So we always over invest ahead of it, not too much, but always ahead of the big trend. So then what are the big trends? If the three layers are horizontal. The three vertical sites are data center first, we are very well positioned. And I believe physical AI will be huge because these are all trillion-dollar markets. See if AI is good enough to reason and talk and see -- imagine what could happen in cars and robots and drones. And these are trillions, trillions of dollars of market. And then the third slice, I always believed is science AI, which is life sciences and other deep sciences. I think what happens is people confuse that all these three are happening at the same time. And to some extent, they are, but they have a peak of their each cycle. So I think data center is in peak. I think physical AI may peak in the next 3 to 7 years. And then life sciences and all will maybe 5 to 10 years from now because that's another important thing. It's very difficult. So we want to invest in all these three slices. So we, of course, do life sciences, as you know. And physical AI, we did acquisition in Hexagon to get the best kind of middle layer for that. It's the best robotic simulator. So the opportunity for the physical AI is not just -- the AI model will be different. It will be a world model, right? If you go back to the three layers of the cake and put the physical AI slice, the top of the slice is different because it's a world model, not LLM. And there is no data for the world model. You have to do a lot more simulation. So therefore, we invested in Hexagon D&E business for simulation. But also, it will drive a lot of silicon. So our traditional business and the silicon and physical AI will be more mixed signal silicon for cars and drones, which is anyway cadence is traditional strength. And you can see that in Tesla or Rivian or BYD or Xiaomi, I mean, I just came back from China, it's amazing what's happening in Xiaomi and BYD and NIO, and they're all designing chips, they're all our customers. Same thing with the U.S. -- some of the U.S. companies like Tesla, what they're doing is remarkable, Rivian, and some of the traditional companies are because the criticism has been, oh, this is a very slow-moving market. But I think this self-driving is completely going to change that and then drones and all. So it doesn't mean that we don't love data center. Of course, we love data center, but we just want to make sure we are ready for physical AI, ready for science is AI.

Great. It's a great place to end it. Unfortunately, we're out of time. Anirudh, thanks for being here with us.

Thank you.

Amazing. Welcome back, everyone, to DB's 20th Annual Tech Conference. My name is John Marco Conti, and I'm heading the Hardware Equity Research Division here. Today, we have the pleasure of having Richard Gu, Head of Investor Relations at Cadence.

So before we start, a quick safe harbor. Today's discussion will contain forward-looking statements, including Cadence's outlook on future business and operating results. Due to risks and uncertainties, actual results may differ materially from those projected or implied in today's discussion. So Richard, now we got that out of the way.

Let's frame the time for the room. EDA for the past 30 years has been growing a few percentage points faster than R&D spendings. And then suddenly -- well, not so suddenly, I guess, in the past 3 to 5 years, we've had a lot of companies start to do custom ASICs and custom designs from hyperscalers to large system companies. So I guess what has structurally changed with [ few ] Design chips? And what role does Cadence have within it today?

Thank you for having me, Johnny. So first off, I want to take a step back and just introduce Cadence real quick for the ones newer to our stories. So Cadence is a pivotal foundational player in the semi ecosystem. We provide the semiconductor design tools to all the chip companies, semi companies and systems companies, okay? So -- and it's indispensable kind of role that we play in there.

If you look at the -- what happened in the past, I'd say, 10, 15 years, there are 2 major trends, Johnny, to your question, that's shaping the industry, okay? One is the convergence and the merge between semi and the systems, okay? Because all these semi companies are becoming like systems companies and vice versa, be it hyperscalers or autonomous driving vehicle companies or even frontier and other model companies that designing their own ASIC chips now, okay, which is a great thing to see.

Because what it means for us is not only the aperture has expanded dramatically in terms of the new entrants and the new customers and design starts, which is always a great tailwind for our business, but it also means increasing compounding complexity for those designs. So if you put together these 2 dimensions, it's a fantastic tailwind for the company for the next, I'd say, 10, 15 years, unabated, okay, first.

So -- but the second trend, I'd say, is the AI obviously is a turbocharger for the entire semi ecosystem. And Cadence is a structural winner throughout this entire process. Not only are we supporting and supplying the EDA, IP, hardware systems and system simulation software to all the key players to design their AI accelerators. But also we're applying AI to our own tools to make sure our customers can reap the benefits of the massive boost in productivities and they can design better chips too.

So I think with those 2 together, we're seeing a very strong tailwind for the business. Our most recent Q2 results is a reflection of that, right? You see clearly all the semi companies and systems companies are doubling down in terms of innovation road maps and R&D spend continue to grow, which is a great leading indicator for our business. And in the meantime, I'd say our business is accelerating. We're growing this year at a clip of 19% with 44.25% of kind of non-GAAP op margin. So when we talk about the Rule of 40, this is -- we're going to surpass Rule of 60 this year. So it's a great business, and the Cadence is well positioned to tap into a long-term growth.

Yes. So clearly, it's showing. Maybe we'll just unpack a little bit of that AI developments of Agentic. You acquired ChipStack last November. Within 3 months, you shipped the ChipStack AI Super Agent, which is the industry's first Agentic workflow for front-end design verification. So for those in the room that have not tracked EDA closely, what does it actually mean for an AI agent to design and verify a part of the chip? And what parts of the chip design process can tackle into the next Agentic race?

Great question. So we're very excited about ChipStack and also the other 3 super agents we launched [indiscernible] over the past couple of months, which literally straddles the entire spectrum of the chip design in the back end also, including ViraStack, which is the analog design kind of full flow orchestrator and also InnoStack, which runs the digital flow and RaStack, which runs the packaging, okay?

So now with those super agents, what we can help our customer achieve accomplish is the dramatic improvement in productivity. When we think about the design challenges for our customers, everybody is faced with a big mismatch in terms of what they try to accomplish in their innovation road map and the supply side of the equation in terms of how many designers they can have, okay? And the workload is increasing unabatedly for the next 5, 6 years to the tune of even 30, 40x.

So it's absolutely impossible for any company to hire that many engineers. Hence, the automation, EDA, AI needs to do a heavy lifting and bridge the gap. So that's a massive opportunity. What it means is these super agents, they all like will be endowed and trained with a certain human designer skill, be it front-end design -- take CHIPS Act as an example, right?

So it will be doing the RTL code generation, translating the design spec to the machine code and also create test benches. And what it does is they also invoke and call a lot of the underlying EDA tools, including simulation, verification, which is a constant kind of iteration and looping process.

So what it does is it's going to free up the human designers to a higher level and it allows them to do a lot more designs to be a lot more productive. Even Jensen talked about during the [indiscernible] about 2 months ago, the CHIPS Act, they are seeing 40x productivity benefit. So the opportunity is massive.

I think importantly, Johnny, to keep in mind is also the -- when it comes to the R&D spend in the design realm, right, the EDA spend right now in terms of wallet share, it's still like low teens, call it, 10%, 11%. So the massive 90% of the spend is still in human designers. So we definitely see this irrevocable trend in terms of that wallet share will continue to shift more and more towards tools and automation, which bodes well for our business in the long term.

Yes. So clear -- there's some clear productivity advantages here, right? So maybe speaking a little bit about that. If, say, you have 10x productivity, how does Cadence capture -- commercially, how do you capture a fair share of that value?

And how do we think about what Agentic AI does to a business model that has historically been built around the mix of seats and project-based R&D? Like could the agent stack open the door to those companies that don't have a team of chip designers like hyperscalers, but still wish to do custom designs?

Great question. So the -- from a monetization standpoint, the way we're going to monetize the super agents is through 3 vectors, okay? So first off, those agents, they are human designer surrogation, okay? So we're tapping into the greenfield, okay? This is a complete greenfield for us. What we're going to do is we're going to create -- we're creating separate price books for these 4 super agents.

And in terms of the pricing, it's all going to be commensurate to what a human designer skills could be. So it will be worth tens of thousands of dollars. And once the customer exceeds or surpass the prescribed workload within that super agent, obviously, we want to charge them additional consumption in terms of tokens and extra usage.

And another great avenue, the third avenue for the monetization is the calling and invoking of the underlying tools, okay? And you can imagine these agents, they're not humans, right? They don't need like ARC, like you and me. So they'll be able to kind of explore in a much thorough and bigger fashion than a human designer could possibly do.

So what it means is it's going to be a lot more base to usage, which has come through in our typical traditional EDA model, EDA monetization model. To the second point of your question on the -- what does it do for newer entrants. I think it definitely -- it levels the playing field, right?

Because now with a smaller team, you could do amazing things, right, by leveraging these tools, as long as you have a clear mind in terms of what kind of chip you're going to have, what kind of system you're going to have and then you can leverage the tools, I think there are different business models existing in place already.

Anirudh talked about the 4-story beauty, like going from merchandise to ASIC to hybrid COT to COT. That typically is going good on that path. I think the more companies -- more customers get straddled around that 4-story building, the better opportunity will be for Cadence.

That makes sense. So okay, the bottleneck in AI systems has been moved from the transistor to the system, data movement, memory bandwidth, packaging and thermals, right? So your fastest-growing segment in recent quarters have been FDA, which is the simulation piece and IP rather than the classic EDA. So I guess my question is, is it fair to say that Cadence's growth is now tied to system complexity rather than chip unit growth? And what does it mean for how investors should size the market?

Sure. So the business is actually -- we're seeing broad-based strength, right? If you look at the most recent quarter, we grew -- the revenue is growing like 24%, okay? And the core EDA is growing 18% to 19% and SD&A growing at about like 35% -- north of 35%, IP growing north of 40%, okay? So these are fantastic numbers to see.

So I mean, using the analogy of a chariot, pulled by multiple like 3 or 4 horses. I'd say all the horses are already at top speed, which is great to see. The Codea is always a great linchpin in terms of like 70% of our business is in Codea, right? It's great to have that kind of growth. I think in general, I think if you look at the -- our business is -- workload is important. So the driver of the revenue, workload is always important, okay?

And one unique aspect for our business is our workload is not static, okay? It is growing exponentially. If you think about the complexity of the chip design, the most complex chip these days is, call it, Blackwell or what it is, like [ 10 billion ] transistors. But it's -- we fully expect that to grow. It's going to grow to like 1 trillion in a matter of 5 to 6 years, okay? If you weave into complexity, it's about 30, 40x kind of workload increase in the foreseeable future. So that will be the ultimate driver for our business.

And it will be coming through in both the workload growth and the pricing opportunities, which is still an opportunity for us to flex further. I think Agentic AI just give us so much more in terms of growth levers. But the business is so well positioned that we have multiple irons in the fire. And it's kind of a 4, 5 cylinder engine. All the engines are running well.

It's almost like it's like additive, right? Like any layer that you're able to capture onto that Agentic layer, it's kind of like net new for you guys, right?

Absolutely. The middle layer, I mean we use the analogy of 3-layer cake, right? -- middle layer for the core principal software, hardware, IP, these are unassailable, okay? -- irreplaceable unassailable. So AI is a great overlay on top of that. It's going to orchestrate and help customers reap massive productivity benefit.

But what it does is not only gives us the opportunity to tap into that greenfield opportunity, but also it's going to drive a lot of tool usage in the middle layer. Now we can optimize that with data and the chips and the systems. So I think it's a beautiful 3-layered kind of stack that we're going to continue to leverage and grow.

One can kind of also put the comparison with how Cerebrus was pulling from the back end, the multiple licenses of Innovus, right? It's almost like same parallelism when you think about how a new product can pull legacy tools that are required, that's like the engine behind -- so sounds like a great opportunity.

Yes. I think that's -- it's definitely [ apt ] kind of parallel. Cerebrus. I mean, you're familiar with that. One copy of Cerebrus can drive 10 copies of the full flow digital standpoint. So there's lots of pull-through. I think Agentic AI is not a big opportunity for us to drive an abstract even further up.

That's fair. Okay. So maybe we should unwrap some of the IP developments. You recently displayed wins in the IP business with Cerebrus, LPDDR6, PCIe, UCIe. For the investors in the room, how should they think about the IP developments, the key areas of the portfolio where you're seeing substantial market demand? And how do you juggle basically a higher IP mix but also wanting to keep a pretty steady margin progression, right? Because IP is not as strong as EDA margins, so.

So IP for us is a great business, right? It's certainly situated and positioned in a place where we're seeing like great secular trend and growth trend drivers. But for IP, for us, it's always a balance, right, a balancing act. It's a conversation between the revenue growth and the margin kind of accretion also, okay?

So we -- early on, we devised an IP strategy that we're not going to be everything for everyone, okay? We chose very deliberately to focus on the advanced nodes IP designs, IP titles, HBM, UCIe, PCIe, all these connectivity kind of important kind of IP. That is very much exposed to the AI super kind of cycle, okay? So that is bearing fruits.

You have certainly seen IPs growing at a very fast clip, okay? We're gaining share in the market. We're going to be at a $1 billion clip by the end of the year, so at scale and growing at much faster than market, which is great to see, right? And also, I'd say the foundry ecosystem is helpful, right?

Now it's not just TSMC, Intel, it's Samsung, it's Rapidus. So we're working with them all, okay? So that gives us a great opportunity to continue to tap into that growth engine. But in the meantime, we don't want to be everything to everyone.

We want to make sure like if -- I think if done right and managed right, we have the opportunity to strike that goldilocks in terms of tapping into that high IP growth, but not sacrificing the overarching company margin, okay? Because the margin growth and the EPS growth is always a North Star for us.

So I think $1 billion is a great place to be. And we'll continue to work with the customers and make sure that they're delighted with our products. And the products is getting a lot better, too. I think now we're in a good place that we can really grow well in general, but at the same time, continue to maintain and have and achieving that 50% incremental margin in general for the company.

Yes. So maybe just following on that question on the IP. I'm curious about what your thoughts on -- obviously, we've seen OpenAI coming out with [indiscernible] and that's all like debate about whether you can possibly expedite substantially the tape-out process and the design process of the chips.

So I guess my question here is on the IP side, could you see a future into which pockets of the IP portfolio get a little bit more commoditized. And so you have a bit more of a software layer allowing customers to just churn out faster and better IP. And so maybe the -- I guess, that will be reflected into the TAM of the IP market thoughts.

Yes. So the IP is a great business, but EDA is -- I mean IP is a good business, but EDA is a great business. Because for IP, there's -- the conversation is always the build versus buy. I think the market is so conducive now. It's almost like all the -- I mean, on every customer's mind, the main objective is they want to win the race, right?

They want to go to market a lot faster with a great product that can go to production, okay? Hence, I think IP is going very well. But EDA is a fantastic business because EDA can only buy, you cannot build, okay?

So that's why I think IP will give us good growth if we do it right, I do feel like IP, ultimately, you have to make sure you have the product excellence, right? Because the measurement of successful IP is you have to deliver the PPA benefit to the customers.

As long as you can do that, I think you can continue to have great growth. But over time, that's why, like I said, we need all the horses, all the engines to run well for the business, and EDA is a great business. We should never lose sight and take eyes off that. I think overall, we're managing the business in totality as a portfolio.

I think IP has a lot of growth to be had in the coming years, given what I talked about the AI super trend. Given I talked about the sort of the foundry ecosystem build-out. And also with Intel, we're doing a lot more, right? 14A, I think you probably noted that we signed a meaningful kind of deal with Intel to help them design their 14A on the foundry side. So it's not just for IP, but also it helps with our EDA tools also and Agentic AI products. So I think ultimately, I think we're -- and the company is just firing on all cylinders. We're sitting in a great place.

So it's safe to say that maybe because EDA is a greater business, perhaps that portion is a bit more shielded by any developments of in-housing software to basically replicate the motions of EDA, right? Because as we know, some of the biggest challenges in chip design is verification, right? And it's a problem with which today still requires enormous amounts of efforts to really reduce all the errors prior to tapeout. So would it be still fair to say that visibility in the next 2 to 3 years as far as EDA comes is still like a safe software business and sort of shielded from the fast [indiscernible] way?

Good question. So I think the EDA is unassailable. The position of EDA is unassailable and impracticable. The reason being that EDA is all deterministic, right? It's a physics-based kind of -- it has to be physically accurate, okay? You don't want to take any chances with any of the probabilistic stuff in there at all, okay?

So I think EDA is a great place to be. We can see that the reliance from our customers on EDA to help them deliver against their innovation road map is going to become a lot more acute and then a lot more pronounced in the coming years.

Just given all what we try to do and given the shortages from the labor side. I think it's a fantastic opportunity and tailwind for our business in the long term, as [indiscernible] said. And I think also -- Johnny, I think I want to highlight from an innovation standpoint, we want to innovate on all 3 layers, right? Core EDA, unassailable position, Agentic AI, if we can do things right by embedding that and co-optimize that with the core tools and then you build it on top of that proprietary good data, data set, data moat and then great systems to build on top of that. I think it's a fantastic 3-layer cake supporting the long-term growth of the story.

That's very fair. Okay. So maybe we'll shift a little bit on the geography side. China used to be a big point of contention, right, in the past. Now we're seeing healthy growth again as a mix of competitive displacements and the future politics helping capture back some of that lost growth.

Knowing what you know and given how much China is pouring into the development of new fabs to startups challenging incumbent architectures, what is your view on how will China fit into Cadence's growth equation, say, 3 years from now? Like is it a constant race towards wanting better emulators or perhaps IP is taking more off? Just walk us through the motions there.

China is a good market, I'd say. But I think if you look at our business, the rest of the world is growing very nicely, too, okay? So it's fairly balanced. I'd say the broad-based strength is not -- does not just apply to the product set portfolio, but also apply for all the geos and regions.

I think we're also pretty confident that China is going to grow at least at the company average this year, okay? Because if you look at the dynamics in the China market, it actually mirrors a lot of what's happening in the U.S., right? They have great LLM model companies. They have good hyperscalers, a lot of EV kind of autonomous driving vehicle companies.

We work with a lot of those. I think the strength is across the board. And not only do we sell a lot of emulation system to the Chinese market, but EDA is a big part of that, too. So I think it's a reflection of the excellence for the product set and the tight relationship with the customers.

So we do feel like China, I think the strength in Q2 really is a reflection of a lot of the bookings and add-on deals we had for the past couple of quarters. So it just come through. But I think it's a good market. We'll keep a close eye on it. And -- but I think overall, the company is growing in a very balanced fashion.

What about competition locally? Like we used to have a debate maybe like 3, 4 years ago about the Univistas Imperion of the world. How has that shaped up recently? And firstly. And then secondly, I guess, with your closest competitor, how are market share shifts happening in what product segments? Like are you seeing that those bookings push more on the hardware side? Or I'm just curious to see the market share dynamics happening in the region.

Sure. Just by sheer growth rate, you can tell we're gaining share in the market. So -- and I think the -- it's part of that is just -- is really driven by the product excellence across the Board, right? But I think from a local competition standpoint, it's not our concern at this point, because I think a lot of the local competition, they're still a lot smaller.

They have some point tools, a lot more -- just not up to our standard, and they don't have full flow. And a very important consideration is also they don't have the foundry ecosystem or certification from TSMC -- so it's not -- I think we'll keep an eye on those, but it's not a near-term or medium-term threat for us, okay?

I think versus our peer company, we feel very confident. I think our growth rate speaks volume in terms of our market position in China. Again, it's not just emulation systems, it's EDA tools. IP, we still have a lot of room to grow in there, too. So I think overall, it's a great business. And we keep an eye on China. But again, like overall, all the geos are doing well. Yes. That's fair.

Okay. Maybe we should unwrap a little bit of that $8.1 billion backlog. Can you walk us through what visibility do you have today in what areas of your 3 businesses? And where do you think there's still an opportunity to add more?

Is it like physical AI simulation, more integration of multiphysics flow with EDA? Or are you eyeing up maybe bolt-ons in other areas? I mean, Anirudh used to say robotics was like -- physically, that was like a very big opportunity for you guys in the distant future. So I'm curious to see out of the backlog, the mix between visibility and have like long-term targets.

Yes. So we're very pleased with the $8.1 billion record backlog exiting Q2, okay? So that was accomplished sequentially in 2 seasonally down quarter from a booking standpoint, okay? So what it means is it's a reflection of how strong the underlying business is, especially those add-on like AMC business for us.

So again, it's a reflection of how strong and broad and deepen these relationships are with our top customers, which is the who's who of the world, okay? The quality, Johnny, the backlog is amazing, too. If you look at the $8.1 billion backlog, we look at the cRPO coverage ratio as a percentage to RPO.

Our ratio in Q2 is about 58%. It's much higher than the peer set, okay? It's a great thing to do to have because it gives you clear visibility in terms of how much of that is going to translate into revenue in the next 12 months, right? So it's a great thing to have. I'd say from a visibility standpoint, EDA business, we have a great visibility. Our contract cycle typically runs for 2.5 to 3 years.

So software is all ratably kind of recognized. We have great visibility in there. Hardware is more like -- it's a pipeline business. So it's 6 months, we kind of look at 6 months out. That's why we kind of -- we'll update the guide every kind of 6 months when look into that. And system design and simulation, like you asked, is a great business. We're now like a quarter past the acquisition, the closing of the Hexagon business.

Now we brought it under one roof with BETA CAE. About 2 years ago, we acquired that business. We try to create one like full flow when it comes to physical structural designs, which is going to tap into the next leg of growth for physical AI, what Anirudh has been talking about.

So we feel very good. I think if you look at the SD&A business, we are squarely entrenched and focused on 2 bookends, okay, which is like high growth, high margin, but also like very much attuned to the Moore's Law, so one is closer to the silicon side of the equation, just like the packaging, 3D-IC. Another one is the physical edge like we just touched upon. So I think overall, the $8.1 billion kind of backlog is a great thing to have. We'll continue to kind of drive the business forward with our customers in the next coming quarters or coming years.

So just piggybacking on that on the Hexagon acquisition, are you guys on track with what was planned in terms of both integration of human capital and the tools? Or like how far are we between the full -- this is exactly where we want to be when it comes to the technologies merging together?

Yes, we feel very good. It's tracking well against our expectations. Again, like we're focused on like creating the full flow when it comes to physical AI and structural kind of designs. So everything is tracking there.

Good. Okay. So Again, on multiphysics, I want to just double down. You built the portfolio organically and then bolt on some acquisitions. Your largest competitor spend roughly [ $35 billion ] buying its way to the same conclusion. So my question is for you, Richard, what does the integrated electrical thermal fluid and structural flow unlock that a single physics tool never could? And a follow-up on that is perhaps any updates on Millennium platform. Where you find bias? How is that -- how is Millennium progressing relative to per se, you emulate for your [ pro ] fibers?

Okay. Johnny, so I look at it that way, right? If you have a great -- if you want a great set of suits like you're wear right now, you don't need to buy the entire department store to get it, right? So that tends to be our philosophy when it comes to building the business.

Anirudh saw this opportunity about 10 years ago. In terms of the merging convergence between system design and the chip designs, okay? We've been like building the business step by step, starting from more organically from the front element analysis to CFD to electromagnetic towards the end to the structural, right? I think the business has been growing well.

Like I said, we are focused on the 2 capstone areas in the SD&A, right? Some of the -- a lot of the kind of system design analysis, like if you want the simulation software to design the swimming pool is not where our interest is, okay? We want to be in the areas which is most compute-intensive and which can tap into our computational software kind of capabilities, okay?

So that's our focus areas. On Millennium, it's gone quite well. We launched the tool early on, it was like about 1.5 years, 2 years ago, together with NVIDIA, right? Because what it does is, again, it's that 3-layer cake kind of being applied in different areas, right? In a sense, it's like it's a 3-layer cake, right? In the bottom layer is the GPU from NVIDIA, right? And the middle layer is all the principal software in simulation.

We started with CFD, right? And now I think we're applying that to different areas, even for EA for Clarity and Celsius when it comes to electromagnetic and the thermal simulation. So it is working well with the customers. We continue to engage with customers, but that's a beautiful business model. I think the cadence bakery is we'll continue to come up with all kinds of different flavors of those cakes.

So are you seeing customers come back to you and say, we've actually like improved the workflow by using Millennium. We've actually -- because like I guess how I'm thinking about it is, at the end of the day, it is a computer. It is a very powerful computer that allows you to do some deep level maths and deep level simulation, right? So I'm curious to see if there's any customers which actually came back and said, yes, we've actually seen real improvements into our flows.

Yes. So we talk about -- obviously, we have deep symbiotic relationship with big customers, right, including NVIDIA, right? So NVIDIA kind of publicly endorsed Millennium early on, right? I think we talked about the productivity of 50 to 60x. Because when it comes to these systems, you have to have big leap forward in order to justify the systems, right?

So yes, we're seeing those, and we're now expanding that further, like I said, applying the different kind of algorithms or solvers by coupling that tightly with the GPU and underlying kind of infrastructure, accelerated compute platforms to do things. That area -- that product could be widely applied also in other sectors, right, aerospace and defense, automotive. So the opportunity is certainly grow there, yes.

Interesting. So maybe I would like to go back on the R&D split that you mentioned before. We've known that for many years, the R&D budget of a semi company was 90% people and 10% tools. I believe there was -- I think it was John a few years ago that sort of gave us some trajectory for the path forward potentially going to 15%, 20% software.

And obviously, we also know that the supply of software and hardware engineers combined together, it's getting fewer and AI is pushing that even further. So I guess my question to you is how do you see the trajectory of R&D go from people to software? How -- like how -- when are we going to get to a point where we're going to reach 20% or 25% of the split being in software, if at all.

So I think again, like when we talk to customers, okay? All the customers, their main focus these days is they try to deliver against the road maps, okay? And the arms race in AI is intense, right? It's unremitting, it's intense. They have a lot of designs they want to accomplish, a lot of different flavors of designs and the design is getting increasingly more complex, okay?

And the complexity, again, is our friend, right? It's compounding, okay? So what it means is, again, I talked about the mismatch between what they try to get to, the workload increase of 30, 40x versus what they have in terms of bottleneck, in terms of engineering resources. So the gap in between needs to be filled with automation, EDA and AI, which is happening right now as we speak, okay?

And there are customers of ours telling us they're willing to spend more than 50% of what we spent on a human being, human designer on automation, tokens and chip designs. So what it means is if you do the math, it's almost like 33% of the R&D budget, right? So I think it kind of -- again, it gives you a flavor in terms of where things are headed.

But I think if you look at -- if you draw a long line, look at the arc of where things are headed in terms of the design intensity going unremittingly continue to advance in a lot more designs. Now it's not just the traditional semi companies, right?

It's hyperscalers, car companies and other model companies doing their own designs. So the opportunity is massive. And so I think if you put all these things together, taken all together, it's a phenomenal kind of long-term tailwind for the business in the long term.

Yes. That makes sense. Maybe let's talk about the long view. I believe Anirudh said that the company's competitive position has never been better. Obviously, we've spoken about the 3 layers, the agents on top, the tools in the middle, hardware underneath and they're all reinforcing each other, right? So the long view is it's 2031 Agentic design has matured.

Physical AI is real. some companies designed most of the world's leading silicon. So what does Cadence look like? And what is the one thing you would tell this room to watch perhaps over the 12 to 18 months that we can track to see that we are on track to get there? Yes. It's a hard question. It's a very long question. And I feel like today in AI's world, it's -- a year feels like forever, but it's good to get. Does that make sense?

Yes, sure, Johnny. I think it's kind of all of the above. But I think one thing we're closely watching is the recurring revenue growth. I think because recurring -- if you think about our model, right, Cadence has been -- always been a great compounder, okay?

So regardless of where the market is, like the volume-driven like business up and down going through cycles and things, the Cadence business model is always a very smooth upward trend, growing at a very steady pace with great margin and great kind of cash flow and share buyback kind of program in there, too. So I think a lot of our business, even the AI kind of business is because of flow through the -- I mean, subscription plus consumption is going to flow through our business through the recurring kind of metric.

So I'll probably keep a close eye on that. I think there's one thing that's undoubtedly on our minds that will be true is the reliance and dependency on EDA companies like Cadence will be a lot more pronounced in the next 5 to 10 years, okay? It will be increasingly more, okay? Just because if you look at what these companies try to accomplish versus what they have, I think we are there to support them. And Cadence is a great player.

We're in the midst of everything. And our customers is the marquee most is who's of the world, right? We want to make sure we have a clear kind of vision in terms of where we want to be, supporting them on their journey and they'll continue to execute well. I think with our CEO, Anirudh, at the helm, I think we are fully confident we can support them in the -- on their journey, be it AI or next chapter. So I think it's a fantastic business and we feel good about where things are.

Yes. So that EDA stickiness is basically what gives you certainty or not certainty, but a good amount of visibility to be able to keep up that profitable growth strategy, right? Because right now, you're tracking, as you've mentioned, more than 50% margins and incremental margins. So the strategy is to keep on track to that.

And I guess many investors will come to me and ask about what happens in a world where you have a substantial slowdown in R&D budgets and what happens when -- typically, when you think about the AI infrastructure world right now, CapEx is driving everything, right? And R&D is not really a big point of focus.

But then, of course, EDA tracks R&D. And so would it be fair to say that because of how sticky EDA is to the customers and because of how less volatile the R&D budget is for each of these semi companies, if there is a down cycle, which one I'm not saying there is, but if there is a down cycle, there is a level of bottom almost that Cadence can have with regards to revenue.

Yes. I mean it's a good question, right? You just need to look at the history, right? Semis and our customers will go through cycles. And it's -- I mean cyclicality is kind of the nature of the [ beast ] in a way, right? I think AI could be different.

And -- but one thing is for sure, I think even when companies and our customers go through this up cycle, down cycle, our business, if you draw a line for the past 10, 15, 20 years, the Cadence growth is a very smooth kind of growth as far as revenue, margin continue to expand, EPS will outpace the revenue growth.

So I think we are much more insulated from the volume side of the equation because R&D typically is the most sacred, right, most insulated piece of the spend for the customers regardless where they are. Even in a down cycle, they want to make sure they spend and invest in the right places so they can emerge stronger, right? So I think time and again, it's been proven, and I don't think this time is any different. Hence, I think it's -- this is a phenomenal, great business to be having there.

That's very fair. Maybe just by concluding because I think we only have a couple of minutes left. What is the market getting wrong about this whole EDA debate and perhaps the fear about AI CapEx eventually climaxing and stopping and then spend trickling -- basically spend ceasing and potentially coming into lower chip starts, which would then trickle down into lower EDA spend. So I'm curious to hear your view about -- what is the market getting wrong across the Board about AI specifically?

No, I think the market is the market. And I mean, for us, the most important thing is we continue to -- I mean, we know we have a very, very crisp strong strategy, right? And then like we have strong leadership, great team around that, and we'll continue to execute. And the market will determine where things are headed. But I think one thing is for sure, I think Cadence will be an AI beneficiary and winner regardless of where the ecosystem that the customers will end up being, okay?

Because ultimately, I mean, our business is not driven by volume, right? It's driven by design starts and design complexities. And those 2 things will grow and advance unremittently in the foreseeable future, 5, 10, 15 years. That's ultimately our North Star. I think again, our business, I think if you look at margin, revenue, EPS, cash flow, we're in a great, fantastic place.

I guess people forget that you are investing a substantial amount every year into R&D, right? I think it's close to 30%. And so if anything, if there's any development in AI, Cadence would be probably at least ahead of the curve or on par with the latest start of trying to do the frontier development. So I guess we can make an argument that Cadence would know what is happening ahead of most people because you're sitting in the room, right, with the leaders of the semi companies.

Absolutely. Because all the relationships and the partnerships with all the top customers, key customers are all expanding broadening and deepening, right? Just when you look at the financials and the numbers, it's all trending in the right direction. Absolutely. We feel very good about where the business is headed.

I think we're out of time. Thank you so much, Richard.

Thank you, Johnny. Thanks for having me.

Ladies and gentlemen, good afternoon. My name is Abby, and I will be your conference operator today. At this time, I would like to welcome everyone to the Cadence Second Quarter 2026 Earnings Conference Call. [Operator Instructions] Thank you. And I will now turn the call over to Richard Gu, Vice President of Investor Relations for Cadence. Please go ahead.

Thank you, operator. I would like to welcome everyone to our second quarter of 2026 earnings conference call. I'm joined today by Anirudh Devgan, President and Chief Executive Officer; and John Wall, Senior Vice President and Chief Financial Officer. The webcast of this call and a copy of today's prepared remarks will be available on our website, cadence.com. Today's discussion will contain forward-looking statements, including our outlook on future business and operating results.

Due to risks and uncertainties, actual results may differ materially from those projected or implied in today's discussion. For information on factors that could cause actual results to differ, please refer to our SEC filings, including our most recent Forms 10-K and 10-Q, CFO commentary in today's earnings release.

All forward-looking statements during this call are based on estimates and information available to us as of today, and we disclaim any obligation to update them. In addition, all financial measures discussed on this call are non-GAAP unless otherwise specified. The non-GAAP measures should not be considered in isolation from or as a substitute for GAAP results.

Reconciliations of GAAP to non-GAAP measures are included in today's earnings release. For the Q&A session today, we would ask that you observe a limit of 1 question only. If time permits, you can requeue with additional questions. Now I'll turn the call over to Anirudh.

Thank you, Richard. Good afternoon, everyone, and thank you for joining us today. I'm very pleased to report that Cadence delivered outstanding financial results for the second quarter of 2026, with all key metrics exceeding our guidance. We exited the quarter with record backlog that was above our expectations. We are seeing growing demand for our AI-driven solutions across our expanding customer base. .

AI transformation is driving strong, broad-based performance across both design for AI and AI for design fronts. Given the growing business momentum, and accelerating demand, we are raising our guidance for the year to 19% revenue growth and with higher profitability as we become more central to our customers as a strategic and trusted partner.

John will provide more details on both our Q2 results and the updated financial outlook. Let me start with the overall environment. Design activity is growing as AI drives exponential design complexity and a new generation of system architectures spanning hyperscaler infrastructure and physical AI. Customers are investing aggressively in these opportunities led by AI and HPC, and we are also seeing continued signs of improvement across the more traditional analog and consumer verticals. Chip and system design present demanding engineering challenges that required deterministic, physics-based engines, proprietary silicon correlated data and deep design knowledge.

Our 3-layer cake framework uniquely brings these capabilities together with accelerated compute and data at the bottom layer, physically accurate simulation and optimization solvers in the middle layer and AI agents and orchestration at the top layer. Agentic AI is a demand accelerator for Cadence as autonomous agents expand the design exploration space and call our underlying physically accurate engines more often, creating a durable tailwind that represents a significant long-term TAM expansion opportunity.

We extended our leadership in agentic AI with AuraStack AI super agent, delivering up to 15x higher productivity and 2x faster time to market for PCB and advanced packaging design. Cadence is now the only provider with agentic solutions spanning the full electronic system design flow from digital analog design and verification to advanced packaging and PCB. We see strong early traction across our AI super agent portfolio with initial customer results demonstrating meaningful productivity improvement and better design outcomes.

Our ChipStack AI super agent, enabling higher verification productivity and faster design cycle has more than 20 customer engagements and is already deployed in production across multiple chip designs. At Computex 2026, together with NVIDIA, we introduced the industry's first fully autonomous virtual AI design engineer, extending ChipStack to even higher levels of autonomy.

Early customer results include more than 40x faster RTL validation, reducing a typical 5-week verification cycle to less than a day on a state-of-the-art advanced node design. In analog and custom design, ViraStack is seeing strong customer interest with more than 25 customer engagement, achieving 2x to 10x productivity improvement compared to traditional design flows.

InnoStack is also building momentum as customers adopt agentic AI for advanced node SoC design. During the quarter, Rapidus announced a collaboration to integrate the Cadence Innostack AI super agent into its AI agent design solution, targeting up to a 2x faster design turnaround. We continue to deepen our strategic partnerships across the ecosystem. We expanded our collaboration with Intel through a multiyear engagement focus on enabling its 14A process, leveraging our design IP and agentic AI-based EDA to co-optimize tool, flows and methodologies for next-generation HPC and mobile designs.

This agreement is expected to be a meaningful driver of growth over the next few years. We also deepened our collaboration with Samsung Foundry on 2-nanometer and 3DIC technologies, combining our AI-driven flows and design IP to enable next-generation AI, HPC and mobile systems. Now turning to our businesses. We are pleased that all product groups delivered double-digit year-over-year growth.

Our IP business had an outstanding quarter, growing over 40% year-over-year. AI performance is increasingly constrained by data movement, memory bandwidth and advanced packaging. And our differentiated IP portfolio continued to see strong adoption. This was reflected in the strong demand for our Star IP portfolio in AI and HPC applications, including PCIe, UCIe, HBM and LPDDR6. We also expanded engagement with leading memory semiconductor and aerospace customers. We secured our first ever Tensilica DSP design win with ST Microelectronics, reinforcing our strength in automotive and audio applications.

Core EDA grew 18% year-over-year, driven by growing adoption of our AI solutions, proliferation of our digital full flow solutions continued, and we saw expanded adoption of Tempus and Certus sign-off tools on leading-edge designs with wins across hyperscalers, top semiconductor companies and startups. We also expanded our implementation and signoff footprint at Frontier AI companies as well as at a marquee ASIC silicon vendor, underscoring their differentiated value in enabling the industry's most advanced design.

In analog, we had a significant competitive win with Spectra at a leading semiconductor supplier and our FastSPICE simulators, Spectre FX brought several production wins at leading customers. Our hardware business delivered another record quarter, driven by continued strength in Palladium Z3 and Protium X3. As designs approach unprecedented scale, hardware-assisted design and verification is becoming a strategic capacity layer for our customers' AI road map.

These customers are designing now the most complex chips and systems in the world, and they critically depend on our scalable, high-performance hardware platforms to realize their designs. Demand remains especially strong from AI and HPC customers, including hyperscalers and leading semiconductor companies. We added [ 12 new logos ] and saw a meaningful expansion with several marquee AI customers as well as a notable competitive win with a major AI infrastructure provider.

System Design and analysis revenue grew 37% year-over-year. As AI system complexity increases, Customers are increasingly turning to our advanced packaging and PCB solutions. Allegro X AI was adopted by several customers, driven by significant layout design time reduction. With our 3D IC technology in collaboration with TSMC's 3D Fabric advanced packaging solutions, we are enabling customers to confidently design cutting-edge silicon for increasingly demanding AI workloads.

In structural simulation, our beta CAE business had several competitive displacements. While the integration of recently acquired Hexagon D&E business is progressing well, with key deals closed with top customers. There is strong customer interest in our integrated full flow that combines our multiphysics products across the electrical, PFD and structural domains to best address next-generation system design needs, including in the emerging field of physical AI.

In summary, Q2 was a great quarter for Cadence, and I'm delighted with the continued momentum of our business. With accelerating design activity, we continue to execute strongly and our competitive position has never been better as we lead the transformation to agentic AI in chip and system design.

With that, I will turn it over to John to provide more details on our Q2 results and our updated 2026 outlook.

Thanks, Anirudh, and good afternoon, everyone. Cadence delivered excellent results for the second quarter of 2026, with accelerating momentum in AI and broad-based strength across all our businesses. Robust design activity and customer demand drove 24% year-over-year revenue growth for Q2, with double-digit growth across all our product groups.

with strong execution, we generated Q2 operating margin of 45.5%, and second quarter bookings resulted in a record backlog of $8.1 billion. Here are some of the financial highlights for the second quarter, starting with the P&L. Total revenue was $1.584 billion. GAAP operating margin was 28.4%. Non-GAAP operating margin was 45.5%, and GAAP EPS was $1.33 and non-GAAP EPS was $2.11. .

Next, turning to the balance sheet and cash flow. Our cash balance was $1.440 billion on the principal value of debt outstanding was $2.500 billion. Operating cash flow was $635 million. DSOs were 65 days, and we used $200 million to repurchase Cadence shares. Before I provide our updated outlook, I'd like to highlight that it contains the useful assumption that export control regulations that exist to date remain substantially similar for the remainder of the year. .

For our updated outlook for 2026, we now expect revenue in the range of $6.260 billion to $6.340 billion. GAAP operating margin in the range of 27.75% to 28.75%. - Non-GAAP operating margin in the range of 43.75% to 44.75%. GAAP EPS in the range of $4.76 to $4.86. Non-GAAP EPS in the range of $8.05 to $8.15. Operating cash flow of approximately $2 billion. And we expect to use approximately 50% of our free cash flow to repurchase Cadence shares in 2026.

For Q3, we expect revenue in the range of $1.595 billion to $1.625 billion. GAAP operating margin in the range of 27.5% to 28.5%; non-GAAP operating margin in the range of 43.5% to 44.5%, and GAAP EPS in the range of $1.11 to $1.17 and non-GAAP EPS in the range of $2.01 to $2.07. And as usual, we published a CFO commentary document on our Investor Relations website, which includes our outlook for additional items as well as further analysis and GAAP to non-GAAP reconciliations.

In conclusion, I'm pleased with our strong first half results and the robust pipeline and momentum heading into the second half of the year. At the midpoint, we now expect revenue growth of 19%, operating margin of 44.25%, EPS of $8.10 and operating cash flow of $2 billion for the year.

As always, I'd like to close by thanking our customers, partners and our employees for their continued support. And with that, operator, we will now take questions. .

[Operator Instructions] And our first question comes from the line of Joe Quatrochi with Wells Fargo.

Yes. Maybe first just wondered if you could give us any help. You talked about agentic AI as being a long-term TAM expansion opportunity, is there any quantification that you can give us on that TAM at this point? And maybe how do we think about that as driving EDA as a percent of R&D expense to maybe higher over time?

Yes. Joe, thanks for the question. So like we've said before, I mean the great thing about agentic AI is it opens up a new TAM opportunity. At the same time, it calls more of our underlying physically accurate software. So going back to the 3-layer framework. So it's a new opportunity at the top layer and reinforces the middle layer. And I think we are pleased by the interest. I mean the interest is amazing, actually, almost all the big customers, almost all customers want to engage in our agent stack and now we have 4 super agents.

So I think it's a great opportunity for us. Now in terms of results, what I -- of course, we had great results in Q2 and the year so far. And there's a lot of strength in different parts of the business. But what I want to particularly proud of is the strength in the software businesses. If you look at our recurring growth, and that was particularly driven by strength of add-on business.

And so both we are seeing add-ons driven both basically for design for AI as our customers design more chips and also AI for design, which is agentic AI portfolio. And you can see that in our results. So what is particularly impressive, and this is, I think, the highest raise we ever had is that it is broad-based, including software and AI contributing to that growth. So we'll see how things progress for rest of the year.

Yes, Joe, I'll just add that -- if I could just add, the customer engagement, as Anirudh said, continues to accelerate. We're seeing increased evaluations and pilots and early deployments. And we continue to expect monetization through both new workflow products as well as increased usage of underlying engines. But just to be clear, we're still not assuming a sudden step function in our guidance. The opportunity is continuing to develop well, though.

And our next question comes from the line of Joe Vruwink with Baird.

Staying on this topic I wanted to ask about open source models, designing chips. And maybe if I just take at face value, it seems like an agent sought out EDA tools and then orchestrate the flow when tasked with chip design. So I guess my question is the implication for Cadence from all of those and 2 things come to mind: One, if customers now have agents capable of accessing your EDA tools, does that drive higher usage and more net consumption ultimately. And then two, where do you think the differentiation lies with the customer buying the Cadence models for orchestration versus customers maybe deciding to build on their own?

Yes, thanks for the question. I mean, like I said before, I've said this for years now, like 4, 5 years that the real AI orchestration and monetization will happen through this 3-layer cake. Just to remind everybody, the top layer is AI agent and orchestration, the middle layer is these -- our traditional, physically accurate tools ground truth and bottom layer is compute and data. So the recent news just confirms that framework.

And by the way, this will happen in all markets. The value of AI will go more and more vertical than horizontal and I've said this for a long time. So even in chip design, the value is in the agentic framework and all the mental model, all the knowledge graphs then calling the physically accurate tools on a rich set of hardware. And this latest news in case of [ Kimi ] doing that, I mean I think that I mean they said a chip, but I think it's a small block, which is about technology, which is like 20 years old on frequency that is 20, 30x lower than current frequency.

So even to design a small block at such an old node, they needed kind of EDA tools to do that. So this is going to happen again and again. And there is -- there have been open source EDA tools for a while, I don't know, for decades, and they're used in some university or specialized settings, but to really do real designs, people use cadence to do that. Now the differentiation will be in all. We want to differentiate in all 3 parts of the cake. So our knowledge graph and a mental model and how we do the reinforcement loops at the agent is really differentiated how we call then the middle layer through deep API access and the strength of our middle traditional tools is differentiated.

And then even in the bottom layer, as you know, we have palladium, we have millennium. We have special hardware to do that. So our differentiation will be in all 3. And then all 3 together, we are more differentiated than we have ever been. So I'm very proud of our differentiation of the mode we have. And then the fact that these 3 layers reinforce each other. Now the customers may always have their own agents just like they have their own flows right now, but to really do mission-critical tasks they increasingly depend on cadence as you're seeing that in our engagements. .

Yes. And Joe, Anirudh, always says that like agentic AI actually increases demand because agents invoke EDA tools continuously while exploring more design alternatives, and Kimi was really good example of that.

And our next question comes from the line of Vivek Arya with Bank of America Securities.

Anirudh, IP business has accelerated to over 40% growth. I'm curious what's driving this? How much is organic versus inorganic? And what is kind of the sustainable growth rate for IP? And then if we zoom out, I just wanted to clarify with, John, what the contribution is now with Hexagon and EPS dilution. .

John, do you want to start on that -- yes.

Yes, sure, sure. Just in terms of hexagon contribution. I., mean Hexagon is delivering as we originally expected, and it continues to contribute to SG&A growth. But the strength in our SG&A numbers is much broader. We're seeing momentum in 3DIC in advanced packaging, in PCB, multiphysics and physical AI. As the integration of Hexagon DAD is progressing well. And we see a significant opportunity to strengthen both the technology portfolio and go to market over time. But also, I guess on the IP, so IP had an outstanding quarter driven by AI, HPC, advanced node activity, memory bandwidth, chiplets and advise packaging.

There were strong customer engagements in beautiful waves. But IP revenue can be timely dependent from quarter to quarter.-- we're pleased with the momentum. But I would agilize any 1 quarter. Our competitive position continue strengthening across interface IP, every IP and how they foundation IP -- Hl Andrew called out, it represents another example of customers choosing broader strategic engagement with us. Adrian, would you like to add?

Yes, absolutely. Yes. Thanks,. So Vivek, I'm very pleased with the IP performance and EDA performance. I mean before I get to specific IP, the good thing is, I mean, these things are growing. Of course, IP is growing very well. SDN is growing very well. but also they have enough scale now. So EDA, we are always, I believe, the leading EDA provider with analog, digital verification, packaging 3DIC. But both our roughly speaking, both IP and SDNA are approaching like $1 billion run rate, okay? So at this point, it gives us a lot of strength in our portfolio to engage with our customers. .

No IP, particularly. And I've mentioned this before, as you know, like I think there are 3 big mega trends. One is, of course, our IP is much better than before. The quality of our IP, the PPA, power performance in the area for like TSMC and the leading nodes is better. So we are getting a lot of competitive wins in IP that 2 years ago, we would not participate in. So that's 1 thing.

Second reason is our IP strategy is more focused, has always been focused and will continue to be focused to leading nodes to star IP to AI and HPC segment. So I've talked about these 5 IPs which is interface IP, memory IP. And then we have expanded to Foundation and other, but especially chip-to-chip IP, memory IP, interface IPs are supercritical and they are growing well, okay? And then the third thing is there are more and more foundries we talk about Intel. I'm very proud of this new partnership with Intel and is, of course, much broader than IP.

But I be a part of it. And then our engagement with Samsung, we mentioned last quarter, and wrap it is. So the foundry ecosystem is much more diverse than before. So I think these 3 reasons, our IP business is doing phenomenal. And also most of it is, just to clarify, is organic growth. I mean this great growth we posted most of it is organic growth. Now how does it proceed in the future? We'll see, but all the signs are positive at this time. .

And our next question comes from the line of Siti Panigrahi with Mizuho.

Apologies for the background noise. I'm at DACH conference, and I can tell you the key thing here is the agentic AI, which kind of validate what you said. So my quick question is you talked about some of this agent super agent, ChipStack, ViraStack that your customer has been using. So wondering what kind of feedback you're getting and the cost saving and the value that you bring to the customer. And then I know, John, earlier, you talked about monetization, which might take contract renewal or cycle time. But as you see the usage, are you seeing any kind of accelerating adoption where the time line can be compressed.

Yes, CT, the demand is great, like I mentioned, for these agents. And we have I mean the exciting thing is that the use cases are, I mean we have publicly talked about so many of them and like 2x to 10x to, in some cases, 40x improvement. And this is only the ones that we can publicly talk about. This is a very small subset of our engagement.

So the amount of use cases and the benefit is real, okay? And the interest is definitely real in terms of number of engagements and how many customers want to engage with us. And our strength of our portfolio with the 3-layer cake is very well differentiated. So I'm very pleased. I mean this is like we are maybe 6 months into our launch of these products. We launched them in Q1, but we're working, I would say, roughly 6 months with our customers. And we'll see how it progresses. But like I said, the early add-on business is encouraging, but we have to still in the early days, so we'll see how it goes. But so far, the demand is tremendous.

Yes, Siti, I think we view this as a demand accelerator. The customers are not trying to do any less design work. they're trying to keep up with design complexity, which is accelerating faster than engineering headcount and scale. We've always said that. As agents expand the design exploration space and call the underlying Cadence engines more often that create opportunities for new agentic workflow products and increased use of our core tools.

And our next question comes from the line of Jim Schneider with Goldman Sachs.

Continuing on the agentic AI theme. Could you maybe talk a little bit about some of the add-on engagements you're seeing for those tools and to what extent you're seeing them across more than the sort of 20 to 25 customers you've already noted. And maybe if you could quantify the impact of those add-ons in terms of either the guidance raise or what it could mean for core EDA software revenue in the next year, that would be great.

Yes. I think like you know us, right, we are very careful about projecting future next year numbers. But I think to step back a little bit, I think the -- the 3 things that I'm super excited about is one is that the overall environment is much better. I mean this also helps us a lot. I mean not just the AI companies the hyperscalers are. I mean the commitment to silicon is much higher than like 12 months ago.

And you can see that you're following all the hyperscalers. So the amount of designs and the number of designs each hyperscaler is doing is impressive. And then the AI semi companies are growing immensely. And then like the analog and memory and the consumer semi companies are also doing well now. So overall environment, is much better than 1 year ago, which, of course, helps us, right? So that's number one.

Number two, I think I just want to emphasize our competitive position, I feel has never been better. So we are taking a lot of share in different customers, getting to much, much deeper engagement, whether it's whether it's agentic AI or hardware or IP. And you can see that in the numbers and then the third part is this new TAM expansion opportunity, which is a agentic, which we are clearly super excited about, we're still in the early stages.

So if you combine those 3 things, I think that is what is leading to such good results and such good guidance. Just to remind you, this is the highest we have raised annual revenue in a single quarter, okay? And to about 19% revenue growth with improved profitability. So I think I would like to say that some of the benefit is already there of the agentic and other next year and year after, I mean you know as we are prudent as ever, and we'll see how things progress.

Jim, I think just -- I know we get a lot of questions about agentic AI, but I think it's important to highlight that the raise that we just did for Q2 for the rest of the year reflects broad-based strength across the business rather than any single customer or product, we saw strong Q2 execution across core EDA, IP, hardware and SD&A. That, of course, is all benefiting from continued strength in AI-driven demand as well. But the strength is broad-based across all businesses and across all regions.

And our next question comes from the line of Harlan Sur with JPMorgan.

And as the volume of AI influencing compute workloads surpassed training workloads in the second half of last year. And we know that inferencing is much more memory intensive, right? So you've seen this diversification of different types of memory architectures emerging to address inferencing in addition to HBM DRAM, we've seen development of SRAM-based offload architectures, we've seen CXL-based conventional DRAM offload and even using enterprise SSD or flash-based memory, right?

So given all of the focus on these memory architectures and memory controller architectures, is this translating into some tailwinds for your custom Virtuoso family of EDA tool solutions or tailwinds for your CXL-based or memory compiler IP portfolios or both?

Yes, Harlan. That's a great point. So yes, like John mentioned, the strength is broad-based. And definitely, the analog group, which is part of EDA is also seeing very strong momentum because all of these whether it's memory or analog is all done in Virtuoso, is the leading platform for analog and mixed signal and custom design in the industry. So I'm very pleased to see overall environment plus the special -- the all this innovation that is driven by influencing helping both all our businesses, analog, digital and verification. But what is exciting to me in this -- I mean you know this anyway, with this inferencing is that there is much more varied architectures you mentioned and also much more varied customers.

So all the big customers believe at this point that, of course, they will use standard products from semiconductor companies from the really big semiconductor companies like NVIDIA who are doing great, but also believe that they will have their own custom silicon. And then on top of that, different versions of that custom silicon for memory access and also networking, right? There's a lot of activity in networking as well. And then I would say like over the last 6 months, I see a lot more activity in startups. Startups were kind of dormant.

But in the last 6 months, there are like some very high-profile start-ups that are starting, not just in AI, but in networking and even CPU. So I think the overall environment is good and it is affecting all our businesses. Analog, for sure, verification with hardware, IP business, digital implementation, 3DIC is a big thing where we have leadership. So that's what leading to this broad-based trend. But the conviction of the hyperscalers to do their own silicon and try, like you pointed out, different architectures. And that's bound to happen.

I mean if there is one bottleneck, the customers come up with different memory architectures to solve that bottleneck or different networking architectures. So I expect this to continue. I mean this is -- as the market gets bigger, you know that as the AI infrastructure market gets bigger, there will be more and more innovation to optimize each part of that market. And all that innovation will require Cadence products to make that happen.

And next question comes from the line of Charles Shi with Needham & Company.

I can ask about AI for 100 ways, but I think the most important question on top of many people's mind right now, or I should say, the scenario, a very extreme scenario that people fear about the most is where you actually prompt, I don't know when we can get that, but prompt, very, very powerful LLM in the future with the chip design requirement, and that LLM can autonomously generate codes that's gets send to foundry directly for [ PayPal ] without running them through any of the commercial EDA tools. So this is one of the scenarios some people were envisioning. We strongly disagree, but do you think this end-to-end so-called end-to-end LLM based chip design is a real possibility at all? Or since you mentioned a 3-layer cake...?

Yes, Charles, I mean like I said before -- I mean before, I said this for years, the way this improvement will happen. And of course, there will be a lot of improvements with AI, there will be to this 3-layer cake. So we will have agents like we have Super agents. Our tools are central, will continue to be central to that. And of course, we'll run a varied set of hardware.

I don't see that changing. Of course, some people may get worry about it from time to time, but the ground truth will prevail, okay? This 3-layer framework will prevail. And what -- if you talk about commoditization, I mean, I think what is likely to happen is not, the EDA tools get commoditized. What is likely to happen is at the agentic layer, there will be a lot of choices for LLMs. So if you look at what is really happening right now in the marketplace is that the customers are demanding choice in their LLMs.

And so -- which is give me example of that and GLM 5.2 and Nemotron, of course, is great, released by NVIDIA and then all the commercial models. So what the customers are asking me is like, can you -- can the agent be more intelligent in choosing the right model for the right task given the rapid progress in the LLMs? I think that's most likely to happen. But the 3-layer framework, criticality of our tools will be here to stay, yes.

And our next question comes from the line of Lee Simpson with Morgan Stanley.

Great. I mean I think most of my questions have been asked, but maybe I'll ask a generic sort of competitive one. It does look still Cadence's expanded DTC collaborations now with Intel building out its Samsung road map and you've also deepened relationships with TSMC. So your positioning in stacked die and multichip designs is pretty much equal or better relative to peers, you'd say now -- so its exposure to digital design and IP interface maybe differs from Synopsys.

So I guess the question here is really, where are you seeing the most competitive pressure from some of your peers in contested accounts? And is there another context of some of the other agentic AI push at your rivals? Are you winning or losing share in that digital implementation and verification at the leading edge?

Yes. Thanks for the question, Lee. So first thing, I just want to say that I'm very proud of this new Intel collaboration because Intel is a company we tried to work closer for a very long time.

I mean this is not a 1- or 2-year old problem, this is like a 10- or 20-year old problem, okay? But finally, we have a great collaboration with Intel with Lip-Bu and his new team. And I think we are working on it for a while now, but it's good to announce it in Q2. And it's a pretty broad-based collaboration. Of course, starting with what we had announced a month or 2 ago, 14A and DTCO, our agentic EDA solutions, our IP portfolio, which is much stronger.

But I think it goes beyond that. And you'll see that we are engaging Intel in all parts of Intel, with all parts of our product portfolio. So I'm really pleased to see our position improving at Intel and our collaboration being just like it is all the other leading companies. And same thing happened with Samsung over the last 6 to 12 months. So in terms of what we were weak at before a few years ago was we were doing great with the TSMC ecosystem. We have a great partnership with TSMC.

But I've said for a while, we were weak at Intel and Samsung, and that definitely has changed. And there's still more to go, but at least the trajectory has definitely changed, in my opinion. And then and especially -- and that especially applies to digital and verification businesses. And even in digital, we are always very, very strong in implementation, place and route.

But now as we have mentioned in my prepared remarks, also strong in sign-off. So the depth of our digital engagement is also improving at all customers. So overall, I'm pretty pleased with our position. And we just always believe in simple things, right team, technology and customers. We have the best team, I believe, develop the best products and listen to these demanding customers, and that's how we stay ahead. We're not looking at who is doing -- who else is doing that, but are we really satisfying the demanding workload of our customers. And I believe right now, we are in a great position.

And our next question comes from the line of Jason Celino with KeyBanc Capital Markets.

Great to hear another record hardware quarter. I know, John, you kind of mentioned this always as a pipeline business, and you kind of wait to the middle of the year to get better visibility for the second half. But maybe can you speak to the type of demand activity you are seeing for hardware? I did notice that inventory picked up nicely in the second quarter, both on a year-over-year and a quarter-over-quarter basis.

Yes. Great question, Jason. Yes, we continue to see strong hardware demand, particularly from AI and HPC customers. Hardware-assisted verification is becoming a strategic capacity layer for customers designing the most complex chips systems. There could be quarterly timely effects, but demand remains solid. We continue to expect 2026 to be another record hardware year. And I would profile hardware is that it still remains supply constrained by customer demand rather than demand constrained as I would believe the systems as quickly as we can to deliver against the backlog. And yes, part of the increase was severe was due to hardware strength, but we are seeing strength right across the board.

And our next question comes from the line of Gianmarco Conti with Deutsche Bank.

Yes. So yes, amazing performance on IP. Maybe if you could share a few more words on Intel win exactly what does that entail? What parts of the portfolio? Was that displacement? How big is roughly the contract meant to draw down over how much time and is this in guidance, just kind of like the layout on the details, if you could share any of that, please.

Sure. I can comment a little bit more. I mean -- but this is a multiyear arrangement. And of course, some of the benefit is this year, but most of it is to come, okay? And then we will also invest more, right, in Intel and Intel customers, which is to be expected. But in terms of IP, I mean, it's much broader than IP because it includes EDA and DTCO. But in terms of IP, we have a pretty good portfolio. So we will make that available on Intel process.

Now this doesn't include as Intel foundry gets more customers, they're buying IP from us. This is just our arrangement with Intel right now. As you know, we are always conservative in those projections. But still, I mean, Intel foundry is, as you know, are talking to a lot of customers. So it's the possibility those customers will acquire these IPs and any differentiated tools that come out from this. So we will see how it goes. But IP strength is -- of course, Intel is a part of it, but it's much more broad-based. And even our overall trend, I think I want to highlight and John already mentioned is not coming from one particular thing.

So I mean there are 4 or 5 things that are driving this raised outlook. So Intel is one of them for sure. Ip Is 1 of them. Hardware is a key focus, but it's hardware. And if you look at our recurring growth is very good, right? So hardware is important growth, but so is EDA and agentic solution and 3DIC and [ SD&A ]. So I feel right now, there are like 4 or 5 engines that are driving our growth, but we are definitely very proud of the Intel agreement and the new partnership.

I know your question is primarily around revenue and things like that. But I want to highlight that there is some kind of expense in the second half as well because we're investing around these opportunities like Intel as well as trying to integrate Hexagon's D&E business because we're very focused on improving margins for next year. So you'll notice that the second half is kind of slightly lower margins than the first half, but that's a reflection of our making targeted investments. These are deliberate investments and not a deterioration in the underlying model by any means, is organic incremental -- our organic incremental margins remain very attractive. And we expect kind of acquisition profitability and profitability of IP to continue to improve as we go into 2027.

And our next question comes from the line of Ruben Roy with Stifel.

John, I think you just answered my question. So let me just make sure I understand that. So yes, I was looking at the implied operating margin near 43% and expenses -- R&D expense is up probably 19% year-over-year based on implied guidance for the full year versus around 10% growth last year. Of course, Hexagon accounts for a part of that. But I guess how much of this is sort of the core business. And I guess I was thinking through agentic AI and go-to-market, is that sort of hiring you're already admitted to? Is that driving some of the expense increase? And how do you expect that to roll into 2027?

Yes. I thought just, Harry, but if it's investment in systems and everything that we're trying to invest heavily in making sure we do a full and proper integration of Hexagon's design engineering business as well as some of the other businesses, the smaller business that we pulled into system design analysis. But we're very focused on that in the second half of this year. And I think I highlighted it last year that we had. I think, $20 million, $25 million set aside specifically for investments in the second half of the year. Now there's always some improvements in our expense expectations and I always want to give the team enough scope to be able to invest and capture the increased profitability opportunities that they can get. But our focus is really on in the second half of this year to grab those opportunities and set ourselves up so that we have better operating margins next year. .

And our next question comes from the line of Kilsichia with Citi.

So regarding Intel, is the engagement around working more likely an incremental driver to the sort of 20%, 25% growth that the team has been delivering for the IP business. And also, will it be a meaningful driver to our EDA business in the coming quarters? Or how long should we think about that trajectory into EDA business?

Yes. I mean just to make sure I understand the question. I think the Intel business that we announced is all incremental to our business. 100%. because we already had existing Intel agreement. So this 1 is a new agreement on top of that, and it's a multiyear agreement with multiple parts of that business. .

And then I'm also -- and I think Libo said that publicly also -- I mean we announced 14A, but I think Intel to be successful in the foundry business. has to do more than 14. So we're already talking to other future road map of Intel Foundry. And then, of course, there is different parts of Intel. As you know, the product groups and they're investing in their server business and their client business. So again, we are proud to be working with Intel closely. And just like we work with other big customers. So I think it's more a normalization of our relationship with Intel like we work with all the other household names. So I'm very proud of this development.

And our next question comes from the line of Jay Vleeschhouwer with Griffin Securities.

Anirudh, I'd like to ask you about the practical implications of or requirements for implementing AI and agents and all that you've spoken of this evening. That is to say, when you think about the pre-sale and post-sale support, customer support that you have to provide, how would that compare to, let's say, what you used to have to do for classical EDA? Is there something quantitatively or qualitatively or technically different now that you need to do that we haven't had to do before.

And what I have in mind, for example, is that over the last few months, there's been a very clear uptrend in your AE openings to classic leading indicator for customer adoption. We've also said that Gen AI is a critical path for agentic adoption. So maybe you could talk about that as well.

Yes. Thanks, Jay, for that question. That's a good question. I mean, in general, of course, we are growing. So we will invest, right, both in R&D and application engineering. Agentic AI, it does not require some massive step increase in investments. I just want to be clear about that. It is more of our traditional business because we are, of course, very, very asset light, right?

We don't -- we're not building compute farms. All this is done by our customers, okay, just to be clear. Now of course, all of the skills are different, but our team anyway is expert in competition software, as you know, and they can pick up agentic AI, some hiring we will do. So it's more of a business as usual, I will say. And also, we can make AI -- there are implications that we have not. We are applying a lot of AI internally, okay, to make things even more efficient.

So for example, AE, yes, we are hiring AEs, but AI can dramatically reduce AI workload and make them much more productive. So then more of the AEs can participate in presales activity rather than post-sales support, right? And same thing in R&D. Of course, we are deploying AI for software development.

And of course, deploying our agent tag and all our agents for IP development to make them more efficient. And this is what John was saying earlier. So we'll see how that progresses. I think AI has the opportunity to even reduce our cost in some cases. But we are not -- you should not model in some massive investment. I think the investment we talked about is more for SD&A, right, for the integration? And you've always talked about, Jay, that how we need to have a full flow. And I feel that finally, we have a full flow in SG&A. So investing in Intel. But the agentic AI will go through a regular sales motion and regular AI and RD support.

And our next question comes from the line of Joshua Tilton with Wolf Research.

Maybe just 1 clarification and 1 thematic question for Andrew. On the clarification, side. Could you just maybe unpack for us what's driving the strength in other recurring revenue that kind of stood out to us this quarter? And any commentary there would be helpful. And then maybe on the sematic side, Anirudh, unless I misheard you in your prepared remarks in the beginning, you talked about becoming more of a strategic partner for your customers.

The question is for you, but John, feel free to jump in here. Maybe like help us as financial analysts like understand what that means from a business perspective? Like are you growing wallet share? Are you taking more -- are you able to charge more? Like -- how are you as a company capturing value financially because you are now becoming more of a strategic partner to your customers. That makes sense..

Should I started Anirudh, Josh, I'll take the recurring revenue question. I mean recurring revenue grew about 24% year-over-year in Q2, and that was driven primarily by strong core EDA growth. Some AI drives about there's share gains and healthy renewals and expansions through add-on business. Within that, probably Hexagon contributed roughly 4 points. But even adjusting for that, your recurring revenue is like high teens to 20% on a normalized pro forma basis.

Which we view as a very strong result. As we also -- we would continue to expect the full year mix to be roughly 80% recurring and 20% upfront. On the agentic AI essentially when you look at the way we sell that, first of all, customers continue purchasing our underlying EDA software. I think what they do is they purchase cadence agent licenses that orchestrate engineering workflows. So generally, our economic scale with customer adoption. I don't if you like to add to address the rest of Josh's question.

Yes. I think, Josh, what we are saying is that, I mean, we are always strategic to our semiconductor customers, right? Of course, we are part of engineering. We're part of R&D, right? I mean we are not like other kinds of enterprise kind of software, this is engineering software. So we are central to them making their products and their revenue. I think what has happened lately over the last, let's say, 1 year is even in semi companies, our engagement is there's a much higher level in the company because EDA and chip design and agent AI opportunities are very meaningful to our customers. As there's more demanding road map as Moore's Law is kind of slowing down, this is well not producing enough improvement. So the improvement has to come with design efficiency, better optimization. -- better use of AI and also the middle layer, right, the PPA provided by -- and we do this with the foundries and with the customers and DTCO is part of it for better optimization of power and performance area. .

then it was possible if Moore's Law was delivering, was really moving fast. Right now, it has slowed down. So that's on the semi side. And on the system side, I think the realization that semiconductor is essential has happened now in the last 6 to 12 months or so. So all the MAG 7 companies, all the big really household norms, silicon is a critical part of their road map. So therefore, cadence engagement is super critical at these customers. And the way to monetize that is we provide more value to them and then we can get more value for us as you see in our results.

And our final question comes from the line of Gary Mobley with StoneX.

maybe a question more for John. 1 thing that stands out is what appears to be about a 55% increase in your bookings in the first half of the year versus the same period last year. I assume you're going to build on what is normally a seasonally strong second half of the year. And I thought this was a low renewal period for some more substantial customers. Maybe if you can speak to what's driving that booking strength? Is it a reflection of the strength of the chip cycle? Is it the strength of the function of the strong chip design activity? Or is it a function of some of the AI tools driving increasing usage of more copies of classic EDA tools?

Sure, Gary. I mean it's a great question. I think Anirudh spoke to it a little bit there to Josh's question. But we've been -- we always say it's strategy first, right? I mean we continue to execute against our intelligent system design strategy. Anirudh mapped that out for us for the last decade or so. But -- and what we're seeing is that all the underlying like structural demand drivers continue to strengthen the semiconductor complexity, AI infrastructure investment, engineering productivity, physical AI, agentic workflows, all of those trends seem to still be in their early stages. And I think that's feeding into really solid bookings for us.

And this year is and probably one of the low years when you look at the kind of a 3-year cycle on renewals this year is kind of probably one of the lower of the 3 years. But the -- but we're seeing very, very good strike that add-on opportunities that Anirudh mentioned earlier in the call. I'm very, very pleased with the progress at how things are going. Anirudh, anything to add?

No, John, that's a great summary. I mean, like John said, Gary, that yes, this year is the low bookings here. And also, normally, first half, we draw down on our backlog, but this year has been good growth. So we'll see how that progresses, but we are very pleased. And like John mentioned, the environment is good. And I'd just like to point out that I feel the 3 big reasons are like John was always saying the environment is great, both like the AI -- new AI comers, the traditional AI and the regular companies.

Our products and competitive position is fabulous and then this new TAM opportunity with Agentic-AI. So if you combine all these 3 things, I mean, the first half has been great. It sets up nicely for rest of the year, and then we'll see how things progress, right?

And I would now like to turn the call back over to Mr. Anirudh Devgan for closing remarks.

Yes. Thank you all for joining us this afternoon. It's an exciting time for Cadence as we enter the second half of 2026 with AI-driven product leadership and strong business momentum. On behalf of our employees and our Board of Directors, we thank our customers partners and investors for their continued trust and confidence in Cadence.

And ladies and gentlemen, thank you for participating in today's Cadence Second Quarter 2026 Earnings Conference Call. This concludes today's call, and you may now disconnect.

Okay. Thanks, everyone, for joining us back after lunch. We're going to keep this lively so that nobody goes into the proverbial as we call in the U.S., the food coma and make sure that we can really get some good insights here from Cadence.

So before I get into talking to Richard Gu, who is the VP of IR at Cadence, I'd like to read the disclaimer that today's discussion will contain forward-looking statements, including Cadence's outlook on future business and operating results. Due to risks and uncertainties, actual results may differ materially from those projected or implied in today's discussion. Everyone understand that? Good.

Richard, thanks for being with us. We appreciate you -- we always appreciate you coming back and you are in such great demand for the meetings today. Let's start by -- if you wouldn't mind giving us a quick overview of Cadence and what you think differences -- differentiates Cadence today from just a few years ago. I used to remember it as Cadence design. That shows how far back I go.

Thank you, Bob. Great to be here, and good afternoon, everyone. So Cadence, we've been around about 30 years, and the company was founded by engineers, for engineers, okay? And we play a very pivotal and foundational role in the entire semiconductor ecosystem, which is one of the most critical, most dynamic in the world. And our technology portfolio consisting of IP, EDA and system design analysis are really essential for our customers to design the most advanced chips and electronic systems from anywhere from the AI accelerators to smartphones to autonomous driving vehicles to aerospace systems.

So it's fair to say any electronic system in the world has always has a component of cadence technology in it. The company has grown by leaps and bounds over the years, and we've always been a great compounder. And this year, the revenue has accelerated to 17% year-over-year growth. And our non-GAAP op margin is going to push and reach 44%. So in a Rule of 40, if you will, we're talking about exceeding the Rule of 60 this year, which is going to be a company record.

I think from -- if you -- on your second part of the question, Bob, comparing us now versus a couple of years ago, I think a couple of things have changed, okay? One thing is worth pointing out is obviously, AI is a big inflection points, right? And we, as a company, under the leadership of our CEO, Anirudh Devgan's Intelligent System Design strategy. We are thinking about AI from really two major vectors, and we have a massive AI beneficiary.

One is design for AI because our technology -- we're one of the few companies where our technology is embedded and used and it's so essential to design all these AI chips, okay, NVIDIA, Broadcom and everybody, all hyperscalers. And we also talked about the AI for design. So we apply AI, the reinforcement learning, the agentic AI to our own product set to make it better. You probably heard about [ Jason ] talking about the -- our agentic AI product ChipStack, providing over 40X kind of productivity improvement for his engineering team during the most recent Computex in Taiwan. I think that's a strong validation to our product road map and our pole position really when you think about the agentic AI in our industry, okay? So I think that's one thing.

The second thing is the overall environment and the customers' environment in our operating kind of environment is getting a lot better and improved, okay? Not only does the 10, 15-ish top AI companies are going gangbusters, kind of really lock up in this dead heat to one of each other in the AI race. But also, if you look at the broader set of traditional semiconductor companies like the analog designs and mixed-signal companies, they are getting stronger, and they have a big role to play in the data center with power and everything. And they're also going up cycle. So these are great things for us. It bodes well for our business.

The last point I want to mention is competitively, we're very strong, okay? We're the strongest ever in our company history. So we feel very good about where the business is right now.

Yes. You don't sound too dissimilar to what we talk about in our business, which is AI in the business and AI in the product. And so differentiating between those 2 and how we use them. We can get into more of that later on. But can you further highlight some of the kind of secular trends that you think are driving this long-term growth? And tell us about maybe some of your top customers and your partners that you use in the semiconductor ecosystem that are working with you on this?

Sure. Yes. Our business, if you look at our top, say, 60, 70 customers, which is maybe a majority of our revenue, 60%, 70% of our revenue, they're the [ hoosives ] of the world, okay? And we also work very closely with all the major foundries in the world, TSMC. I mean we have made a great announcement to collaborate with Intel on its 14A journey yesterday. So it's a big step forward. And we collaborate with Samsung and everybody in ARM also, okay? So really just very strong, sticky kind of ecosystem.

And when it comes to the secular trend driving our business, I would probably break it down into, say, volume and pricing, okay? So volume -- from a volume standpoint, there are a lot of designs to be had, right? All these major companies they are competing to capitalize on this big AI megatrend and try to take advantage of that fantastic opportunity could be a lifetime opportunity for all of us. So there are lots of designs to be had. Not only do all these semi companies are launching a variety of advanced silicon and chips and systems. But also if you look at the hyperscalers and the systems companies, they are entering into the space very strong and their demand is very robust.

They're designing their own ASICs, right? We kind of use the analogy of the 4-story staircase. So as companies, they go about the custom silicon, they'll move from the merchandise silicon to ASICs using a third-party vendor to go after that journey. And as they mature, they go do some sort of what we call hybrid COT, it's customer-owned tooling. And then towards the end, they could do like COT, okay? So as they go down the stairs, what it means for us is it's not only more designs, but also the EDA and IP content will grow steadily,okay?

So it's a great business opportunity for all of us. And I think the last thing I want to mention is all these -- I mean, pricing is an important component of our business, right? And over the years, the industry has consolidated to really two major players, okay? And we are very disciplined in terms of the pricing conversations. I want to make sure we can capture that value. So overall, the company and the business is well set up to capitalize on the next wave of growth.

So we touched on AI earlier. AI disrupting Cadence? Are your customers going to use less of your tools because the agents are going to do more of the work and they're not going to need Cadence as much as they need today?

That's a good question. I think the beta was few years, a couple of quarters ago. But we -- our CEO, Anirudh, likes to use the analogy, which is I think is at in terms of our business is like a 3-layered cake, okay? It's not like we have a Cadence bakery or anything. Our core business is the middle layer which is the principled software, be it EDA, IP or system design analysis or hardware business, okay?

When you think about it, it's really grounded in the immutable kind of ground truths, be it physics or mathematics, right? So the relationship with the customers is very deep. It's super embedded. We are completely vested and committed to their journey. The conversation is R&D to R&D multiple times a day. So the business is irreplaceable in the middle layer, okay? But we are continuing to innovate on that, too. And when you think about the upper layer in terms of agentic AI, we're launching a slew of super agents. The example I gave just now on the CHIPS Act with NVIDIA is one of the many products we're launching, okay? We have like 3 or 4 super agents launched in the past couple of months, and the opportunity is massive, okay?

So the opportunities for us is not only as a new TAM expansion, right? So we can monetize and capture that with great pricing, and we can actually shifting more dollars from the labor budget in the R&D bucket to more automation and tools. Because one of the key things you have to realize and keep in mind for is there is this big mismatch between the design demand from our customers and the engineering supply, okay? TSMC have been talking about this 48 to 50x kind of transistor growth in the next 5 years, okay?

If you think about complexity of those designs, in terms of the volume and it's -- in terms of the workload volume, it's going gangbusters, okay? It is absolutely impossible for any of these companies to keep up by throwing bodies of the province, okay? So with our products with automation, AI, agentic AI, we have the opportunity to help bend that engineering hiring curve to help them meet their ultimate goal and objectives in the design process. So I think there are massive opportunities. We're seeing really early signs, very encouraging. The business -- the core business is doing great, okay? So I think we can be patient in terms of monetizing the top layer, making sure we can capture the full value. But the opportunity is very, very exciting for us.

So I guess I want you to remember a few things, right? You got to step down the stairs, 4 stories. You got the cake, the 3-layer cake. So there's going to be a quiz later. So got to remember all this, right? So I'm glad you jump right into my next question, which is really, it seems like you're benefiting on the other side of AI in terms of your -- these new Agentic AI tools. And so will there be monetization and revenue impact? And how should this group think about the potential timing of that?

Sure. It's a great question. So we -- in terms of monetization for this agentic AI products, one of the great things is, is a new category, right? And it's a TAM expansion. By pivoting more R&D dollars from labor to automation and AI. So it has a lot of promise. And we are thinking about a business model where we want to make sure it's a combination of subscription plus consumption, okay? Think of it as a rental car, car rental, okay?

The base subscription model is we obviously, we price these AI, call it, the virtual engineers, okay? And the value is commensurate to what a physical human engineer can do. So it's definitely not priced like an LLM token, okay? It's worth tens of thousands of thousands of dollars, okay? And if you drive in those car rental example, if you -- on a daily rate, you have a difference embedded 100 miles. -- okay? But if you drive more, like if you drive 500 miles, incremental 400 miles will come with, say, 4 tokens, each is worth like 100 miles per se, okay? So this is how we are thinking about it.

Another key point of monetization, we definitely should not lose sight for, and we're very excited about is, as you think about these virtual engineers, right, is agentic AI agents, a big difference with the human beings, they don't rest, right? They can't work 24/7, okay? So they can help explore the design space a lot more far away than a human engineer could do. So what it means is they're going to call a lot of the baseline underlying tools, which is the middle layer of the cake, I was referring to and talking about. So I think the monetization opportunity is enormous. It could come from all these different factors, and the natural question is where is the limit, right? I think we obviously -- we're still exploring, experimenting, but the early signs are very, very encouraging. [ Jason ] actually mentioned on stage with Anirudh during our cadence live back in April.

He mentioned he was willing to spend 50% of the human engineers cost on the tokens, okay? So what it means for us is it's 1/3 of the R&D budget, which is about 3 -- call it, 33%. Just as a reference point, right now, EDA is only 11%, 12% of the R&D budget. So the headroom is massive. So I think, obviously, we have -- the most important thing for us is want to make sure the products are strong, right? And then we're providing values to our customers. And that's how we can share and ultimately capture value accordingly. Okay.

Good. So let's shift gears a second, go into your IP business. You touched on that earlier. Seems to be growing well and well ahead of the market for the third year in a row. What's driving this?

Yes. The IP has historically with, I think, we deliberately underinvesting IP. Because I think Anirudh wanted to make sure EDA is solid and is world-class. I think we are at this point, right? We have the most comprehensive and strongest EDA platform in the world, okay? So -- and a couple of things have changed in IP, too. When you think about IP, I think AI definitely is a game changer.

With AI, I mean, these are really disaggregated architecture, right? Because a lot of the AI chips is not just one SoC. It's multiple chips all connected together in a chiplet or 3D-IC kind of fashion, okay? So what it comes with it, what it means for us is there are lots of high-value, high-growth IP, especially those connectivity IP, like the UCIe, PCIe, the SerDes and also storage to the memory, right, HBM, the DDRs of the world. And so our strategy is we want to focus on the, we call it star IP, the high-value, high-growth IP, which we are, okay? And I think a second important thing we did right is Hard hired this phenomenal leader from Intel, Boyd Phelps a couple of years ago. And he surrounded him with fantastic engineering leaders, okay?

So I think we're always a product-first company. As long as we have the right people in place, the product is getting better because ultimately, people's buying decisions for IP is based on the value of those and the PPA benefit, okay? And competitively, we're very strong in terms of the PPA kind of benchmarking. So IP is gaining ground. And now it's all very much exposed to the AI megatrend. And the third growth driver for IP is there is definitely -- you're seeing a foundry ecosystem expansion, right? TSMC is phenomenal. but you're seeing a lot of the other foundries too, right?

I think the Intel conversation and the announcement we had on 14A is a clear example, right? And then in Japan, they're building Rapidus, which is new foundry. Elon is talking about Terafab, all these things. So I think as the foundry diversify further, there's more demand and more opportunities because not only do we have to help them set it up, to enable the foundries and make sure the EDA tools can work seamlessly with them, just like the Intel situation. But also as they capture the end customers will have more revenue streams on that front. So IP, I mean this is the third year in a row. We are growing way above the market, okay, above like 20%, 23% this year. So we foresee the IP will continue to have very strong growth and continue to gain share in the market.

So let's talk about competition for a minute. How does Cadence view your competition? EDA has predominantly been a duopoly between you and Synopsys. Can you talk about how you differentiate yourself?

Sure. So I think we are under Anirudh's leadership, I think one of the main thing is we are a product R&D-centric company, okay? And if you get a product right, ultimately, you're going to win, okay? The strongest product always wins in the market. So I think that's the most important point to take away from. And I think at this point, we feel very, very comfortable in terms of our competitive situation. We are gaining share across the board.

I think EDA, EDA, we have the strongest platform. Analog is our market. digital, we are very strong, and we are gaining on -- with the Intel announcement and everything. Samsung, we're collaborating with Samsung SF2, okay? And when it comes to verification with our own ASICs, our Palladium platform is the gold standard in that market because we use our own ASIC, okay? So this is a clear differentiator. So we feel very good about core EDA is growing double digit strongly. And we touched on IP already, okay, IP where we're gaining, and we are much more focused and much more profitable also.

SD&A for system design analysis, we chose to focus on the two bookends of the market because not all SD&A is all tied or exposed to the AI megatrend, okay? So there are two bookends. One is closer to the silicon is the packaging and the 3D-IC and the chiplet, okay, which we have a strong footing because our Allegro is the market leader. And then on the physical AI front, closer to physical AI and robotics and autonomous driving and drones, we have built a strong business and platform with the most recent Hexagon acquisition, which really brought us the two key platforms. One is called Adams and another is called Nastran, these are the leading software simulators for multi-body robotics. And then when you combine that with the pre and post capabilities from Beta, CAE, which acquired a couple of years ago, I think we're full flow for physical AI.

And when you think about the physical AI, it's a phenomenal opportunity for us because not only do we innovate on that core server kind of realm, but also what it means for us is there's a lot of great silicon too, underneath that, okay? Because the physical AI, what it means naturally, it will be mixed signal, low power, and that's really our core strength when it comes to analog and mixed signals. So I think competitively, we feel very strong where profitability is great. We're talking about the Rule of 60. And I think continue to focus on our own execution and satisfy and delight our customers.

Yes. I want to talk more about the Rule of 60 in a second, but certainly in the right space now as we hear more about, obviously, drones with this little conflict going on in the Middle East. And then robotics and some of the robotics companies that we've seen, some of us have seen and all you have to do is go to Asia these days, and there's plenty of them being built. And I know that more and more in the U.S., robotics is going to be a big focus in the future. So it would seem like you're very well positioned there.

Yes, because I mean these are massive markets, right? I mean when you think about the AI, again, another 3, okay? We like 3x3 kind of model because for us, the infrastructure AI build-out is massive, right? It's happening right now and here, right? It's got so many more -- so much more growth to be had in there with all these investments and opportunities. But the next wave is emerging, too, right?

The physical AI is real. And they started with the autonomous driving. We live in the Bay Area and then you're seeing Waymo everywhere. I mean this is phenomenal. It's like when you think about the that the combination of the silicon plus the world models plus the physical cars, I mean, they've done a great job in there. So I mean, plus these are multitrillion-dollar opportunities. So massive, massive opportunities. We're very excited about this.

Yes. We hear more and more about -- you talked about data centers, but the data center build-outs and using robots to build the data centers. And it becomes kind of virtuous and the opportunity that exists.

Well, that's a great point. I think, again, like when we think about the physical AI, it's actually in the cars, right? These are data center on wheels in a way, right? A lot of the data, which they collected has to be trained in the data centers, right? So it's a reinforcing kind of mechanism with the data center build-out in the physical AI world. So I completely agree, there's a flywheel.

Yes, because the more places that you put a Waymo, the more data that needs to be -- you can't have data from the Bay Area in Indianapolis that doesn't really work or Knoxville, Tennessee or whatever -- I mean, you need to have local data in order to really to make it function in those jurisdictions. So it will be more and more important for those -- to have more and more of these data centers and how they're going to be constructed and be constructed quickly using your tools.

Absolutely.

So you've touched on it a couple of times, you talked about the Rule of 60. So Cadence has been one of the few companies out there delivering on the Rule of 60. How should we think about the long-term revenue growth and operating margins under that scenario?

So we are a financially disciplined and fiscally responsible company, okay. We don't go out and guide multiyears, okay? We look out 1 year at a time. And it's also very, very important for us to make sure we are growing in a very profitable way, okay? So I think there is a trifactor of our operating philosophies. We strive to deliver and drive growth. And it's accelerating, right?

Clearly, you're seeing like 17% as it stands right now for the year. And also, we wanted to continue to drive 50% -- north of 50% incremental margin, okay? In our core business, organic business is actually close to 60 okay? And on top of that, we wanted to continue to spend and give back use like more than 50% of the free cash flow for share repurchases. So I think those were quite well for us for years, and we don't have any intention to deviate from that philosophy. So we certainly feel very good about the long-term trajectory of the business because it feels like all the long-term drivers, if you will, tailwinds, it all stays intact if you look at a very long horizon in ARC in that way. Yes.

It. So one of the things you mentioned earlier was your acquisition of Hexagon, D&E which I believe fits into your system design and analysis business. How does that increase the TAM? And what was the strategic rationale behind that acquisition?

Sure. So Hexagon is really is a fantastic technology, and it's a carve-out for us. What's valuable for us is, I mean, again, like I mentioned, they got two great platforms. One is called Adams, okay? It's a multi-body simulation tool and another is called Nastran, okay? So when you think about the physical AI challenges, part of that is just the simulation is not accurate enough, okay?

So these tools could be kind of interposed or inserted in the simulation loop to make it work, okay, to make the simulation lot faster and more accurate, okay? So that is a great kind of set of tools we are acquiring. And then like I said also, -- by combining that with the beta CAE's technology, we have a physical AI simulation full flow, okay?

So with that, I think we are well poised to capture some of the opportunities as it emerges. I think in the -- so in the autonomous driving segment, robotics is coming up very strong, right? So I think certainly, I think these are great business. Integration has gone quite well. So we closed the deal about a quarter ago. And then our team is just heads down, focused on making sure we'll continue to capture the upside and the opportunities.

So you think of that TAM, how do you think about that TAM?

Yes. So the -- I mean, if you think about the EDA TAM in the -- I mean we've talked about the convergence between silicon and systems, okay? So it's fair to say that the simulation and SD&A TAM is as big as the silicon side of the equation. So it could double the TAM over time. And then not to mention, now AI is another lack of growth. So I remember years ago, the TAM is, I call it, $10 billion, $15 billion. Now it's much higher now. So it's a great opportunity for us to continue to prosecute, I say, yes.

Sure. So we'll have time for some questions in a minute or 2, if any of you have them. So I just want to preview that for you. Are there any areas of the portfolio, Richard, do you think we could benefit from more M&A, more focus on that? And what's the philosophy? You talked about how you return cash through share repurchases, but how does the philosophy look between M&A and share repurchase?

Sure. So our philosophy has always been organic is the first order of business, okay? We always invest first and foremost, in our own R&D capacities and capabilities. This is a very R&D-intensive business. And organically, we have to say organic is delicious, okay? So that's has the greatest highest return for our shareholders. So that's our core focus area. And we don't do major transformative deals, and we don't have any EBITDA to do that, okay?

We will supplement the core organic business with some tuck-in acquisitions opportunistically if the right asset is in the market with a good price or a good talent out there. But we don't feel any need to do any major deals, okay? And I think the third order, obviously, is the share buybacks. So -- but I think we feel at this point, the portfolio is fairly complete and comprehensive and the market is growing very nicely. So I think we wanted to make sure we have the best product in place to support the growth and the innovation agenda for our customers.

So focusing inward rather than outward.

Focus on organic. First and foremost, I'd say..

Do we have questions? Yes, sir?

Thank you very much for all today for your attendance. This session is now concluded. You may now leave the webinar. Thank you.

Good afternoon. Welcome back to this BofA Global Technology Conference. Really delighted and honored to have Anirudh Devgan, the Chief Executive Officer of Cadence Design Systems, join us. We'll go through the usual fireside. Please feel free to raise your hand if you would like to bring something up.

But before I start, let me read a quick disclosure statement from Cadence that today's discussion will contain forward-looking statements, including Cadence's outlook on future business and operating results. Due to risks and uncertainties, actual results may differ materially from those projected or implied in today's discussion.

So Anirudh, welcome. Very, really happy to see you at our conference.

And maybe at the start, what I would love to get from you, I think you have been in the CEO's role for roughly about 4.5, 5 years or so. And what's sort of the master plan as you look over the next 5 years? You have gone through this transformation of the company from kind of just EDA, right, to kind of diversifying it into system design, multiphysics, right, more IP. But what's the kind of the grand strategy? What should investors expect over the next 5 years?

Yes. Thank you, Vivek, and it's great to be here. By the way, I'm a big fan of Vivek and the CadenceLIVE in April, I gave the talk and your numbers just came out, your semi market numbers. So I quoted Vivek and put it and then all the analysts that I should also quote them, but Vivek always have very good numbers and analysis.

So it's great to be here, Vivek. And then in terms of Cadence, yes, last 5 years have been phenomenal growth, I think. And because we have about 15% CAGR in a tough market. Parts of the semi market was good, but part of the semi market was not that good. And then also our margin has improved. We crossed the rule of 60 this year. Now going forward, I think the environment is improving. I think our customer environment is improving. Our competitive position has the best it has been.

So going forward, I think -- and of course, I started all this SDA in 2017, 2018. And SDA is still good, but the value of EDA and IP is much higher because of resurgence of semiconductors, resurgence of AI. I mean, not just in semi companies, but in hyperscaler companies. So I think it will be more of the same, but I feel that we are well positioned in all the 3 areas: EDA, IP and SDA. And EDA and IP have potential to grow meaningfully in the next 5 years. SDA still will be good, don't get me wrong. But I think the value of EDA in 2026 is much higher than 2018.

Got it. The one question that has emerged this year for not just Cadence, but for the entire -- anyone connected to the software industry has been the potential of disruption from AI. So give us kind of your measured view, are there certain parts of your design flow where AI can be disruptive? So how do you think about just the emergence? Of course, it's giving you opportunities for more engagement with the customers. But are there parts of the design flow that we should kind of watch out for any kind of disruption?

Well, AI is a net positive to us, and there are multiple reasons for that and I've said this a long time, so sorry if some of this is repeat.

So first of all, in terms of software, we are both involved in the building of AI and the consumption of AI. So this is designed for AI and AI for design. That is not true for any other software. Okay, we love all our software colleagues, but no other software is directly involved in building of NVIDIA chips or Google CPUs. So I think that's one benefit. And then when we apply AI to us, it can be transformative in terms of productivity. And then people worry that, okay, if you are going to be 5x more productive, does that mean less usage of your base tools.

But what you have to remember is even in that part, that's why I'm so optimistic about not just about SDA, but EDA and IP. When we apply to AI to our products, our workload is exponential. See, I'm talking to one big customer, and this is true for all of them. They're saying that every next chip, they need 2x more engineers. That's an unrealizable headcount curve. So AI is needed. AI will blunt the headcount curve, right?

And if the workload is constant, if it blunts the headcount curve, there's a worry that it reduces usage. In our case, because the workload is -- demand is exponential, the blunting of headcount curves makes it possible. So the market expects 5x, 10x improvement, just like we have delivered 100x over the last 20 years.

I just came back from [indiscernible], I was telling some people, and this is supposed to continue. Kevin gave a good talk from TSMC, and he's showing TSMC road map, the number of transistors on these systems, chip plus package will go up 48x or 50x in the next 5 years. So to design all these things, you need that productivity, okay? And that you can see that even like when -- like, for example, recently, Jensen talked about Cadence at COMPUTEX. By the way, that's the work internally at NVIDIA and they highlighted it is that -- so when the customer is writing RTL, and this is the kind of thing that we never had tools before. It's a new TAM for us. They were using Xcelium and Jasper to write it because we need to verify that the RTL is correct.

So now when ChipStack is writing it, it actually uses more Jasper and more Xcelium. So actually, the number of base tool usage is going up versus a non-agentic world. So that's first thing that happens.

And second thing is this new TAM expansion of this kind of capability that the users were doing manually, and we have now this agent product that we can provide to our customers. So in net, because of all these things, because we, first of all, participate in the build-out of AI, that's one. Second, when we apply AI it's an exponential workflow, exponentially growing workflow.

And third, the way we apply it, and this is what I've said forever, is that it uses the base tools more. I've talked forever about this 3-layer cake. So the Agentic AI, then the ground truth tools and then the base layer, which is compute and data. Now if there is a certain application in which the base layer or the middle layer is very trivial or not as complicated, then maybe the LLMs can do everything themselves. But in this kind of complicated physically accurate workload, you have to use the ground truth tools. You cannot do it without that. And actually, Jensen has a very good analogy, okay? He was saying that -- just to -- is that he was saying that, let's say, agent is like a robot, right? And let's say, robot comes in your home.

If what you were doing was relatively simple, that's picking like the bottle from one place to another place, then maybe the robot can do that. But if you're going to prepare food like or use a microwave, is the robot going to warm the food with its hands or it's going to use the microwave that's already there. And in our case, we are not even like a microwave. We are like a nuclear reactor.

So when a bunch of robots come, they're not going to build nuclear reactor. They're going to use nuclear reactors in a more effective way. So I think the complexity of the task is also there. So for these 3 reasons, one, complexity of the task, two, the exponential nature of the workload; and three, that we also participate in the build-out, I think AI is a very big net positive for us.

Got it. So just to kind of close that, there is no genius sitting at any hyperscaler who is writing software that can completely remove their use of any part of the EDA process.

Well, there are a lot of geniuses sitting at hyperscalers. All our customers are very smart, but they will develop all these things working with us. I think that's what we see. They want to enable -- again, they want to use our tools and build on top of that and use our agents.

Makes sense. Now one other thing I also find fascinating is at the leading edge, right, TSMC, right, their ability to drive the advancements in silicon and you are very strongly engaged, right, with the TSMC ecosystem, right, a lot more than some of the other ecosystems. So as they raise their pricing, is it fair to think that, that actually puts a lot more pressure on the upfront design process. So that's a good thing. But does it also perhaps limit the number of design starts because these things are getting so much more expensive. So how are you seeing those 2 kind of trends evolve, right? That one is complexity is growing? And second, is it then becoming a headwind to the number of design starts, which often tends to drive your business?

No, number of design starts is still very good. It's increasing. And of course, I've said this for a while, not just for data center, but for physical AI. So right now, the design activity is very, very strong. And I think that it has a chance to further improve as -- because 2 years ago, I mean, there were some design activity inside the big system companies. Of course, some of the really big phone companies have done this for 10 years.

But what is also new, you know all this, in last 1 year, the success of Google is a big shiny example of verticalization. And same thing in China, I just came back from China, success of Xiaomi. Xiaomi is very impressive, okay? They have their own car, they have their own model, they have their own chips. So once this kind of verticalization has happened in like physical AI with Xiaomi or they also have LLM models or with this big data center with Google, I think the more is happening to compete with that.

So the amount of design activity has picked up independent of this pricing dynamic of the foundry because the value is very high. So I see that increasing. And then the traditional semi-analog memory have also improved. So overall, the environment is much better than a year ago.

Got it. I'm glad you mentioned, right, the environment. There were a few years in between, Anirudh, where we saw the EDA, IP industry have a more modest type growth pace, right, closer to kind of low double digits rather than kind of the mid-teens. But that, of course, followed some years where it was high teens, but I think that was the time when China was also growing much, much faster.

If you look over the next 3, 4 years, are we now at this mid-teens kind of growth rate? What were the headwinds that kind of made the growth rate slow down? And now what are the tailwinds that you think can sustain the kind of growth rates that you're at right now?

Yes. I mean the environment is -- I think there are 3 main things for us. And of course, we guide 1 year at a time. And this year, I think, is very strong, and we'll see how the next years go. So first, the design activity is much stronger. And for all the reasons I mentioned, hyperscalers are back in and other system companies like Tesla or BYD.

Okay.

The parts of semi that were weaker, relatively weaker have improved. So that's number one. The market is much better. And we can drill down more into that. But overall, I think that's good. The second part is competitively, we are in the best position we have ever been.

Agentic, we are ahead. Base EDA, we are ahead. IP, we are taking share. Hardware, we have a unique platform with Palladium. So competitively, we are in the best position.

Okay?

And then thirdly, we have this new opportunity with TAM expansion with agent. So if you ask me like last few years, there were some -- there's always some issue or other like in the past, we were not as strong in IP or some of the semi was weak or we will not have full exposure to Intel and Samsung.

So I think all the -- we -- I feel good that right now, a lot of things are aligned well. And that is a change. We have big opportunity at some of the companies, we never worked before the parts of the semi has improved. We have new TAM expansion. So we'll see how it goes. But the environment is probably the best it has been.

Yes. Got it. We had -- to your point, we had ON semiconductor, right earlier, we had Microchip and Analog Devices yesterday. And it's so interesting to see that in the past, 90% of the questions were about industrial and automotive and now half the questions are about the data center, right? So that is also, I assume, helping these companies come out the downturn. So are you starting to see that because you have a very strong exposure, right, to kind of the analog market as well. So are you starting to see R&D activity pick up there as well?

Yes, absolutely. Absolutely. And then also physical AI also benefits that part of the market. Look at ADI or ST just to give some examples. So I know -- and I think the design activity is strong. And also, not only that, everybody wants to do more things more effectively with this agentic flows. So not only is the customer base doing well, but the interest in more automation is there. So we have the amount of engagement we have with like ChipStack or ViraStack or InnoStack is throughout our big customers, whether they're traditional semi or their data center or analog mixed signal or system companies, they all want to do things more efficiently, right?

Got it. Is that part of the agentic workflow, right? You mentioned some of these tools. Maybe talk to us about are they means of efficiency for you? Or are they a means of efficiency for your customers? Or can they also become a source of more pricing power or stickiness, right, like some of these Agentic tools that you described? How do they fit into the equation?

I mean, first of all, we are a technology company first. We want to make sure we deliver value to our customers. So like what NVIDIA was showing or we have so many other engagements with ChipStack like Qualcomm and MediaTek, if we can make things more efficient for what was a manual process, then that's a good TAM expansion opportunity for us. So that's the main thing we focus on is providing value to our customers.

And then the way the ChipStack works, it also calls more of the base tool. Now in terms of productivity benefit, okay, it applies to customers and us internally as well. So we have about 15,000 people roughly, okay? About 4,000 are like customer-facing or what these days would be called forward deployed engineers, what we call like AEs, application engineers. And then 10,000 people are in R&D, okay?

So out of 10,000; 3,000 are in IP. This is like broad numbers, okay, without getting into -- so I think with IP group, we are also applying our own agentic solutions to them, right? That's the best way to prove it. And I am expecting at least a 2x productivity from the IP group. So those 3,000 people should operate like 6,000 because anyway, we have so much demand for IP solutions, and we will hire more, but they should be at least.

So the way I look at it is at least 30% reduction in headcount per project and at least 30% reduction in schedule for a given project, okay? So that's like a 0.7 x 0.7, that's 0.4, that's 2x. That's like a Moore's Law kind of productivity, okay? And sometimes we can go even better than that. Some customers told me they want 0.5 x 0.5, okay? So that's like 4x. So I think with this agentic flow, there's opportunity of 2x to 3x improvement, and that's huge, right?

So now we will make sure that we deliver that to our customers and get this opportunity of the new TAM. And then we apply it internally on the -- and then this is just the IP group. AEs can be more efficient. And then the 7,000 people who are writing code, they can be more efficient with like other regular AI tools like Claude and Codex. So it's both providing value to our customers, which is the main thing and then adding efficiency internally.

Now internally, you see some of it already. Our incremental margin was 60%, okay? That's pretty good. And our operating margin is about 44%, 45%, but incremental is at 60%. So we will always try to drive that up and also provide this 2x to 3x to our customers. And again, this 2x to 3x or more is needed if the number of transistors is going to go up 48x, you need much more than 2x to 3x, but we'll start with 2x to 3x to our customers.

Got it. One thing, Anirudh, that has come up more frequently is, is the EDA industry utilizing its pricing power in the [indiscernible]. Interestingly, I find that the same pushback comes with the semi-cap equipment industry, right? Both are kind of ways of complexity. The usual pushback, right, which is your customers, their customers are making so much more money. Are we seeing it already and we just don't notice it as much? Or have we not yet seen it and that goodness.

So talk to us about has this increase in complexity, the use of all these AI tools, has it actually helped you improve your pricing power versus what it has been historically?

Well, we try to always get the right value from our customers. And first, like I said, our culture is first to deliver value, and these are like the biggest customers in the world, right? So it's not that they are short of money, okay? All these big 60, 70 companies that drive 60%, 70% of our revenue. So my philosophy always is if we can provide value, they are always fair to us, okay?

Now a lot of the growth in the past has been driven by volume more than price because they are doing more and more. And the way to get the right value for us is to deliver more value. So I do think this agentic flow in which they are able to substitute more of human tasks with agents is a unique opportunity because we will deliver a tremendous value like 2x, 3x, 4x, and this is the right way to capture more value for us.

Because it drives more consumption of base tools and your value is kind of more levered to the consumption of your tools, right?

Not only that, it's a new TAM, first of all. It's a new TAM, right? Because like if we are having tools to write RTL, like I was giving the previous example, there were no tools like that. So it's a new TAM to buy. And then they can -- they always have exponential workload. So I think if something can give them 2x to 4x productivity or 40x in case of NVIDIA, yes, they will use that.

So it's a new TAM. And then that TAM also drives more of the base tools. So it's a multiplicative effect. But our philosophy always is we are very thankful to all these big 70 customers and relationships we have for years. If we provide value to them, they will provide value back to us. So my focus is having the best super agents for ChipStack, ViraStack, this is InnoStack. This is a new way of monetizing because if they get so much benefit, they are more than willing to share that with us.

Got it. Makes sense. You mentioned growth in IP. So why did it lag historically? And what is helping it now pick up and it's now become one of the faster-growing parts of your business?

Yes. I mean IP, first of all, I mean, I also intentionally in the beginning, didn't invest as much in IP, just to be honest, because we wanted to make sure that we are good in EDA first. Cadence has been -- if you followed Cadence, there has been a transition of one business to another. So I also personally invested more in EDA because that's the core of our business. And if you are strong in EDA, then everything follows from that.

So first 4 years when I was doing, I was more focused on EDA. Then I think a few things have changed in IP. One is because of this AI and disaggregation, some star IP is a lot more valuable. And we are focused on these 5 star IPs, which just -- without getting too DDR, the memory subsystem, PCIe, UCIe, which is chip-to-chip, HBM and then SerDes, okay? So we are always focused on advanced node star IP because, again, we -- and part of the IP business was Tensilica, which is more profitable. But we always thought if we focus on few star IPs and do a good job, you can get a better margin and better behavior because our customers see -- when we have these really big customers, we always focus on win with the winners, the really big guys because once you win with the winners, the other stuff just happens naturally, okay?

The big with the winners, they always buy best-in-class. They're not interested in -- they're not buying something for -- they want to buy best-in-class. So it's better to focus on a few things and do them well rather than have a big portfolio. And so that's the second difference, which I think has played out well now. It took a few years for that.

And third thing, which is probably the most important is I've over the years, changed all the R&D teams in Cadence, and we always have very technical leadership and very good R&D teams, okay? And finally, in IP, I finally believe that the team is world-class. It wasn't world-class before. And -- but I think over the last few years, they are world-class. And if you look at all the GMs in Cadence, they are all technical there because that's what our customers want, okay?

So as a result, these are standard-based IPs, even these 5 things I mentioned, DDR is a standard, right? So somebody chooses or not depends on how good the PPA is for performance and area. And we are using TSMC or whatever. So it's the same process, same requirements, but how good your R&D team is, it depends. So our PPA now is pretty good. The R&D team is good. So if the product is good, it sells, right? So I think -- and then there are a few other things that happened. Disaggregation helped us and then now Intel and Samsung and other rapiders. But the main thing is, I think we have the right IP strategy, which is more focused at these 5 key IPs and advanced nodes, and we have a great R&D team now.

Got it. For better or worse, investors will always look at the 2 kind of leading companies in the EDA space, right, and try to -- both are high-quality businesses. What do you think gives Cadence the edge in terms of potentially gaining share over the next few years? Is it that you have a specific mix of businesses? Or within those businesses, you have an opportunity to take share in certain things?

I mean, first of all, we are very strong in core EDA, okay? I mean there's no doubt about that. And we were -- we didn't have as much opportunity. And you can see that in the TSMC ecosystem. So in terms of PPA and also in terms of scope, right? So we have analog. We're the only company that has all the tools. So we have analog, we have digital, we have verification, we have packaging.

So core EDA, we are very, very strong. In IP, we are getting much stronger with this kind of strategy I mentioned. And then we have SDA, but what I would consider the right amount of SDA. You don't want to over SDA yourself. The right amount of SDA, which is focused on 3D-IC and physical AI. So I feel very good about the mix. And I think we have a very good R&D-driven culture, okay?

I'm R&D by nature. All the -- I said all the leaders are R&D because see, when you interact with these top 60, 70 companies, they're all very technical, too. They want R&D to R&D interaction. They're trusting their huge road maps to a company. So they want that confidence that the other side because we are deeply embedded with their road maps.

So we have a whole culture of we will enable R&D to R&D interaction, and that's thing we have gained over the years. And I think that is unique to Cadence. And so I think we are well positioned as the market improves. So I mean it's a good industry. What I tell to investors is, yes, you can invest in both, but just invest more in Cadence.

You say that very objective.

Yes exactly.

Of course.

It's also true, right, if you go back and look at the last 5 years, yes.

Okay. So on -- one other, I think, interesting thing about Cadence [indiscernible] space is you were sort of the early leading indicator of this growth that we are now seeing in cloud AI. What do you think, Anirudh, about the whole excitement and interest in kind of physical AI, edge AI. They still seem a little bit further out. But if anyone has kind of early visibility around that, so do you think investors should be paying more attention? Like are those going to be real markets, real players? How much of a design activity are you seeing in physical AI, robotics, right, edge AI?

Yes, I'm a big fan of physical AI. And I've been for a few years. By the way, I say the same thing year after year for a few years, and people used to tell me, what are you saying? What is physical AI. But now I think it becomes more -- because we do have some early view of what is happening. And it's still like I always said like 3 to 7 years, okay? So it's still like -- now it is already -- but if it's 3 years, then they already start designing for that. So we already -- if you look at all the car companies, and of course, a big famous thing is what Elon is saying, right? I mean they're investing a lot in their AI chips and also then with robots and all.

And then if you see what is happening, like Rivian is doing things. And then in China, there is a lot of activity already. So BYD and Nio and XPeng, they're all Cadence customers, okay? So that's why Xiaomi now not just making cars, but making robots, and they're very impressive. So I think you're already seeing signs of that in Tesla and in China and then some of the traditional companies, too. And then you're starting to see the signs of that with like ADI and TI and all the traditional semi companies. And then even NVIDIA, Qualcomm, MediaTek. So I think it will be a big market.

Now what I want to make sure is that as a company, we are well positioned for physical AI. I believe we are well positioned for data center AI. And also, we are well positioned physical AI. So there is simulation part, but also the base silicon content will go up a lot. If robotics is going to be the biggest market ever, then the amount of chips that go into robots will be high. If the chips is high, you need to design them.

And now exact timing is very difficult to say. Like there's data center, there's physical AI, it's like when it happens. But for us, we want to be best prepared for that. Now it may happen sooner, it may happen later. This may correct, may not correct. I think one thing I want to tell investors is because I believe we are well positioned with data center and physical AI, and we are not directly tied to there is pros and cons of that directly tied to the actual silicon volume. If things are good, it should be good. Even if things turn south a little bit, it should still be good for Cadence Design. So the physical AI part is to make sure we don't miss the next big thing while focusing on the current big thing, which is Agentic and data center.

Makes sense. With that, Anirudh, thank you so much. Really appreciate your time.

Thank you.

Thanks a lot.

Hello, and welcome to the Cadence Design Systems 2026 Annual Meeting of Stockholders. [Operator Instructions]

It is now my pleasure to turn today's meeting over to Anirudh Devgan, President and Chief Executive Officer of Cadence. Dr. Devgan, the floor is yours.

Good afternoon, and welcome. I'm Anirudh Devgan, President and Chief Executive Officer of Cadence. On behalf of our Board of Directors and our 15,000-plus employees around the world, I would like to welcome you to Cadence's 2026 Annual Meeting of Stockholders.

I will chair this meeting. And Cadence's General Counsel and Corporate Secretary, Marc Taxay, will act as Secretary. I will now turn the floor over to Mr. Taxay.

Thank you, Dr. Devgan. We're holding Cadence's 2026 annual meeting today in a virtual live audio webcast format. Please bear with us if we have any technical glitches or delays during the meeting. We thank everyone who is in attendance today.

We are conducting this meeting in accordance with our bylaws and the meeting rules of conduct. The rules of conduct, annual report, proxy statement and agenda of this meeting are available on the meeting website.

Now I'd like to introduce you to our Board nominees who are with us virtually. They are: Mark Adams; Ita Brennan; Lewis Chew; Anirudh Devgan; Moshe Gavrielov; M.L. Krakauer; Julia Liuson; James Plummer; Alberto Sangiovanni-Vincentelli; Young Sohn; and Luc Van den hove. Also in attendance are Sachi Patel and [indiscernible], representatives of PricewaterhouseCoopers, Cadence's independent auditor.

As a reminder, stockholders attending the virtual meeting can vote their shares or change their votes online from now through the closing of the polls by logging into the meeting website as a stockholder and clicking the link provided on their screen. If you have previously voted by proxy and you do not wish to change your vote, your vote will be cast as previously instructed, and no further action is required.

In order to log in as a registered stockholder, you will need to input the 15-digit control number that you received from Computershare with your proxy materials. In order to log in as a beneficial stockholder, you will need to input the control number provided to you by your broker's proxy distributor, likely in a communication from either proxyvote.com or proxypush.com. Alternatively, a beneficial holder could have obtained a control number from Computershare by submitting a legal proxy from your broker, all as described on Pages 107 and 108 of the proxy statement.

We will begin by attending to the formal business of the meeting. After the formal business is adjourned and to the extent time and format permits, we will conclude with a general question-and-answer session. Participants who are logged into the meeting website as a stockholder will be able to submit questions online for the general Q&A session by clicking on the Q&A icon on the right side of the screen.

I now call your attention to the rules of conduct for today's meeting, which can be accessed by clicking on the Documents icon on the right side of the screen. In order to conduct an orderly meeting, we ask that you abide by these rules.

Now at the request of the Chair of this meeting and our Board, I will conduct the business portion of this meeting.

The 2026 Annual Meeting of Cadence's Stockholders will now come to order. We will proceed with the formal business of the meeting, as set forth in your notice of annual meeting and proxy statement.

A list of the holders of record of Cadence's common stock as of the close of business on March 9, 2026, which is the record date set for this meeting, has been made available for inspection by stockholders at our corporate headquarters in the 10 days prior to this meeting. I also have affidavits certifying that as of March 25, 2026, notices of this meeting and Internet availability of proxy materials were deposited in the U.S. Mail to stockholders as of the record date in accordance with SEC rules and Delaware law.

A representative from Computershare, who will be acting as the inspector of election for this meeting, is also in attendance and has taken his customary oath. I now ask the inspector of election to advise whether a quorum has been reached for this meeting.

We have present, in person or by proxy, shares representing approximately 88% of Cadence outstanding common stock, which constitutes a quorum for the conduct of business.

As I indicated in the meeting introduction, the polls are open for voting on all matters to be presented and will be closed after we go through all of the matters up for vote. After the business of the meeting is concluded and the meeting has adjourned, a question-and-answer session will follow to address questions that have been submitted to the company during this meeting.

The first order of business is the election of directors, as described beginning on Page 19 of the proxy statement. The Board recommends the election of the following individuals: Mark Adams; Ita Brennan; Lewis Chew; Anirudh Devgan; Moshe Gavrielov; M.L. Krakauer; Julia Liuson; James Plummer; Alberto Sangiovanni-Vincentelli; Young Sohn and Luc Van den hove. In accordance with Cadence's bylaws, stockholders are required to provide advance notice of their intent to nominate candidates for directors. No such notice was received.

The second item of business is the approval of the amendment of the Omnibus Equity Incentive Plan to increase the number of shares of common stock reserved for issuance. This proposal is discussed beginning on Page 33 of the proxy statement. The Board recommends stockholders vote in favor of this proposal.

The third item of business is the approval of the following advisory resolution: resolved, that the compensation paid to Cadence's named executive officers as disclosed pursuant to Item 402 of Regulation S-K of the Exchange Act, including the compensation, discussion and analysis, compensation tables and narrative discussion in the proxy statement is hereby approved. This proposal is discussed beginning on Page 44 of the proxy statement. The Board recommends stockholder votes in favor of this proposal.

The fourth and final item of business is the ratification of the selection of PricewaterhouseCoopers LLP, Cadence's independent registered public accounting firm, for the fiscal year ending December 31, 2026, as described beginning on Page 45 of the proxy statement. The Board recommends a vote in favor of this proposal.

That concludes the matters to be voted on as outlined in the notice of annual meeting. I propose that the foregoing matters be put to a vote at this meeting. If you have not voted or wish to change your vote, may do so now by clicking on the link provided on the meeting website. Any stockholder who has already voted and does not want to change their vote need not take any further action. Will the common stockholders and proxies please conclude their voting.

[Voting]

It is now 1:08 p.m. Pacific Time on May 7, 2026, and every stockholder has had the opportunity to vote. As of this date and time, which will be recorded in the minutes and in accordance with our bylaws, I hereby declare the polls for online voting at our 2026 annual meeting closed. The inspector of election will complete his tabulation of the voting results after the close of this meeting.

I'll now turn the call over to the inspector of election to announce the preliminary results of the voting.

Each person nominated as director has been elected. The amendment of the Omnibus Equity Incentive Plan has been approved. The advisory resolution to approve named executive officer compensation has been approved. And the proposal to ratify the appointment of PricewaterhouseCoopers has been approved.

The final vote count with respect to the matters voted on today will be reported on Form 8-K as required by the SEC.

This concludes the 2026 Annual Meeting of Cadence Stockholders. And on behalf of the entire Cadence Board and management team, I would like to express our gratitude to all of the stockholders for their continued support. This meeting is adjourned.

It is now my pleasure to begin the Q&A session. Before I do, I will go through the safe harbor statement and Regulation G reconciliation announcement. The Q&A session, including any responses provided after the meeting on the Investor Relations website, may contain forward-looking statements. Cadence's actual results may differ materially from those expectations discussed here. Additional information concerning factors that could cause such a difference can be found in our recent reports on Form 10-K and 10-Q, our future filings with the SEC and the cautionary statements regarding forward-looking statements in our recent earnings press release.

Today's Q&A session, including any responses provided after the meeting on the Investor Relations website, may also contain certain non-GAAP financial measures. You are encouraged to review the reconciliation of any such non-GAAP financial measures with their most recent direct comparable GAAP financial results, which can be found on the Investor Relations page on our website.

Just as a reminder, on process, you may submit up to two questions by clicking on the Q&A icon at the right of the meeting screen. Questions should be relevant to the business of the meeting.

We have no further questions from our stockholders. So that concludes the question-and-answer portion of the meeting. As needed, we will post responses to any unanswered questions that relate to the business of the meeting on our Investor Relations page as soon as practical after the meeting.

I want to close by thanking everyone who participated in the virtual meeting. On behalf of the Board of Directors and employees of Cadence, thank you for your interest in and support of our company.

Thank you for participating in Cadence's 2026 Annual Meeting of Stockholders. The webcast will now end, and you may disconnect.

Ladies and gentlemen, good afternoon. My name is Abby, and I will be your conference operator today. At this time, I would like to welcome everyone to the Cadence First Quarter 2026 Earnings Conference Call. [Operator Instructions]

Thank you. And I will now turn the call over to Richard Gu, Vice President of Investor Relations for Cadence. Please go ahead.

Thank you, operator. I'd like to welcome everyone to our first quarter of 2026 earnings conference call. I'm joined today by Anirudh Devgan, President and Chief Executive Officer; and John Wall, Senior Vice President and Chief Financial Officer. The webcast of this call and a copy of today's prepared remarks will be available on our website, cadence.com.

Today's discussion will contain forward-looking statements, including our outlook on future business and operating results. Due to risks and uncertainties, actual results may differ materially from those projected or implied in today's discussion. For information on factors that could cause actual results to differ, please refer to our SEC filings, including our most recent Forms 10-K and 10-Q, CFO commentary and today's earnings release. All forward-looking statements during this call are based on estimates and information available to us as of today, and we disclaim any obligation to update them.

In addition, all financial measures discussed on this call are non-GAAP, unless otherwise specified. The non-GAAP measures should not be considered in isolation from or as a substitute for GAAP results. Reconciliations of GAAP to non-GAAP measures are included in today's earnings release. [Operator Instructions]

Now I'll turn the call over to Anirudh.

Thank you, Richard. Good afternoon, everyone, and thank you for joining us today. I'm pleased to report that Cadence had a strong start to 2026 with accelerating AI demand and disciplined execution, delivering one of the best Q1s in company's history. Our record backlog of $8 billion was ahead of plan, reflecting strong customer confidence in our AI-driven portfolio and its pivotal role in enabling delivery of their increasingly complex chip and system design road maps. Given the accelerating momentum of our business, we are raising our 2026 revenue growth outlook to 17% and expect to achieve the Rule of 60 for the first time. John will provide more details in a moment.

Agentic AI era is here, and Cadence is leading the transformation of semiconductor and system design. At CadenceLIVE Silicon Valley 2026, we took a major step towards fully autonomous chip design, pioneering the industry's most advanced and comprehensive agentic full flow platform. We introduced AgentStack, the head agent framework for our AI Super Agent, which enables knowledge sharing across the design flow and extend autonomous designs from chips to 3D-IC to systems.

Building on our revolutionary ChipStack AI Super Agent for RTL design and verification, we introduced two new breakthrough AI Super Agents, ViraStack for analog and custom design and InnoStack for digital implementation and sign-off. Together, these solutions span the entire chip design flow, creating a connected continuous learning platform that brings the industry closer to comprehensive automation.

As the industry begins transitioning to agentic AI, the need for physically accurate and highly mathematical EDA solutions become even more critical. Our agentic AI solutions are built on decades of domain expertise, proprietary data and tightly integrated physically accurate engines, delivering high fidelity results. We continue to review our platform as a 3-layer cake, with accelerated compute and data as the base layer, principal simulation and optimization as the critical middle layer and agentic AI as the top layer. As I've said before, we believe the greatest value comes from the tight coupling of these layers, reinforcing each other to deliver much better results.

As these super agents invoke our simulation, verification and implementation engines at scale, we expect them to materially expand EDA consumption and drive higher usage across our platforms. We announced a strategic collaboration with Google to optimize the ChipStack AI Super Agent with Gemini on Google Cloud. By combining LLM reasoning with GCP scalable compute, this collaboration delivers a cloud-native platform for next-generation chip development.

In Q1, we furthered our long standard partnership with MediaTek through a wide-ranging expansion across our new agentic AI offerings and core EDA, 3D-IC and system analysis solutions. Physical AI is emerging as the next big wave of intelligence as AI moves into autonomous systems, autos, drones and robotics, and Cadence is uniquely positioned to lead this transition. The addition of Hexagon's D&E leading structural and multi-body dynamics technologies transforms our system analysis portfolio to a leadership position in physical AI, enabling customers to build and train fundamentally new AI word models by narrowing the critical sim to real gap.

At CadenceLIVE Silicon Valley, we announced an expanded partnership on AI and robotics with NVIDIA. By combining our agentic AI-driven solutions with NVIDIA's advanced technologies, we are accelerating engineering workflows and boosting productivity across chip design, physical AI systems and hyperscale AI factories.

Now let me provide an update on our businesses. Our IP business continued its strong momentum, with 22% year-over-year revenue growth driven by accelerating demand of AI, HPC and automotive workloads. Growing complexity of advanced node designs and chiplet-based architectures is driving strong demands of our differentiated Star-IP portfolio across interface, memory and foundation IP.

We achieved meaningful competitive wins and customer expansions at marquee accounts, reflecting the breadth of our portfolio and more importantly, the differentiated performance of our solutions. We closed a record deal with a leading global foundry, marking our largest IP engagement with this customer to date and reinforcing our leadership at the most advanced nodes. With strong market tailwinds, focused strategy and expanding customer proliferation, we remain very well positioned for continued growth in IP.

Our core EDA business delivered another strong quarter, with revenue growing 18% year-over-year, driven by increasing proliferation of our solutions at market-shaping customers. Our AI-driven solutions, and increasingly, our agentic offerings are becoming an important part of customer renewals and expansions. Demand for our hardware accelerated in Q1, resulting in our best quarter ever, led by AI HPC customers and increasing demand in automotive and robotics.

Palladium Z3 continues to be the gold standard for emulation and drove multiple competitive displacement. Momentum on verification software grew, particularly in Xcelium and Verisium SimAI. And ChipStack generated tremendous customer interest, with a large number of evaluations underway. Led by AI-driven Cadence Cerebrus solution, our digital platform continues to gain share, especially at the most advanced nodes.

A global semiconductor design leader significantly increased their Innovus usage and adopted our digital signoff solutions, and a marquee AI infrastructure company expanded their usage of our signoff solutions in their leading-edge ASIC designs. In custom and analog, our AI-driven Virtuoso Studio continued its strong momentum in design migration and layer automation as it gets increasingly deployed by analog and mixed signal leaders seeking greater productivity.

Our System Design and Analysis business delivered 18% year-over-year revenue growth as AI-driven multiphysics simulation and 3D-IC become essential to addressing growing system challenges. We have strong momentum in 3D-IC, where our unified multi-die integrated design to analysis flow is helping customers address their rising chiplet and advanced packaging complexities. We also saw strong momentum in security and clarity, with multiple memory and advanced IC packaging customers expanding their deployments as they move to higher-speed interfaces. Customer adoption is increasing as they look to address signal integrity, power integrity and thermal challenges earlier in the design flow through deployment of a full Cadence signoff flow.

In closing, I'm pleased with our strong execution and the broad-based momentum of our business. As the agentic AI era unfolds, Cadence is leading the charge to realizing much higher design productivity, increasing design complexity, and the growing need for productivity is creating a compelling long-term opportunity for Cadence. With our differentiated solutions and expanding agentic AI portfolio, I believe we are very well positioned to lead this transition and continue delivering meaningful innovation and value to our customers.

Now I will turn it over to John to provide more details on the Q1 results and our updated 2026 outlook.

Thanks, Anirudh, and good afternoon, everyone. I'm pleased to report that Cadence delivered excellent results for the first quarter of 2026, with accelerating momentum and broad-based strength across all our businesses. Robust design activity, coupled with our solid execution, drove 19% year-over-year revenue growth and 45% operating margin for Q1. First quarter bookings were ahead of expectations, resulting in a record backlog of $8 billion.

Here are some of the financial highlights from the first quarter, starting with the P&L. Total revenue was $1.474 billion. GAAP operating margin was 29.3%. Non-GAAP operating margin was 44.7%. GAAP EPS was $1.23, and non-GAAP EPS was $1.96.

Next, turning to the balance sheet and cash flow. Our cash balance was $1.407 billion, while the principal value of debt outstanding was $2.925 billion. Operating cash flow was $356 million. DSOs were 67 days, and we used $200 million to repurchase Cadence shares.

Before I provide our updated outlook, I'd like to highlight that it contains the usual assumption that export control regulations that exist today remain substantially similar for the remainder of the year. For our updated outlook for 2026, we expect revenue in the range of $6.125 billion to $6.225 billion; GAAP operating margin in the range of 27.5% to 28.5%; non-GAAP operating margin in the range of 43.5% to 44.5%; GAAP EPS and in the range of $4.39 to $4.49; non-GAAP EPS in the range of $7.85 to $7.95; operating cash flow in the range of $1.875 billion to $1.975 billion, and we expect to use approximately 50% of our free cash flow to repurchase Cadence shares in 2026.

With that in mind, for Q2, we expect revenue in the range of $1.555 billion to $1.595 billion; GAAP operating margin in the range of 28.5% to 29.5%; non-GAAP operating margin in the range of 44.5% to 45.5%; GAAP EPS in the range of $1.07 to $1.13; and non-GAAP EPS in the range of $2.02 to $2.08. And as usual, we published a CFO commentary document on our Investor Relations website, which includes our outlook for additional items as well as further analysis and GAAP to non-GAAP reconciliations.

In conclusion, Cadence is off to a strong start for the year. We are raising our 2026 revenue outlook to approximately 17% year-over-year growth. As always, I'd like to thank our customers, partners and our employees for their continued support.

And with that, operator, we will now take questions.

[Operator Instructions] And our first question comes from the line of Charles Shi with Needham.

Anirudh, I think I have a pretty high-level question, but this is probably top of the mind for a lot of investors. We obviously learned agentic AI is probably good for EDA, good for license consumption, et cetera. But we're still hearing some concerns around AI's ability to actually write the software, and there are some doubts around whether AI can actually write better EDA-based tools like [ base ], I mean, Virtuoso universe, those kind of tools. So -- and obviously, there are always many EDA start-ups happening at the same time. And so the question is, is AI's ability to write software worries you about the defensibility of the EDA-based tool business? Obviously, once again, we understand agentic AI is good for consumption of the base tool business, but I want to get your thoughts.

Yes. Charles, thanks for the question. So I mean, there are multiple parts to this. Of course, I'm super excited about agentic AI applied to chip design and EDA. And your question is more specific to the base tool and whether AI can write those base tools. So first of all, I'm very confident in our position in the base tool and our competitive advantage, okay? And just to remind everyone, I mean, we have about 15,000 people now in Cadence and about 10,000 are in R&D. We have -- more than half of them have advanced degrees. I think more than 1,000 of them have PhDs from the top universities.

So we will, anyway, deploy AI internally like we are to write our software better. But I'm not worried that some of the party will be able to write any better base tools. So -- and our competitor of the base tool is anyway best-in-class, and I don't see any reason that will change going forward, okay?

Now what I'm super excited that we launched in CadenceLIVE is the agentic part and the interplay of the agentic tools with the base tools, the AI orchestration combined with physical accurate base tool. And that creates new opportunities for us, both in terms of TAM expansion. Because what agentic AI allows us is to sell products in spaces we didn't have products before, like RTL generation, verification, plan generation. And those products, I think will be consumed more on a subscription plus consumption model. So this is an entirely new category for Cadence.

And then in turn, like you said, agentic AI will drive more of our base tools. So I feel pretty good about this kind of 3-layer framework we have talked about and confident going forward.

And our next question comes from the line of Jason Celino with KeyBanc Capital Markets.

Great. Thank you so much. Maybe just a clarifying question. So I noticed that the operating margin guide is coming down by a little bit. Curious if -- like what are the main drivers of that, John? I know we're layering in kind of the Hexagon acquisition, but on like an absolute basis, it's relatively small entering in that OpEx. So maybe you can just help us understand the guide on the margin?

Yes. Sure, Jason. Thanks for the question. What you're seeing there is primarily the impact of including the Hexagon design and engineering business in the current outlook. The strategic opportunity there is very large, but the 2026 P&L reflects the timing of integration that we announced in the press release when we closed the deal, that we expect $160 million of revenue this year. That's in the guide now.

We expect it to be dilutive to the tune of about $0.28. The margin impact on the $160 million is kind of in the 5% to 10% range. But the dilution comes from -- because we paid 30% of the acquisition price in shares and 70% in cash. So the interest component on the -- or the lost interest income on the cash causes a lot of the dilution impact in the short term. We'd expect it to be accretive in 2027.

The -- yes, so I think the way to think about it is financially, 2026 is an integration year. And the guide includes the acquired cost base, the financing impact, the acquisition-related integration costs and kind of near-term dilution. And that's why revenue moves higher, while EPS and operating margin are lower than the February guide. So yes, $160 million.

And I think in Q1, the impact was slightly less on the EPS that we had about $20 million of revenue from Q1 from Hexagon. So only about $0.01 kind of dilution impact. So EPS would have been like $0.01 higher if we didn't have Hexagon.

And our next question comes from the line of Vivek Arya with Bank of America Securities.

Anirudh, in the last year, [ all have been hitting on stop ] or different news about chip shortages and growing kind of price of chips and just the pricing power that many of your customers have. And my question is, what affects [ new ] shortages and the fact your customers have more pricing power? What effect does that have on their engagement with Cadence? Does it restrict chip starts? Does it shift them towards higher ASP products? Just what impact do semiconductor shortages have on your growth and engagement trajectory? What has changed? And what are you observing in your customer behavior?

Yes. Thanks, Vivek, for the question. So I would say a few things. So first of all, I mean the environment is pretty healthy, both for the system companies and semi companies. So that's always good. Like you know, I mean, some of the hyperscalers and AI semi companies who are already doing well last year, but now the memory companies are doing well, even analog companies are doing well. So we, of course, want to see our customers doing well, and that creates a positive environment for engaging, especially with these new solutions we have. So that's actually a pretty marked improvement over the last 3 to 6 months. So that's number one.

Number two, the shortage is it doesn't directly -- I mean, the customer is still committed to long-term R&D road maps. And sometimes, they may like do -- like I've seen in a few cases, the customers, for example, may do multiple foundries or nodes to make sure there is capacity at a particular node or foundry. So that would directly lead to more design activity for us. So in general, if the customer is healthy because the revenue is going up, they will do not only more in the current designs to accelerate them, but also may start new designs. I think that's the second thing, I would say.

And third thing, which is more exciting for us is, as we have these agentic solutions, it can give more productivity for our customers, and we can deliver more value ourselves. And the more value we deliver, the more opportunity we have to capture part of that value. And the customers are very open to those discussions as there is more automation.

So we are actually -- like I mentioned, there's a lot of engagement with ChipStack and also the new AgentStack, InnoStack, ViraStack. There is no pushback at all. If we can deliver productivity, the customer is more than willing to engage. So that's I would say, Vivek, at least the 3 broad areas I see in the current environment.

And our next question comes from the line of Jim Schneider with Goldman Sachs.

I was wondering if you could maybe unpack your commentary on the agentic solutions, specifically around your indication they would drive increased consumption for base tools. Can you maybe talk a little bit about the pricing for those tools, how the agentic solutions are being priced specifically? And then on net, how -- if you could frame for us maybe how you might be able to capture more revenue value overall on net between agentic and conventional licenses?

Yes. Thanks for the question. So I think the opportunity is significant, I believe, and especially with agentic because what -- and this happened over the last, let's say, 6 to 12 months, in my opinion, and more so in 6 months is -- not only the agentic tools have evolved, but agentic tools are able -- we can embed skills in them so they can do a lot more automation.

For example, we launched ViraStack, which is analog automation. Analog has been a long problem to automate, right? It's very difficult to automate. But now with these agentic flows and skills, we can automate that. So what does that mean in terms of pricing or how these things are consumed?

So first of all, like I said, this kind of automation was not possible before. So all this work used to be done by the customers themselves, right? And in that case also, I talked to 1 big customer. Like, for example, they said for analog or even for digital, every new design, they require 2x more engineers. And anyway, it's not -- it's like unrealizable headcount growth because they can't hire 2x more engineers every time.

So the way we plan to monetize and the early signs are positive is that, first of all, we'll sell new tools that we never sold, which is more like this was manually done by customers like doing analog design or doing RTL. So that will be priced as a subscription plus consumption model, very similar to other kind of leading AI tools. So that's a completely new category for Cadence. And that will kind of bend the headcount curve for our customers, but the expected headcount curve was never realizable anyway.

So this is the history of automation, as you know, in EDA. We always need to do that. But this time, we can do that with the agentic kind of AI flow. And then once the agent runs, like when a user designs a chip -- and this is pretty common, right? Like let's say that chip has 100 blocks, just to keep it simple. And there are 100 engineers, 1 engineer is running 1 block. So 1 engineer will run like 1 or 2 experiments, he or she, to see which settings or which design is better.

But when the agent runs those blocks, they may try 10 or 100 variations of those things. And anyway, AI does a lot more exploration than a human would do. So not only agent can give more productivity, it by nature runs more of the base tools. So that's why if you look at -- our usage of base tool is going up pretty significantly in this kind of environment. So this is the 2 ways -- and those environments is a traditional business model, but -- in the base tools, but there will be more demand for it. And then the new business model, which is more automating which was manual with agentic flows.

Yes. And I would just add, Jim, that what we saw from Q1 is -- I mean, the overall pricing environment has improved. Pricing obviously remains value-based with us. We provide tremendous value to our customers, especially with our agentic flow. And we stand to benefit from our customers' success in that area. Also, any shift that you see from customers' labor spend to automation, that's likely to be irreversible and likely to accelerate over time.

And our next question comes from the line of Siti Panigrahi with Mizuho.

Great. I want to switch to the IP business. Anirudh, you talked about IP entering now, third year of strong growth. Could you give an update like what you saw in Q1? And are the HBM, LPDDR6 and all that remaining still the key drivers? Or -- and the newer foundry like [ Rapidus ], Intel Foundry, are they contributing meaningfully to the IP demand yet? And John, just to clarify also on your EPS guidance, you said $0.28 dilution, but you lowered only $0.20. Just want to clarify that your organic basis, you raised by $0.08 EPS?

I'll take the last part first. Yes, yes, we did. We raised by $0.08.

And it is a great start to the year, okay? And not just in IP across the board. And I was looking at with our team. I think this is 1 of the strongest raises we have had in Q1. We only gave you guidance in February. So 2 months later, I think is one of the strongest raises we have had.

Now all the businesses are doing well, and especially IP is off to a great start, okay? And I think it will do well going forward from what I think I see. And there are at least 3 big reasons in my mind for IP growth. And like I said, it's the third year now. So we don't like to talk about things too early, but after 3 years of strong growth, I think that is a good trend.

So the first thing is our IP quality and performance is just better. We have a new team, just the performance, just -- because these things are standard-based IPs, right, like DDR or PCIe. So the spec is same, but if our power area is better than the competitor or what the customer can do, then they will buy our IP. So the most promising thing to me is because the strength of our R&D team, our PPA is better. And that is leading to a lot of competitive wins at pretty significant major customers. And I highlighted some of them in CadenceLIVE. So these are like really big kind of marquee names. So that gives me strength that the team is operating well. So that's number one.

Number two, our portfolio is expanding, like we have highlighted with -- like HBM. And some of it is organic. Some of it is acquired, like HBM, we acquired from Rambus and then we improved it. But UCIe, which is a critical chip-to-chip technology, was all developed organically, okay? So the second reason is that our portfolio is expanding.

The third reason is these new foundries okay? And it's very encouraging to see. Of course, we want to make sure we are best-in-class in TSMC, which is the leading foundry. But now there are at least 3 other major foundries, as you know, Samsung, Intel and [ Rapidus ] at advanced nodes and then Global and others at mainstream nodes. So the amount of design activity with AI and number of increasing foundries requires more IP.

So that's why I'm actually pleased to note today like in the prepared remarks that we had a pretty significant IP deal, one of the largest ones at a leading global foundry, okay? And just to clarify, that is not Intel, okay? We are actually pleased with our discussions with Intel, with [ Liban ] team on 18A and especially on 14A. I think Intel realizes they need to invest more in 14A, and this time, be more ready because the availability of IP and EDA solutions as 14A is critical as they go talk to their customers. So we are making very good progress with Intel. And we will have -- soon, we'll have more to say on our engagement with Intel.

But I'm also pleased with this engagement with the other global foundry. So overall, IP growth seems robust. And I'm very pleased where we are. And we're already -- always very strong in EDA. But historically, last few years, we have not done as well in IP. But right now, I think we are very well positioned and also well positioned in SDA.

Our next question comes from the line of Joe Quatrochi with Wells Fargo.

Yes. Maybe just to kind of follow-up on the discussion really on EDA. I mean, I guess, would you take a step back and you think about EDA's share of R&D expense. And clearly, we're seeing an acceleration of R&D expense across a number of different companies. How should we think about EDA's contribution to that or a percent of that? And where could that go given the value maybe you're providing from AI? Because we're also seeing, right, memory costs are increasing, things like that, that also need to flow through that R&D line.

Yes, good question. And we have to observe it closely, right, as we rather like print things than kind of predict what will happen because it's better to show than to -- but as you know, historically, we have said EDA used to be 7% of R&D and now it's more like 11% of R&D. So it has gone up, and R&D spend itself will go up significantly.

But I think there is a real potential, especially with agentic AI for that 11% to go up. And all the big CEOs I talked to, they are not only willing, they want to see that happen. They want to invest in more automation and compute to make it happen. So I'm pretty sure right now, I think it will go up. Now how much it will go up, we will see, right? But I think there is a meaningful opportunity for automation to be a higher percentage of R&D, plus R&D itself to go up.

Our next question comes from the line of Ruben Roy with Stifel.

Yes. John, I want to go back to the operating margin discussion. It's great to see that you guys are targeting a Rule of 60 by the end of the year here. Just thinking about that though, it's driven on revenue acceleration. Obviously, we've got the Hexagon integration costs here.

But how are you thinking about the operating model relative to operating margin as you get over $6 billion in revenue? Does the operating model look a lot different than the $5.3 billion? Is this sort of a 43% to 45% range, how we should be thinking about the operating margins? Or -- and I ask that because, obviously, you're investing in agentic AI and other sort of new product areas. Just wondering if you can give us a little bit of an idea of how you're thinking about the operating margin structure at this revenue run rate longer term as you integrate Hexagon.

Yes. Sure, Ruben. Thanks for the question. Yes, I think when we look at our like organic incremental margin is closer to 60% these days than 50%. And as we get our arms around these acquisitions, it typically takes us 12 to 18 months to improve the profitability up to kind of something close to our expectations at Cadence.

And I would liken the profile to the [ way ] BETA. So in '24 and '25, you kind of had an operating margin profile where we had the dilutive impact of the BETA acquisition in '24, but then margins improved dramatically in '25 as we got the synergies and we've got the benefits of making that more profitable.

I would expect a similar pattern for '26 and '27 when it comes to Hexagon. We have a slight headwind in the short term, but there's plenty of opportunities to improve the profitability there. And also with the benefits that we're seeing in terms of customer engagement, accelerating on the agentic AI front, I think there's even more opportunities to stretch that incremental operating margin going forward.

Our next question comes from the line of Harlan Sur with JPMorgan.

If I take your 2Q guidance and look at your implied second half guidance, the average quarterly revenue run rate in the second half is actually slightly below the 2Q level. Is there some lumpiness in the Hexagon business in the second half maybe moving customers to multiyear license agreements? Or is it due to some lumpiness in the core business, maybe a more first half-weighted hardware or IP shipment profile?

Yes. Thanks for the question, Harlan. Yes, sure, the first half is very strong. And the second half, I described is containing appropriate prudence. Your comment on Hexagon, Hexagon's D&E business is correct. They are more kind of first half weighted in terms of their profile. When I looked at last year's revenue, the -- for Hexagon, the I think Q3 and Q4 were their worst 2 quarters of the year. They tend to have a lot of early year kind of dated contracts.

But overall, I think the second half -- I mean, it doesn't -- Hexagon doesn't impact the first half, second half that much. It's really -- I think we had such -- as Anirudh said, Q1 guide represents one of the highest rates we've had at this time of the year. And we normally like to wait until we had 2 quarters under our belt to raise the guide. We couldn't help but raise the guide given the strength of Q1 bookings and the strength we saw across the board. So we just wanted to wait until July to update the second half.

Our next question comes from the line of Lee Simpson with Morgan Stanley.

Great. I just wanted to ask about physical AI. I mean you've made some pretty good acquisitions. You now announced collaborations, especially with NVIDIA. So I'm just trying to get a sense for the momentum here and what really is still the early years in this breakout. And I think, in particular, the take-up of your emulation tools, especially as it relates to closing the sim to real gap in robotics and probably even self-driving chips as well, whether or not that's going to really lead to an outsized value capture for Cadence? And when do we actually see this in the numbers as well?

Yes. Thanks for the question, Lee. So I mean, like I talked about it forever now that we look at this thing as a 3-layer cake, right? And there are multiple slices of the cake, and the first slice was data center AI or infrastructure AI. And the second big slice is physical AI. And of course, I've said this for 5 years now, but I believe physical AI will be bigger than data center AI by a long shot because you're talking about like trillions of dollars of product opportunity. And it will reconfirm the data center layer with the data center slice because to deploy, for example, an AI model in the car, you need to train it on the data center anyway. So I think it will even help the data center slice.

Now for our portion, yes, we made this acquisition we are super excited about, and we have this training flow, forward models and also more complete simulation environment. So what is exciting about Hexagon is with a combination of our previous technologies like Millennium and Cascade and, BETA, we do have finally a complete solution for physical AI in the middle layer, kind of principal simulation and optimization layer. And then that can be used to do these word models, which will be different in the top layer.

But other thing I want to emphasize, apart from the SD&A and the AI part, that physical AI itself will drive larger silicon design. So it is also good for EDA and IP. And this is -- you're starting to see that, of course, companies like Tesla mentioning that they don't have enough silicon because of physical AI. So physical AI not only is good for SD and AI, it is also really good for silicon. And it also is the sweet spot of Cadence because Cadence always had both analog and digital solutions. And that's why we are always good with all the major semiconductor companies for automotive. And now with all the system and OEM companies for automotive and as that translates to drone and robots, it will also turbocharge the silicon business. That's why I have been always been excited about physical AI, not just for the AI and SDA, but also for EDA and IP.

Our next question comes from the line of Gianmarco Conti with Deutsche Bank.

Perhaps on hardware, another strong quarter, of course. But as we think about the next refresh cycle for Palladium and Protium, historically, you've roughly been on a 2-year cadence. Should we expect [ Z4 X4 ] within the next 12 to 18 months? Or is the bar to upgrade higher now, given how recently customers absorbed the third generation? And perhaps related, are you seeing any of your own agentic AI tooling materially compress the internal [ hardware ] development time lines to the same extent that customers are reporting that same next productivity on RTL?

Yes, absolutely. Great question. So first of all, like I said, we have -- most of our headcount is engineering, right, whether it's R&D or customer support. So we always want to use our own product in both our hardware groups, which is the significant design team. We do both software, hardware and all the system design in Palladium and Protium. And also, just to remind you in our IP team, it's a great -- they're working very well together. Our IP team and EDA teams. Because IP, we have so much demand. And instead of, again, increasing headcount, we're always sensitive about how much headcount we'll increase, and we are increasing headcount in all areas including IP, but we can make them a lot more productive with agentic AI.

Now on the hardware part, yes. I'm very pleased. I mean, it's a remarkable start to the year. Our competitive position is amazing. We are the only company that does its own chip, as you know. We have at least a 10-year lead in that in Palladium. And then Protium also is doing now in which we use the FPGA solution.

Now just to be clear, we always design next-generation systems. And because we control the whole stack, including the system design and silicon design, one thing to remember is we will do it much faster than what the FPGA cadence will be. FPGA companies will also do next-generation FPGA designs. But because we are own chip, we do our own design, it will be much faster than FPGA.

So what that means is the lead of Palladium over FPGA systems will only continue to increase as we introduce new products, okay? But I'm not going to get into like when we're going to introduce new products because the current products are doing amazingly well. Of course, we are designing Z4 and Z5. But what you have to remember is the current Z3 system has the capability to design 1 trillion transistor systems, okay? And right now, the biggest systems in the world are 100 billion to 200 billion transistor. So we have a lot of leeway. The industry is supposed to reach 1 trillion transistor by 2030. One thing I'll assure you is we'll have a Z4 system before 2030.

So there is no issue of whether Z3 can handle the capacity and requirements. So we're just happy to work with our customers. At the same time, we want to assure our investors and customers, we have a very, very good road map on hardware systems.

Our next question comes from the line of Jay Vleeschhouwer with Griffin Securities.

Anirudh, now that you've completed Hexagon MSC acquisition, it would appear that you are the fourth largest non-EDA simulation company, let's call it industrial simulation with multiphysics. Your share is perhaps 1/10 of that total market, again, aside from EDA simulation.

So the question is, now that you've assembled all these pieces, invested over $5 billion over the last 5 or 6 years, can you speak in some detail about what your principal technical and/or go-to-market objectives or executables are going to be for the next year or so? Synopsys talked about what they're doing with ANSYS, perhaps you could do the same for your pieces? It also seems you're becoming a little bit more vertically integrated in go-to market with the acquisition of a longtime channel partner. So maybe talk about some of those critical elements here to grow your revenues and share in that business?

Yes, Jay, that's lost there, right? There's a lot there. So let me try to unpack some of it. I'm sure we can talk more if I don't get to all the pieces there.

Well, first of all, we are satisfied with the scope of our SDA business now after this acquisition. So I mean, this is a rough number. So I think it will be roughly $1 billion of run rate. And what is more exciting to me is that it is focused in the two important areas of SDA. I'm a fan of SDA for a while now, I don't know, maybe 8 years now. But not all SDA is created equal, okay? To me, we want to do the part of SDA that is either growing well or is closely related to EDA.

So the part of SDA that is closely related to EDA is, of course, 3D-IC, okay? So we have an inevitable position in 3D-IC with Allegro being the leading packaging platform, and then we completed that with Clarity and Sigrity and Celsius. So all the thermal electromagnetics. So at Integrity, so I'm pretty happy with the 3D-IC portion, which is like the closest to chip design, the part of SDA that is closest to chip design and the part that is growing the most because of AI.

Now the other part now with Hexagon is all this physical AI and for design of cars and robots. So that, with this acquisition, is complete, and we can do a much better integration of that part of SDA. And there are multiple things happening there, okay? They are at least 2, 3 key things. So first thing is -- we will integrate the whole solution. I know you asked me this before, when will you integrate. So I think now that we have all the pieces of critical mass, this is the right time to integrate because we have CFD now, we have structural, we have multibody dynamics, we have pre and post, okay? So we have a lot of effort to make a full flow solution, integrate them. And I kind of hinted at that at CadenceLIVE.

The other thing, the way to integrate these solutions, which is true for EDA, what will be true in this area is the agentic flow. So you will see from us, agentic flow to do system design. And that part of the market has not seen that much -- it's even worse automation than chip design that I had a lot of automation. But there will be agentic flow which will integrate all these things in a better way.

The second thing we will do is that there is a lot of room for improvement of these solvers. And especially in our history of improving the base solvers, adding GPU acceleration, adding physical AI or AI surrogate models. So for example, there is a potential for at least the order of magnitude improvement of performance of these new solvers. So that's the second thing we'll do in terms of R&D.

And third thing, what I'm also pleased with Hexagon is we did get like a good go-to-market team. That's one area we have not been as strong because we were -- most of the others was mostly organic. And we did move some of our people into go-to-market. But with Hexagon D&E business, we get a much stronger go-to-market team. And then as we mentioned, we also acquired some resellers to strengthen go-to-market, okay?

At this point, I'm very confident of our R&D solution, and it will get improved by agentic solutions. It will get improved by speeding up the solvers. But we also need to invest in go-to-market, and Hexagon gives us a good start. So you will see that, too. So these are the 3 kind of focus areas of improvement of SDA.

Our next question comes from the line of Kelsey Chia with Citigroup.

Anirudh, you mentioned that the AgentStack helped address talent gaps for chip designers. It sounds like the AgentStack that adoption is just accelerating from here. Based on your composition, is that the case? Or are you seeing cases where customers prefer to build or use their own agentic stack versus adopting Cadence's? And so as Cadence is able to sort of charge for AgentStack or the increased base licenses as an incremental add-on within an existing [ fee ] contract? Or is that monetization tied to renewals?

Yes. Thank you. There's a lot of good questions there. Okay. So make sure I -- and I'll start, and John can add to that. No, first of all, I think just to be clear, the customer will always write their own agents as well, if I understand the first part of your question. Even in our pre-agentic flow, we would have given a lot of flexibilities to our customers. We had a [ tickle ] or a Python interface to our tools, and they would always have their own flows. I mean this is natural for big customers. I mean these are who's who of tech companies. So they always want to have some differentiation from 1 flow to the other. So -- and that will happen in the agent world itself. So I think most of our customers are writing some of their own agents.

But the key thing is that the critical agents, okay, like these big super agents we talked about like RTL design and verification, analog design and physical design, these are like super categories. And also, the value of the agenda flow is not just in the agent itself. It's always the coupling of the agent with the base tools. Because we operate the agent at a much lower level of interaction, this API calls, which is not possible for customers to do.

So what has happened as an example, as we showed InnoStack or ViraStack and ChipStack to our customers, they realize, oh, there's no point writing these kind of agents, okay? So they would rather use the super agents we have because not only we are good in agentic flow, we are good in the coupling to the base tools.

Now they will still write some agents to customize things which are specific to them, and we naturally welcome that. And the AgentStack allows the environment to -- for the customer to write its own agent, but also the customer to write its own skills. We want the customers to write their own skills in InnoStack, which may be specific for a part of design. So this has always been our strategy to be more open to customer kind of customizing their own environment, okay? And I think the second question is on renewals versus new -- I mean, it's a combination of that always. John, maybe you want to comment on that?

Yes. Yes. Thanks, Anirudh. Thanks, Kelsey. Our subscription model remains the anchor arrangement with our customers. The add-on monetization then comes incrementally through agentic workflow products that are kind of usage-based or consumption-based for capacity and through our token and card models.

What's different about agentic AI is that it doesn't replace the core EDA engines. It calls them more often and it calls them intelligently. So the monetization opportunity is twofold, really. So you've got like the new agentic workflow products, and then you've got the increased usage of the underlying base tools through more exploration, more verification, more optimization and more compute.

Now that said, we're obviously being disciplined in our 2026 outlook. We're not assuming a sudden step function in AI monetization in the guide, but we do believe agentic AI expands the long-term growth opportunity for Cadence.

Our next question comes from the line of Andrew DeGasperi with BNP Paribas.

I just had a 2-part question. One is, marquee, I think you called out in the prepared remarks that a marquee AI infrastructure company expanded the use of signoff solutions. I just want to clarify, was this a cloud provider? And then second, at CadenceLIVE, you discussed about physical AI in terms of the time line of adoption being around 2 years. But yet, you called out that automotive and robotics companies have adopted hardware. I was just wondering, does this mean that, that physical AI time line has been brought forward? Or is this just a natural evolution of how these new markets will adopt EDA? And if so, when would we see that kind of software benefiting from that?

Yes, I think with physical AI and also agentic AI in general, I mean, yes, I've said for a long time, 2 contract cycles, and that is generally true. Though I think because of this new category of TAM expansion, which is more labor productivity related along with the base tools, I think there is a potential that the monetization of agentic AI could happen sooner than 2 contract cycles, okay?

I don't want to predict too much. And like John said, we are not putting it in our guide. But I think definitely, the more opportunity is there because of all the shortages, because all the build-outs because of physical AI. So we are -- and like the previous question, we always can add in the renewal, but we always have capability to do add-ons, which we have already seen, okay? So that's what I would like to say.

On the signoff, we are very happy. You know what has been the leading solution for implementation, especially at TSMC and now increasingly with Samsung, Intel and [ Rapidus ]. But signoff is coming on strong at TSMC and other customers. And we are working with all the leading AI players. And I think the one we mentioned specifically is a major kind of AI infrastructure/ASIC company. And we are glad to see that adoption.

Our next question comes from the line of Gary Mobley with Loop Capital.

John, I think, if I'm not mistaken, 2026 is going to be a low renewal period by then. I mean, existing long-time customers scheduled to renew this year, kind of like 2022 was. And so was the strong bookings in the first quarter a reflection of some add-on sales as salespeople trying to meet their quota? And do we expect that type of behavior to last through the balance of the year?

Thanks for the question, Gary. Yes, I mean, 2026 is kind of later than 2025 for actual renewals on an annual value basis. But we often see that, that's the -- those are some of the strongest growth years for us because of all the add-on activity. Yes, we were really, really pleased with the Q1 booking strength, and it was right across the board across all lines of business.

So Gary, I mean, it bodes well for the year. But look, it's just 1 quarter. As you know, we like to wait for a couple of quarters before taking up the guide in the second half. And although the last few years, Q1 has been strong. And this one has been very, very strong. So we had to take up the guide at the end of Q1.

Our next question comes from the line of Clarke Jeffries with Piper Sandler.

I just wanted to ask around the largest IP arrangement today with the global foundry. Was it really the extension of that agreement to additional nodes the scope of more content or the addition of agent ready AI flows that make the biggest difference to get that to the largest arrangement you've ever seen?

Yes. That's a particularly IP contract. So that 1 particular is focused on IP. And the 2 things that drove it is that it is a new node, new advanced node, more specifically 2-nanometer and more content in IP because we have a much broader portfolio.

Our next question comes from the line of Joshua Tilton with Wolfe Research.

Maybe just a 2-parter, a little unrelated, so I apologize. But anything to call out on what drove sort of strong quarter for China? And then maybe just a second part to that. Can you help us just bridge what is driving such a great organic raise for the full year relative to the organic beat in the quarter? I know you mentioned the record backlog. Is there anything 1 level deeper you can give us? Especially in the context of it sounds like you're trying to tell us that even though you raised by a pretty solid amount, that there still seems to be some conservatism in the guide for the second half. So any help there would be greatly appreciated.

Sure, Josh. Thanks for the question. I'll take this one. So Josh, yes, China, it was 13% of Q1 revenue. The -- and that was just kind of broadly consistent with what we were expecting. Yes, we still expect China to be about 13% for the year. I think it can be lumpy from quarter-to-quarter. So I think the year-over-year comps probably look generous because Q1 in 2025 wasn't that good in China. So the -- being 13% revenue in Q1, probably the growth rate looks strong, but it's just -- it's a really important region for us that -- yes. And we were very, very pleased with the 13%.

The -- in relation to the guide, yes, I mean, we're -- look, Q1 was a very strong start to the year. We exceeded all our metrics. And I guess when we back out the Hexagon, the $160 million of Hexagon and the $0.28, we're basically raising the year by $65 million at the midpoint for revenue and about $0.08 for EPS. Also on the cash flow front that's operating cash, the way we paid for Hexagon. The reported guide includes approximately $180 million of pre-close Hexagon tax liabilities that are economically part of the acquisition consideration but are classified in operating cash flow. I think the -- just the geography and the accounting forces us to put it through operating cash. If you adjust our operating cash guide for that underlying -- for that [ pre ] Hexagon tax liability that we're paying, the operating cash flow outlook is approximately $2.1 billion, which should be about $100 million above our original guide.

So there's a lot of strength we saw across the businesses. So the $65 million is what we took revenue up by, but we're seeing $100 million extra in cash, but there's potentially strength in the second half, but we thought it was too early to raise the second half right now.

And our final question comes from the line of Blair Abernethy with Rosenblatt Securities.

Just want to ask about the Millennium platform. How is the adoption going there, Anirudh? And just in general, the health in some of your non-semi verticals like automotive, aerospace, industrial equipment, and so forth, just any commentary around that would be great.

Yes, absolutely. So yes, Millennium is doing great. I don't know if you saw, Jensen was there at CadenceLIVE and did a nice autograph on [ Millennium Box ]. So we are pleased with the partnership with NVIDIA there.

And I mean, there are 2 ways to 2 kind of high-level applications. We are working on this kind of CFD or SDA application for a while. And that's going well, especially in auto, and also in drones, okay? There's a lot of what cascade acquisition we made is very good at very high accuracy CFD, which also applies to aerospace and defense. So -- so there is autos, but also A&D is Millennium offtake. And we have several customers. Some we can talk about, some we can't, okay? So that's in the traditional Millennium.

And the other part, this year, like I mentioned in CadenceLIVE, we have all kinds of EDA application now on Millennium. It's super exciting. And the most exciting part of EDA application in Millennium is 3D-IC signoff. Because right now, the biggest issue is the complexity of the 3D-IC systems. Not just to design them, which we can do in Integrity and Innovus, but to sign them off. So there's this huge system that need to do thermal simulation, electromagnetic simulation, power delivery simulation. And they are more naturally like a matrix without getting too technical. They're closer to a matrix multiply and numerical solver, which is great for GPU acceleration.

So right now, I see Millennium as applying to more traditional areas like autos and then new areas like aerospace and drones and then applying to 3D-IC signoff. So we are super excited about the Millennium opportunity along with our traditional hardware systems.

And I will now turn the call back to Anirudh Devgan for closing remarks.

Thank you all for joining us this afternoon. It's an exciting time for Cadence as we begin 2026 with product leadership and strong business momentum. And on behalf of our employees and our Board of Directors, we thank our customers, partners and investors for their continued trust and confidence in Cadence.

And ladies and gentlemen, thank you for participating in today's Cadence First Quarter 2026 Earnings Conference Call. This concludes today's call, and you may now disconnect. Goodbye.

Okay. Good afternoon, everyone. Welcome to San Francisco. We're on the stage with Anirudh Devgan, CEO of Cadence. Welcome, Anirudh.

Maybe if I just kick things off. I mean, no pressure, but we were in here, all of us earlier. Jensen gave a call out to Cadence, talking, I think, really around the sort of emulation space. And of course, you talk a lot about that sim to real gap and how you guys can plug that. So maybe help us understand, how does that segue? And how is this an opportunity for you guys, looking at physical AI in particular?

Yes, yes. Well, thank you. It's good to be here. Thank you for the interest. And we love working with Jensen and NVIDIA too. We have a long-term partnership with them, of course. And then I think what I've talked forever -- so sorry for people who are familiar with this -- is like the 3-layer cake. Because there's all this worry AI is going to replace software or something like that. And the thing is that there are different kinds of software, right? There's a whole range of software.

And for us, the reason I call it a cake and people say, like, "Why do you call it a cake? It's like a Cadence bakery or something?" I'm not a good cook, by the way. I'm a horrible baker, but definitely not a great. But the thing is unless you are like 2 years old, normally, when you eat a cake or you consume a cake, you consume all 3 layers of the cake together. At least that's what I do. So -- and then you bake it together.

And then what are the 3 layers are -- is AI at the top, and I can get into more detail, which is more like data science algorithms, AI at the top. The middle layer is more ground truth, physics and the good old stuff of how things actually work, like molecules and transistors. And then the bottom layer is compute and data. And now it's accelerated compute and data with NVIDIA and others.

So -- and then people who graduated last few years, they say, "Well, I just need AI. Give me like input and output. I'll create a model, and it will do everything." And people who graduated 30 years ago say, "Well, what you know is the real truth," how transistors actually work and all that. But the reality is that you don't need to take side of that, any side -- it's both together, and then running on top of data and compute. So by the way, this is going to happen in all markets. All markets. And then the slice of a cake is, of course, domain dependent. It could be chip design, it could be self-driving cars, it could be robots, right?

And now -- so first thing to remember is, in our case, the middle layer is very scientific, numerical. Physical -- if you're designing like 100 billion transistors at 2-nanometer, it better be accurate and you really do need to know the fundamentals. So -- and then when AI runs on it, it uses more of the middle layer, which is what I think Jensen is talking about also. So when you do more physical AI or Agentic AI, so there are at least two.

And then there are -- I also talked about this for years about the 3 main slices of the cake, okay? And so the first slice is what is happening now, which is driven by data center, deployed in software. So that would be a lot of like even for us, like chip design flows or other flows. But I always believe that -- and for years now that the second slice will be huge, which is physical AI, which is cars, robots, drones. And then the third slice of the cake would be sciences AI, which is, of course, life sciences, material sciences.

So in all cases, the 3 layers are different. So in case of the current one, of course, we have Agentic -- LLM-based Agentic AI at the top level, our basic kind of tools at the middle level and then GPUs at the bottom level, right? So that's how we -- so that, we can talk more about. We have all these new products for AI.

Now the second part which you asked me, which is more specific to physical AI. So because physical AI will be huge, right? I mean, we can talk more about cars, drones and robots. So we're also building a flow on that. So there, there is a sim to real gap. There's more opportunity for simulation.

Okay. Yes. It sounds like a huge cake because that's 3 industrial revolutions, one after the other.

Yes, it's 3 by 3.

Yes. That's phenomenal. So maybe if we go -- first of all, I should have read out a disclaimer, so I will apologize. So let's imagine we've done this from the top. Today's discussion will contain forward-looking statements, including Cadence's outlook on future businesses and operating results. Due to risks and uncertainties, actual results may differ materially from those projected or implied in today's discussion. I apologize, that's on me.

If we think about, however, the wider ecosystem in relation to what you talked about, silicon physics and then AI as the 3 layers, what is your position, particularly around that physics layer that we just mentioned?

Yes. I mean, we have a great -- I mean, first of all, again, the physics, I mean, the ground truth is different in different slices of the cake. So if it is chip design, of course, we have the strongest position. See, Cadence, you have to remember, has the biggest portfolio in core EDA, core chip design. Digital, analog, verification, packaging. And you are seeing that in the market, right? Competitively, we are doing great in our core business, okay?

So now on the second slice of the cake, physical AI, we had to do some M&A. So the one we did recently was Hexagon. And the reason I did that was in the second slice of the cake, like I was saying -- so if you're going to build -- so in the second slide, the physical AI model, the AI model is different. The AI model is no longer an LLM model. It's a world model, like W-O-R-D and W-O-R-L-D. My wife says my Ls are difficult to -- so it's a world model in the second case.

So in the world model, there's not enough data on the Internet. The LLM model, you can train with data in the Internet. But in the world model, you need to generate data, synthetic data. And so either you capture the data by sensors, but that's too difficult, right? It takes too long to do it. Or you simulate it. But if you simulate it, you have to make sure that it is very accurate. So that's called the sim to real gap.

So in that case, Hexagon had the most accurate robotic simulator with Adams. So we're going to put Adams in that loop to improve the accuracy of simulation for physical AI. And then the third part, which is the silicon, is going to be different because the silicon is more mixed signal and more low power. The silicon for physical AI is different than the silicon for data center AI. So -- like silicon used in cars and robots. So that is actually also in Cadence's core strength because it's more mixed signal and lower power.

So we have historically worked with all the big semi companies that make auto chips, right, or like these kind of embedded chip, mixed-signal chips, and then now with OEM players like Tesla and Rivian or BYD that are designing their own chips. So again, with the physical AI, the critical thing in these slices is that all 3 innovate together, and we want to make sure we are well positioned for that.

Got you. Okay. Very clear. Maybe take us back to some of the discussions we've heard in and around this whole conference is the sort of worry that AI volatility is disrupting traditional software business models. And maybe if you could just help us understand how you would stand apart from that disruption? And where indeed you would be moving to change your business model or augment it against this volatility?

Yes. I think the one thing to remember is that I think for us, it's not disruption, it is amplification. And I can talk more about -- and the question is how do we monetize that amplification. Because what AI will do is that it will naturally drive more usage of the middle layer. The top layer drives more usage of the middle layer.

And there are a few things that are different for chip design. Because if -- so what I think what people get worried is if something is 10x more efficient, does it reduce the usage, okay? But in chip in EDA, going back to 20, 30 years ago, we are 100x more efficient. It just -- when I was in IBM in the late '90s, we would have 500 people design a CPU in 5 years. It is a real thing, by the way, okay? And Intel, same thing or [ DEC Alpha ]. Now you can have 50 people, sometimes even less, design a CPU in 6 months. So it's 100x more efficient.

And we have even more usage of our tools. The reason for that is that the workload is exponential because our customers are designing bigger and bigger chips. If the workload -- that argument only applies if the workload is constant. I don't know, like you are doing something like -- I don't want to pick on anybody, but like if your workload is not growing, workload is linear to the number of people, for example, then if you are 10x more efficient, you may use 10x lower.

But in our case, we are doing 3-nanometer chips now. It will be 2-nanometer, then 1.4 nanometer, then 1-nanometer. Then there will be 3D-IC. So there's a wide projection in 5 years, the chip size will be 5x to 10x bigger. Complexity will be 20x, 30x bigger. So you need that 10x to keep up because our customers don't want to hire 30x more engineers. So AI will modulate the headcount growth for sure, but instead of 30x, it will be 2x, 3x. And the remaining will be with automation. And this is the history of chip design industry because of Moore's Law. So this is a very different thing.

So one part that is different for us is that the cake, the middle layer is critical. You have more scientific software. Second part that is different is the workload is exponential. So as a result, like we have customers that will spend months optimizing things to get a few percent better power because they're going to have like millions of these devices. So as things get more efficient with AI, they run more things. Like NVIDIA, they will run more optimization to improve the GPU further or more optimization to improve the mobile CPU or more optimization for the car. So if you look at the license count, I think that is going up nicely. We just have to make sure that we get our value for that. And the way to do that is to demonstrate the value to our customers.

Got you. And talking about more optimization and some of the things that are different. A few weeks ago, you did launch the ChipStack Super Agent. So maybe just help us understand, how does that differ from the GenAI tools that you had out in recent years? And how could that accelerate the growth for you in the next couple of years?

Great question. So I'm super excited about ChipStack. This is a new product category, okay? So if you look at LLM or Agentic AI, what is the biggest use case right now? The biggest use case in the general market is coding, right? C, C++, Java, you can just talk to it and it writes code. Now -- which is great. Now one issue is -- and we use it internally for our coding. We are a software company. So we can use that to become more efficient in R&D.

Now one issue is if you write like 80% of the code is good and 20% is not good, then you spend a lot of time figuring out which 20% is not good. So this is one issue with LLMs. Now if you go to chip design, it's actually the opposite. If you look at our history over 30 years, chip design also has a language. I don't know if you -- for those of you who did engineering in undergrad or -- there is RTL is registered transfer language or system dialog is the language that all our customers will define the chip with, okay?

Now so far, they manually write that language. So they not only write the design manually, they also write the verification plan manually. And then we have all kinds of tools to verify that it is correct. This is our core business is that we have automated the 80%, 90% of once you have RTL, how to design a chip. Because it was so -- these things are so complex and expensive, you had to automate that.

Now what we have never done is ability to write RTL or test bench. But now with ChipStack, we can do that because that is the core engine of LLM. So we have these -- and we have a new method -- and ChipStack has a new way of doing it using a mental model and knowledge graph. So it's a much better use of LLMs. And we can write the RTL, and then we can write the test bench because verification is as important as design. So this is an entirely new product category where there was no automation. So there's a lot of customer pull to deploy that. And then to verify that this RTL or test bench is correct, of course, it runs a lot of the middle layer or the base tools. So then we will monetize as an agent plus the use of the base tools.

So the optimization here is really just in relation to the test benching as verification and the process flow. But equally, it's a pull-through on the base software layers as well on your tool sets. Pretty clear.

Maybe let's jump to IP because it's been something that's been something of a focus. And I've heard you earlier today talk about this as being super hot as a category area, maybe not so much for others in the field. But maybe help us understand what are the dynamics behind the growth in IP with Cadence at this point? And is this supported by recent acquisitions? Or is this a moment in time perhaps driven by SerDes and other standard libraries?

Yes. So IP is doing well. And actually, this is the -- we normally don't talk about it if it is a onetime thing. We only talk about it now. So it's the third year of very good growth we will have. So that's our style anyway. We don't want to say things unless they are fully verified. So I feel very good about IP. We are at third year of very strong growth.

And there are multiple reasons for that. One is that our products are better. We finally have a good team. We always say team, technology, customers, right? So we have a great team finally in IP. So our products are doing pretty well, especially in advanced node TSMC, which is the most exciting part of the market.

Our portfolio has grown. That's the second reason. And more on AI HPC side. So there, we want to focus on some high-value IPs like HBM. Now that, we did acquire from Rambus. That's a great acquisition. But then DDR is organic, UCIe, PCIe, SerDes. So I think the portfolio is better. That's the second reason.

And third reason is there are more and more foundries. Like -- of course, TSMC is doing remarkable, but there are at least 3 major advanced node foundries of -- sorry, 4 with Intel, Samsung, Rapidus and TSMC. So that's also driving more demand for it.

Yes. Okay. Pretty clear. And maybe just with the chiplet coming into focus, and I've heard you talking about COT and hybrid COT sort of designs coming through, how does that all make a pull on the IP business for you as well?

Yes. I think that trend is good for both EDA and IP business. Because as customers do more and more of their own chips, they use more EDA tools. And also, because these things are so big and they're moving so fast, right, every year, every other year, the customer, of course, wants to focus on their key part. So if you can buy a standard-based IP which is good from Cadence, they would rather buy it and focus on the CPU part or the AI part or the auto chip part. So I think as long as we can deliver good perform PPA, the customers will rather buy that. I mean not all of them, but enough of them want to focus on -- some customers will do IP themselves because they think that's a differentiator. But a lot of them will buy it because they want to focus on some other part of the...

Got you. Makes sense. Maybe just turn to the core EDA business. I mean, I think you've guided up something close to 12-plus percent for this year. Last year, you grew about 13%. So clearly in that sort of low teens, moving back into that sort of category. What's the durability on the growth here? And what should we be thinking about as indicators for growth in the next couple of years?

Yes. I mean, if you look at it, we always look at growth plus margin together. We have, I think, world-class margins. Because that's what our investors want, right? We want to grow at a certain rate, but we want the profitability to be better than that. And then we buy back some stock, so we want EPS to be even better than that.

So last year, we grew like 14% and EPS grew around 20%, right? So this formula, we have done for several years. And -- but if you look at a Rule of 40 metric, I think we are in the high 50s, right, last few years. So I feel good about that. And I think we will definitely -- my goal is to crack 60.

60?

Yes, yes.

Okay. Make sure somebody noted that. There we go.

So that's a combination of growth, but also, of course, we need to make sure the margin is good. And if you look at our incremental margin -- I mean, our margin last year was 45%, but incremental margin was 59%. Like if we add $100 million more in revenue, we added $59 million in profit. And that's also making our internal operation more and more efficient with AI and things like that. So yes, we always look at both. But yes, my goal is to cross 60. And that should be good for our investors.

Yes. You would have thought. You did mention Hexagon, and congratulations on closing that deal. Maybe just help us understand how this all fits into systems design and analysis? And when do you think that can make an impact, particularly on the margin side for Cadence?

Yes. I think any M&A, normally, whatever company -- I mean, Hexagon or the simulation business of Hexagon is a great group, great company. They're one of the original simulation companies. It was just not ideal in Hexagon because Hexagon is more hardware company than software, and they realize it would be better with Cadence.

But anything we buy is never as profitable as Cadence. So -- but it takes us about a year or so to get it to that -- to better profitability. So I think definitely, this year, there is some hit. I mean, most of it is not on operating part. Most of it is on financing side because there is some dilution or there is some debt. But we will take care of that. And next year, it should be accretive.

Okay. Makes sense. Maybe just turning to China. We did see pretty decent growth last year. And this was despite, I think, some of the concerns you had expressed that this could be a strange year, '25. It turned out to be 18% growth. And this year has gotten off to a good start. How do you think that will continue to grow through this year? And what are the dynamics that you're looking at in China, certainly by EDA, but also in the IP space as well?

Yes. China is doing -- I mean, did well. It was very turbulent in '25, and we wanted to be prudent in our guide in the beginning of '25. I mean, I don't know that all those things would happen, but there was just a lot of uncertainty so we want to be more careful in beginning of '25.

This year, I think -- I mean, it's difficult to predict, but the environment seems more stable than beginning of '25. So this year, we think China will grow. And we'll see how much it grows because it's difficult to predict by region. The growth rate by region is like double derivative. So we'll see how much it grows, but the environment is good. There's a lot of design activity. Physical AI is big in China, of course. And even a lot of the other parts of the market. So I feel good about China right now.

Got you. And no competition from the local guys in that market either?

Well, there's always some competition, I think. But again, we want to make sure our tools are best-in-class. And EDA, we have a very good position in China. Hardware, we have a very good position, Palladium. IP, we do less in China historically. Because IP, we are focused more on really advanced node and AI. But in EDA and hardware, yes, it's good, and it should grow.

Got you. Before I move to further questions, I'll just maybe give the floor an opportunity to ask Anirudh directly anything.

I'm curious about the [indiscernible] and the SRAM-related chips. Do those all need your EDA tools to design, Anirudh?

Absolutely, yes. Any kind of chip, you need. I mean, there will be a lot of innovation on the hardware side and software side. So yes, no, all of these will -- you can't design them by hand. They have to use our tools. And they will need Palladium, they will need our EDA software. They will need IP.

I mean, one thing I want to say, it's very difficult to predict, but I mentioned this earlier also that -- like some people -- I talk to some of our customers, they say, "Oh, the inference demand will go up by 1,000x in the next 5 years." And that's amazing, and maybe a little more than that, right? So -- but then we have to normalize that with the improvements in hardware and software. So this is from current levels. So I think there was that customer or that one already assumed that the hardware will improve by 10x. They assume software will improve by another 10x. So the actual improvement is 1,000 divided by 100, okay? So which is 10x. By the way, 10x over 5 years, 60%. Even if that gets modulated by power and other things, maybe it's 30%.

So what I'm trying to say is that you know this already that it will not be static. The hardware and software will improve dramatically, whether it's this new hardware architectures or advanced nodes. Software will also improve, right? I mean, software has already improved a lot. And all these new CS algorithms will be applied to AI, right, whether it's partitioning, abstraction, latency, also the lower precision. All those things that happened in CS over 30 years will apply to AI. So I think it's going to be very exciting.

Now of course, it's possible the software improves even more, right, than 10x. But then a lot of times, the software improves more than 10x, the demand can go up even more, right? So it's like a very exciting double exponential, but I think it will be great to see all this innovation.

Yes. Makes sense. One area we didn't touch on was hardware. I know it's an area that you've done really well in, particularly with the launch of Z3 over a year or 2 ago. Any updates you've got there? Because it does look as though there's quite decent growth. It's represented well in backlog. And by the way, congratulations for your record backlog as well. So maybe walk us through that. How does hardware grow this year?

Yes. So hardware, when we say hardware, I know you know that it's like a full stack. So we make our own chips. So we also make our own chips to accelerate logic verification. So what happens is at the verification level, in chip design, there are 2 kinds of software. So one is like a more Boolean, if you remember, like 0, 1, Boolean logic, like how GPU or CPU will work. There's a lot of Boolean logic.

And then the other part is numerical. Like simulation or timing, power, noise, it's more numerical. So for numerical, we can accelerate it with CPU and GPU, okay? For Boolean, we build our own custom processor. It's a Boolean supercomputer. It's as complicated as any processor in the world. And when we do that, it runs like 1,000x faster than standard silicon. So we are our own kind of designer, using our own products.

So when we put that together in hardware and software, those things -- this is called Palladium -- become indispensable to design of modern chips. So all the big chips right now are designed by our product. Because you want to verify -- see this is what happened in the old days, you would design a chip and then do software development, and it would come out of -- to production, right? Any CPU or GPU. But the issue is that you don't want it to be wrong because if it is wrong, you have to iterate, okay? That's one problem. Second problem is the customers want to overlap hardware and software development. You don't want to wait until hardware silicon is ready and then start writing software because that takes too long.

So the demand for Palladium is driven by these two things. So what happens then is we overlap hardware and software. So the customers are writing software when no silicon exists. So that's why they use Palladium to emulate the silicon. So there are two advantages. One, you can start writing software. You can boot like Android or Windows or iOS, whatever you want. And the second is you can make sure that the silicon is correct. So it became like -- and to do that, you need to run 1,000x faster because if you run on a regular CPU, it's not going to be fast enough. So this is the reason that Palladium became like indispensable tool for chip design.

So now as the chips get bigger, you need more and more Palladium capacity. And then as there are more and more software, as the system companies start doing silicon, I mean, there are system companies because they have software and hardware, they need to run more Palladiums. So I mean, we have like 6 years in a row of record growth in Palladium. And so I think this year will be another record, and we'll see how it goes. But whenever we start the year, we are more prudent in the assumption. But I'm pretty bullish on Palladium this year as well.

Got you. Maybe with a couple of minutes to go, I have to ask about how you're going to monetize the Agentic EDA. So when we go back -- if we go back to ChipStack. And John was pretty clear on the callbacks, that this could be -- this is on a value-based basis. And this will be perhaps based on tokens. So maybe help us understand, how does that work? And could this be margin accretive in a couple of years' time for the group?

Yes, I mean, we always want to be margin accretive. You know us, right, over all these years. Every year, we're trying to be. But this is another big thing, right, that can help us. So again, yes, I think it will be token-based. And there are all these -- I mean, we want to have a base subscription plus tokens on top. That's how we want -- because these new tools will be new product categories. And then as they consume work, they can use tokens. And this kind of model in AI that is well established now. So we would hope to have a base subscription plus tokens. And that also gives good visibility to our customers, and it's good for us.

So they can see the meter, basically. Yes. Makes sense. It looks like we've run down the clock. Anirudh, thank you very much.

Thank you. Thank you very much.

Ladies and gentlemen, good afternoon. My name is Abby, and I'll be your conference operator today. At this time, I would like to welcome everyone to the Cadence Fourth Quarter and Fiscal Year 2025 Earnings Conference Call. [Operator Instructions] Thank you.

And I will now turn the call over to Richard Gu, Vice President of Investor Relations for Cadence. Please go ahead.

Thank you, operator. I would like to welcome everyone to our fourth quarter of 2025 earnings conference call. I'm joined today by Anirudh Devgan, President and Chief Executive Officer; and John Wall, Senior Vice President and Chief Financial Officer. The webcast of this call and a copy of today's prepared remarks will be available on our website, cadence.com.

Today's discussion will contain forward-looking statements, including our outlook on future business and operating results. Due to risks and uncertainties, actual results may differ materially from those projected or implied in today's discussion. For information on factors that could cause actual results to differ, please refer to our SEC filings, including our most recent Forms 10-K and 10-Q, CFO commentary and today's earnings release. All forward-looking statements during this call are based on estimates and information available to us as of today, and we disclaim any obligation to update them.

In addition, all financial measures discussed on this call are non-GAAP, unless otherwise specified. The non-GAAP measures should not be considered in isolation from or as a substitute for GAAP results. Reconciliations of GAAP to non-GAAP measures are included in today's earnings release. [Operator Instructions].

Now I'll turn the call over to Anirudh.

Thank you, Richard. Good afternoon, everyone, and thank you for joining us today. I'm pleased to report that Cadence delivered excellent results for the fourth quarter, closing an outstanding 2025 with 14% revenue growth and 45% operating margin for the year. We finished 2025 with a record backlog of $7.8 billion, well ahead of plan, reflecting broad-based portfolio strength and increasing contributions from our AI solutions.

I would like to emphasize the essential nature of Cadence's engineering software. As I have stated previously, our platform is best viewed as a 3-layer cake framework, accelerated compute being the base layer, principal simulation and optimization as the critical middle layer and AI as the top layer to drive intelligent exploration and generation. This holistic approach ensures that our AI solutions are not just fast, but physically accurate and grounded in scientific truth. Building on this foundation, we are deploying Agentic AI workflows powered by intelligent agents that autonomously call our underlying tools.

AI flows act as a force multiplier, enabling our customers to significantly expand design exploration and accelerate time to market, while driving increased product usage and deeper engagement across our entire platform. We see growing momentum on both AI for design and design for AI fronts. On AI for design, our Cadence AI portfolio continues to gain traction with market-shaping customers. Last week, we launched ChipStack AI Super Agent, the world's first Agentic AI solution for automating chip design and verification. It is built upon our proven physically accurate product and provides up to 10x productivity improvement for various tasks, including design coding, generating test benches and debugging.

ChipStack has received compelling endorsements from Qualcomm, NVIDIA, Altera and Tenstorrent, among others. Our other AI products such as Cadence Cerebrus, Verisium and Allegro X AI are proliferating at scale. And our LLM-based design agents powered by JedAI data platform are delivering impressive results.

On design for AI, the infrastructure AI phase is in full swing with AI architectures growing in scale and complexity. Customers are increasingly standardizing on Cadence's full flows to address their performance, power and time-to-market challenges. We continue to closely collaborate with market leaders on their next-generation AI designs spanning training, inference and scaling. We deepened our long-standing partnership with Broadcom through a strategic collaboration to develop pioneering Agentic AI workflows to help design Broadcom's next-generation products.

We also expanded our footprint at multiple marquee hyperscalers across our EDA, hardware, IP and system software solutions. And we are particularly excited by the emerging physical AI opportunity, and our broad-based portfolio uniquely positions us to enable autonomous driving and robotic companies to address multimodal silicon and system challenges. In addition, we are increasingly applying AI internally to improve efficiency across engineering, go-to-market and operations.

In 2025, we also furthered our partnerships with leading foundries. We expanded our collaboration with TSMC to power next-gen AI flows on TSMC's N2 and A16 technologies. We strengthened our engagement with Intel Foundry by officially joining the Intel Foundry Accelerator Design Services Alliance. Rapidus made a wide-ranging commitment to our core EDA software portfolio across digital, custom analog and verification solutions. And Samsung Foundry expanded its collaboration with Cadence, leveraging our AI-driven design solutions and IP solutions.

Now turning to product highlights for Q4 and 2025. Accelerating compute demand driven by the AI infrastructure build-out and demanding next-generation data center requirements continue to create significant opportunities for our core EDA portfolio. Our core EDA business delivered strong performance with revenue growing 13% in 2025. Our recurring software business reaccelerated to double-digit growth in Q4, a testament to the strength and durability of our model.

Our hardware business delivered another record year with over 30 new customers and substantially higher repeat demand from AI and hyperscalers. 7 out of the top 10 customers in 2025 were Dynamic Duo customers, underscoring the differentiated value provided by our hardware systems. With a strong backlog entering 2026, we expect this year to be yet another record year for hardware. Our digital portfolio delivered a strong year, driven by continued proliferation of our full flow solutions as we added 25 new digital full flow logos in 2025.

We expanded our footprint at a top hyperscaler, growing our AI-driven synthesis and implementation solutions, including our 3D-IC platforms. A marquee hyperscaler embraced the Cadence digital full flow for its first full customer-owned tooling AI chip tape-out. Broad proliferation of Cadence Cerebrus continues and adoption of our Cadence Cerebrus AI Studio is accelerating. Recently, Samsung U.S. used it to tape out a SF2 design, achieving 4x productivity improvement. In custom and analog, our Spectre circuit simulator saw significant growth at leading AI and memory companies.

Our flagship Virtuoso Studio, the industry standard for custom and mixed-signal design saw continued traction in AI-driven design migration across its vast installed base. A top multinational electronics and EV customer reported a 30% layout efficiency gain using our AI-driven design migration. Our IP business saw strong momentum with revenue growing nearly 25% in 2025, reflecting both the strength of our expanding IP portfolio and the critical role our STAR IP solutions play in the AI, HPC and automotive verticals.

We achieved both significant expansions and meaningful competitive wins at marquee customers, demonstrating the superior performance and capabilities of our IP solutions across HBM, UCIe, PCIe, DDR and SerDes titles. We are seeing particularly strong adoption of our industry-leading memory IP solutions, including our groundbreaking LPDDR6 memory IP, which is enabling customers to achieve the memory performance and efficiency required for next-generation AI workloads.

In Q4, we launched our Tensilica HiFi IQ DSP, offering up to 8x higher AI performance and more than 25% energy savings for automotive infotainment, smartphone and home entertainment markets. Our System Design and Analysis business delivered 13% revenue growth in 2025. Earlier in the year, we introduced the new Millennium M2000 AI supercomputer featuring NVIDIA Blackwell, which is ramping nicely and with growing customer interest across multiple end markets. Our 3D-IC platform has become a key enabler for the industry's transition to multichip architectures, which are increasingly critical for next-generation AI infrastructure, HPC and advanced mobile applications.

Adoption of our AI-driven Allegro X platform is accelerating. Earlier in Q3, Infineon standardized on Allegro X and in Q4, STMicroelectronics decided to adopt our Allegro X solution to design printed circuit boards. Our reality data center digital twin solution continued its strong momentum and was deployed at several leading hyperscalers and marquee AI companies. BETA CAE continues to unlock tremendous opportunities, particularly in the automotive segment. With our previously announced acquisition of Hexagon's D&E business, we'll be poised to accelerate our strategy around physical AI, including in autonomous vehicles and robotics.

In closing, I'm pleased with our strong performance in 2025, and I'm excited about the strong momentum across our business. As the AI era continues to accelerate, our AI-driven EDA, SDA and IP portfolio, powered by new AI agents and accelerated computing positions Cadence extremely well to capture these massive opportunities.

Now I will turn it over to John to provide more details on the Q4 results and our 2026 outlook.

Thanks, Anirudh, and good afternoon, everyone. I'm pleased to report that Cadence delivered an excellent finish to 2025 with broad-based momentum across all our businesses. Robust design activity and strong customer demand drove 14% revenue growth and 20% EPS growth for the year. Productivity improvement across the company helped us achieve an operating margin of 44.6% for the year. Fourth quarter bookings were exceptionally strong, and we began 2026 with a record backlog of $7.8 billion.

Here are some of the financial highlights from the fourth quarter and the year, starting with the P&L. Total revenue was $1.440 billion for the quarter and $5.297 billion for the year. GAAP operating margin was 32.2% for the quarter and 28.2% for the year. Non-GAAP operating margin was 45.8% for the quarter and 44.6% for the year. GAAP EPS was $1.42 for the quarter and $4.06 for the year. Non-GAAP EPS was $1.99 for the quarter and $7.14 for the year.

Next, turning to the balance sheet and cash flow. Our cash balance was $3.01 billion at year-end, while the principal value of debt outstanding was $2.5 billion. Operating cash flow was $553 million in the fourth quarter and $1.729 billion for the full year. DSOs were 64 days, and we used $925 million to repurchase Cadence shares during the year.

Before I provide our outlook for 2026, I'd like to share that it contains our usual assumption that export control regulations that exist today remain substantially similar for the remainder of the year. And our current 2026 outlook does not include our pending acquisition of Hexagon's design and engineering business.

For our outlook for 2026, we expect revenue in the range of $5.9 billion to $6 billion, GAAP operating margin in the range of 31.75% to 32.75%, non-GAAP operating margin in the range of 44.75% to 45.75%; GAAP EPS in the range of $4.95 to $5.05, non-GAAP EPS in the range of $8.05 to $8.15, operating cash flow of approximately $2 billion, and we expect to use approximately 50% of our free cash flow to repurchase Cadence shares in 2026.

For Q1, we expect revenue in the range of $1.420 billion to $1.460 billion. GAAP operating margin in the range of 30% to 31% non-GAAP operating margin in the range of 44% to 45%; GAAP EPS in the range of $1.16 to $1.22 and non-GAAP EPS in the range of $1.89 to $1.95. And as usual, we published a CFO commentary document on our Investor Relations website, which includes our outlook for additional items as well as further analysis and GAAP to non-GAAP reconciliations.

In conclusion, I am pleased that we delivered strong top line and earnings growth for 2025, and we finished the year with a record backlog and ongoing business momentum, setting ourselves up for a great 2026. As always, I'd like to thank our customers, partners and our employees for their continued support.

And with that, operator, we will now take questions.

[Operator Instructions] And our first question comes from the line of Vivek Arya with Bank of America Securities.

Anirudh, I'm curious, have you seen any disruption or change of thinking whatsoever at your customers in terms of them using AI to reduce or eliminate demand for EDA or IP or any other computer-aided engineering tools. Is there a scenario at all that you have discussed, right, or your customers might contemplate where they can use more of their internal tools or AI to displace what you're doing right now?

Yes. Vivek, thank you for the question. I know this is a topical question on top of mind for investors. But like I said before, I mean, for us, we always look things as a 3-layer cake. And there's different kinds of software. There's a lot of discussion in terms of will AI replace some form of software. But you know well anyway, there are different kind of software. Our software is engineering software, you're doing very, very complex physics-based mathematical operations. So any AI tools that we are developing or our customers are using basically in the end, call our software to get the job done properly.

So what we are saying instead is that -- and you can see that in our results, we can see this in our discussion with customers is there is -- as we move to these Agentic flows, it uses more of our software to get the job done than the other way around. So as we -- even like our own super agent, which is ChipStack, it is doing a part of the flow, first of all, that was not automated. Even in regular AI, there is a lot of automation in coding. That's one of the big applications.

But if you move that over to chip design, if you look at our flow, there is an equivalent of coding, which is RTL code, which describes the chip or the system. But that part has been mostly manual. And then after that, our tools kick in to optimize the RTL to simulate, verify the RTL. So what we are doing with our AI flows, the top layer is we are adding extra tools that will automate the writing of RTL, but then still, it calls a lot of middle layer tools, a lot of the base tools to implement and verify that.

And I've said before, like what we are seeing at our customers, they want to use more AI. And I think they will invest more in R&D. I think they will also hire more engineers. But as a percentage of spend, the more spend will go to automation and compute because the other thing which is unique to our end market is that the workload is exponential. If the chip goes from $100 million now to $1 trillion in a few years, they need to do a lot more work and then some of the work will be done by AI agents calling our base tool. So overall, to answer your question, we have seen absolutely no discussion with customers of reducing the usage. On the contrary, all these AI tools are increasing the usage of our tools. And of course, then the AI build-out also, as customers design more and more chips, that is also increasing the usage of our tools.

And our next question comes from the line of Joe Vruwink with Baird.

I maybe wanted to ask about how you're approaching the outlook for 2026. It looks like recurring revenue is set to accelerate, and that's normally well supported by backlog. Maybe can you talk about the key contributors to the recurring improvement? And then just on the 20% or so of revs that come from upfront sources, you obviously had an incredible 2025 with your hardware platforms and it sounds like you're expecting growth there again. I think we're in year 2 of that platform now. Can you kind of see a repeat of what you observed back in 2023. That was a very strong year 2 for the second-gen product. How maybe are you thinking about that product and just where it is in its life cycle?

Yes. Thanks for the question, Joe. This is John. As usual, at this time of the year, our guidance will reflect what we believe to be a prudent and well-calibrated view of the year. We finished the year with very strong momentum on backlog, and we saw that strength right across the board across all lines of business. And as Anirudh says, our view of the AI era is that it increases workload faster than headcount grows and Cadence monetizes workload through broad portfolio proliferation across EDA, IP, hardware and SDA. And we're seeing that flow through into all lines of business for us.

Now typically, at this time of the year, our hardware is a pipeline business. We're expecting a very strong first half for hardware. But because we only typically see 2 quarters in the pipeline, we're quite prudent in the second half of the year. in this current guide, but that's no different to what we normally do. Same -- we typically try to derisk the guide for things like hardware and China at this time of the year. And if you look at how China has performed in the last 2 years, I think it was 12% of our revenue in 2024, 13% in 2025, and we expect it to be in that kind of range, 12% to 13% of our revenue as well for this year.

But yes, we're seeing absolutely huge strength across the board, delighted with the strength of the guide. And just a key transparency metric you'll see in the CFO commentary that around 67% of 2026 revenue is coming from beginning backlog. And that gives us strong visibility into the multiyear recurring base. So we're very, very happy to see that recurring base get back to kind of double digits, kind of low teen growth.

And our next question comes from the line of Joe Quatrochi with Wells Fargo.

Just kind of curious, maybe following up on that. On the verification and emulation hardware cycle, any sort of help on just kind of where you think you are at in that cycle? And then is there anything we should think about just in terms of memory availability from that perspective or just anything about margins given pretty significant price increases that we've seen across the DRAM spectrum?

Yes, good question. So hardware, like you know, is in a multi -- every year is a record for hardware, and I expect that trend to continue. And the reason being, of course, these hardware systems become indispensable to the design of complex chips and systems. Actually, no complex AI chip or any other mobile or automotive chip, any complex chips are not designed without hardware systems, and we have the best hardware system on the market because we design -- just to remind you, we design our own chips made by TSMC, and we sell full racks. These things have trillions of transistors to emulate other chips.

So even though it is reported upfront, as you know, because the customers will buy and use these systems for multiple years, the big customers are buying them almost every year, okay? And I don't see that trend changing. And like even I indicated like when we launched Z3, even Z2 was a very good system. So the fact that the system is second year now, I think it's still -- it has capacity to design systems of 1 trillion transistors, okay, which will last for several years to go. And in a few years anyway, we'll launch our next system. So we're always ahead of what the market will need.

But in terms of demand, we don't see any difference versus -- if you ask me this year versus last year, the demand is only stronger, and you can see that in the backlog. And then how much this will grow, we will see. Like John said, beginning of the year, we are a little careful with hardware. But we'll update you middle of the year depending on how things are going. But hardware systems are performing well. We are taking share. And actually, what I feel is we are taking share in all our major product segments. So we are taking share in hardware. We are taking share in IP, which is really good to see now. This will be almost third year of strong IP growth.

You know us, right? We don't -- 1 year doesn't make a trend for us. So -- but after 3 years, I can see that I feel good about our IP business. Hardware has been strong for a while. EDA, our core business is doing phenomenal, okay? 3D-IC, we are taking share. Agentic AI, we are first to market. We already have a lot of customers using our Agentic AI flow. So not only I feel good about the hardware business and where it is, actually, I feel really good of our overall portfolio and how we are performing.

And our next question comes from the line of Jim Schneider with Goldman Sachs.

I was wondering if you could talk about a little bit more about your -- your AI workflows. And if it's possible to quantify any of the benefits that your customers are getting from those workflows today, whether that be time to market, enhanced productivity per seat or so on? And maybe separately kind of address how you're able to monetize that and how broad that is across your portfolio today?

Yes, Jim. I mean, first of all, the results are quite remarkable with AI. And like a few years ago, there was some skepticism of how much AI can benefit. But now, I mean, this is true in other areas, too. But definitely, in chip design, the results are fantastic and real. And I think there is a difference, I believe, in chip design versus other industries because one of the issues with AI flows is that you really don't know whether the AI result is correct or not. And this has been one issue even in wipe coding or software, like, okay, generate some code, but you spend a lot of time verifying that it is correct or not.

And in some other industries, there is no like formal languages to design things. But in chip design, first of all, we have formal languages to design things, which is RTL. Plus over the last 20, 30 years, we have built all these products whose job is to make sure that the RTL is correct, okay? So all our middle layer tools, verification, simulation, optimization. So therefore, AI can be a force multiplier and accelerant to chip design versus other areas, okay?

And so the way -- and the results, just to highlight, like we talked about Samsung getting 4x productivity. This is code from the customer or Altera talking about 7 to 10x productivity improvement. Now they're on parts of the flow for like RTL writing, which has been kind of manual, there can be massive improvements in productivity. And in the back end, for example, physical design, there could be 7%, 10% PPA improvement, 12% in that range. So just that you know that when you go from one node to another node, like 5 to 3 or 3-nanometer, 2-nanometer, the gain could be like 10%, 20%. So you're getting half the gain or almost the same gain as a node migration through better optimization with AI, okay?

So I think the results are real. We have demand from almost all customers now to engage rapidly because they want to deploy AI in their R&D function. And you have to remember the way our customers deploy R&D in the -- apply AI in their R&D function is through Cadence and Cadence tools, right? So they are all very anxious to try all these things. We have all these engagements with all the top customers. And our monetization, and I've always said in the past that it takes some time for monetization to happen. It takes 2 contract cycles, and I think we are well into that now.

So I think we are seeing the monetization now, which is reflected in our results is reflected in our record backlog. And Agentic AI can give further monetization. So the way we go to market with Agentic AI will be different because this is a new tool category of something that EDA never automated. Writing of RTL or test benches was manual, right? So we will price it as like a virtual engineer or agent. So that would be extra business. And our customers are willing to spend on that because it is productivity improvement for them.

And then on top of that, just like before, it will call the base tools and they become a lot more licenses or usage will happen our base tools. And the reason for that is like in the non-AI flow, this is a misnomer that we are like seat count limited. We are exploration limited. Even if a user, like a manual user is running our tool, they will run like 3 or 4 or 5 experiments in parallel to see what is the best PPA. But with the Agentic AI flow, it could run 10 or 100 experiments in parallel. So our plan for monetization, which is working well, we'll add the Agentic AI part. We will charge for the Agentic flows from a virtual engineer, things like RTL writing and then, of course, for the licenses in the base layer and see how that goes. But from a customer standpoint, I mean, there's a lot of demand to try all these new tools.

And our next question comes from the line of Gary Mobley with Loop Capital.

Let me extend my congratulations on the strong finish to the year. John, I believe there's been an effort to move your SD&A customers into 1-year license terms. And if I'm not mistaken, that's been an impediment to growth. So the question is, is that the reason why SD&A revenue grew only 13% in 2025? And what's the consideration for 2026? And then what's the consideration for Hexagon when you roll that business? And I believe they were at a $240 million revenue run rate. Does that see a more limited -- is that number limited because of this 1-year license term transition?

Yes. Thanks for the question, Gary. Yes. And you're right in terms of SD&A, we lapped some tough comps in SD&A in Q4 2025, partly due to the multiyear business. So we did some multiyear business in Q4 2024 through our BETA subsidiary, and we have deliberately been moving to more annual subscription arrangements for BETA in 2025, and that impacts the year-over-year numbers. In saying all that, we're very pleased with SD&A's strategic trajectory and its role in the chip-to-systems thesis.

From a mix standpoint, SD&A was like 16% of revenue in '25, consistent with '24 when you look at the year. and we expect it to grow. We expect all product groups to grow, but we're not guiding by segment. In relation to Hexagon, I think the annualized -- I think we've said this before at some fireside chats that the annualized revenue for Hexagon is about $200 million on a year basis. Now what that means, of course, that it's kind of like BETA where BETA did a lot of January 1 deals that -- like if that deal closed by the end of Q1, you're probably looking at $150 million revenue for the year. But we're not guiding. We don't have final numbers for anything like that now. But -- so we haven't got anything to do with Hexagon in this guide.

And our next question comes from the line of Charles Shi with Needham.

Anirudh, I thought the best highlight of the quarter was the announcement around the marquee hyperscaler customer adopting Cadence digital full flow. I think you characterized it as for the first COT chip that they're going to tape out. So it sounds like we should expect that particular hyperscaler having a COT chip coming out in 2 or 3 years down the road. And just kind of want to ask a question like how many hyperscaler customers right now are doing COT and even for that particular customer having the first chip on COT, wonder what's your -- what do you think the ramp is going to be?

Like how will they proliferate COT for the other chips they are developing? Because every hyperscaler these days have more than one chip. That's my understanding. And I just want to get some sense from you where you are in terms of that whole COT proliferation. And I believe this is one of the great stories about the Cadence about EDA in general, but I want to get your sense.

Yes. Thanks for the question, Charles. I mean, without getting into like specifics of a particular customer, but I have said for some time now because we work with our customers confidentially. We share our road map with them. They share road map with us. And we are in a unique position to work with all the leading companies across the globe, right? And so I have said for a while that this trend of -- first of all, the trend that the customers, especially these big hyperscalers will do their own chips is even more firm now than 1 or 2 years ago.

And it's evident now with some of the big hyperscalers, the success they're having with their own chips, right, especially in the last 6 months, that has become evident because it was not clear like 1 or 2 years ago, people thought people will not design their own chips. It doesn't mean that the merchant semi will not do well. A merchant semi will do fabulous, but the big customers will design their own chips, okay? And then this is also true that over time, the big customers will do more and more things in-house, starting with ASIC to hybrid COT to COT because these chips are -- I mean, this is more -- there's another step these days versus the old days, which is hybrid COT because these chips have multiple chiplets in them. So the customers can do some of the chiplets themselves, some can be outsourced and then they can do all of them themselves.

So I think this trend is going to happen. And the reason we talk about it, it is happening and different customers will do it at different pace. But eventually, I think there will be multiple customers with their own chips. There will be multiple, of course, very significant semi-standard general purpose chips. And almost all of them will, over time, do more and more COT. And like you said, they do multiple chips now, at least 3 major platforms for each hyperscaler.

So all this is good for us, good for more EDA consumption at the system companies, more IP being used internally, of course, more hardware, more system tools because they are nature -- system companies in nature. So we just want to make sure we are well positioned for that, but the trend is only accelerating of these big companies doing more themselves. And then as you know, this will also then apply to other verticals like automotive and robotics and things like that.

And our next question comes from the line of Siti Panigrahi with Mizuho.

You talked about robust design activity. Can you give us some color in any kind of improvement on your traditional semi segment versus AI or automobile. If you could give some color, that would be helpful. And Anirudh, on the physical AI side, that was a big focus at CES recently. Have you started seeing any traction in that space? When do you think that will be a significant contributor?

Yes. Thanks for the question, Siti. On both, I mean, the design activity is accelerating, like I was saying, and that's true for system companies and semi companies. And actually, I mean, a lot of the projections are that we might hit as the industry, semi might hit $1 trillion this year, which is like it used to be 2030, and we are like 4 years ahead of that. So this is very good news for the industry. And of course, we have deep partnerships with all the major semi players and definitely the AI leaders like with NVIDIA and with Broadcom. Actually, in this prepared remarks also, we highlighted our new collaboration with Broadcom, which are, of course, doing phenomenally well and so is NVIDIA. And then, of course, all the memory companies are doing phenomenally well.

So overall, I think the semi companies, along with system companies are doing great. And I do see, especially in AI and memory, but we do see the general market, I'm sure you follow that, the mixed-signal companies, the regular, let's call it, the regular semi companies are also, I think, have a better outlook for '26 than '25. So it's good to see a broad-based strength in the semi business, which is about 55% of our business. And that just creates a better environment for us to deploy our new solutions. And they all want to deploy AI like we discussed earlier. And that's true for both semi and system companies. So overall, I feel that the environment is much more healthier starting '26 than it was like a year ago.

And our next question comes from the line of Lee Simpson with Morgan Stanley.

I just wanted to go back to ChipStack, if I could. I mean it seems relatively clear that you see the super agent as something that can transform from Verilog to RTL or the coding thereof at least. And then it would pull in basic layer tools for debug and optimization. So you get a more deterministic outcome for customers. But you teased us a little bit with the idea about where the further monetization would come. It didn't sound like it would be on a subscription basis. It would be on a sort of value to customer basis. So I wonder if you could maybe just expand a little on that and how that would be monetized? And maybe in particular, whether or not this would be margin accretive. You're at 45% now already. So could this help kick that on?

That's a great question, Lee. if I might jump in here on the monetization side that we don't see AI forcing a wholesale change from subscriptions to consumption. Our customers still want predictable access to trusted sign-off engines. and certified flows. So multiyear subscription remains at the core of our business. What AI does is it changes how much customers run the tools and where value is created. There's more automation. There's more iterations, there's more compute. So we'll attach more usage-based pricing for incremental capacity and AI-driven optimization. We have card models and token models that handle all those things.

And then in a few areas on the services side, we can offer outcome-oriented packages that's structured around measurable improvements like cycle time, closure productivity with clear scope and governance. And that's kind of how we've been going to market in recent times. And it's worked out well for us. And you can see how it's turning around already our recurring revenue. Now we've been prudent in our outlook, and we're not expecting an uptick in that, but it definitely is -- there's plenty of opportunity for Cadence in AI.

But as Anirudh said at the beginning in his opening comments there, that there's 2 real things that differentiate Cadence. First, we're engineering software anchored in physics and mathematically rigorous optimization. And that's not a nice to have. It's a core truth that our customers require as complexity rises. And then secondly, AI is not replacing our products. It's amplifying demand and accelerating adoption. And you see that in our results for 2025, and I think you see it in our guide for 2026. Anything to add?

And our next question comes from the line of Jason Celino with KeyBanc Capital Markets.

Looks like IP had a phenomenal year. I know you have a slate of new exciting titles coming out, but I just wanted to ask how that translates to pipeline? Like does it take time to sell these new IP titles? And then with the guide overall, it looks mostly first half weighted. Does your visibility into the IP today look more first half or second half?

IP is doing great. I mean, like I said, we want to see multiple years of performance before we call it out. And starting last year, I started to call it out because we saw like multiple years and good outlook into '26, which I think should come true. So our starting backlog and everything in IP is strong.

And then we are also talking to -- I mean, not just our traditional business with TSMC, which is doing phenomenal, but we have opportunity to engage with the newer foundries. So overall, I think IP will be good this year, and we'll see how it progresses. We'll keep you updated, but it should be a strong year for IP in '26.

And our next question comes from the line of Jay Vleeschhouwer with Griffin Securities.

Anirudh, if we think about what's currently occurring with the AI phenomenon in large EDA historical terms, the last time I would argue that there was a major let's call it, generational technical and procedural change in the industry was in the early 2000s. And I'd like to ask how this time might be different from that phenomenon in the sense that the last time, it was fairly narrowly based in terms of the number of products that grew or were newly adopted. We saw the very interesting phenomenon where average contract durations actually shrank.

I think, as customers were looking to perhaps mitigate technical risk and wanted to retain some vendor flexibility or optionality, hence, the shorter durations at that time. Would you say that this time around, the adoption phenomenon might last longer than just a few years of the earlier generation I mentioned that there wouldn't be necessarily an adverse effect on contract durations, perhaps maybe even a lengthening with longer commitments from customers. And maybe talk about how in those big respects, this phenomenon might be broader and more long lasting than what occurred, again, many years ago, but it has some similarities.

Yes. That's a great point, Jay. And I mean, we have to see how it unfolds because each time is similar but different. But we are not seeing any change in the duration, so which is good. We don't want to -- but there is always more opportunity to see more and more add-ons like we have mentioned in the past, -- now it will affect all parts of the flow like in the 3-layer cake, the top 2 layers will fuse together, AI and our core engines. And I think there is opportunity to add, like I said, add new product categories, especially in the front end, this kind of super agent to write RTL, which -- and write -- not just write RTL, which this is different from regular kind of wipe coding.

So what is exciting about ChipStack is it's not just writing RTL, but also writing test benches, writing verification flows because you know that, Jay, anyway, that chip verification is as important as chip design. If you can't verify, then the thing -- because all our customers want things to be first time right. So I think the opportunities of AI and verifications are huge because that's an NP-complete exponential problem. So I think what is also exciting to me on the Agentic AI new tools is the ability to verify much more accurately. And then we go from there. I mean, I think I feel good about the strength of the -- at this point, I feel good about all the 3 layers of the cake. We have been innovating.

We have been first to market in porting our software to new hardware platforms, whether they're parallel CPUs or GPUs or custom chips. Our base tools are performing remarkably well. We are taking share in almost all segments. And then we are first to market with Agentic AI. So I feel good about the portfolio. I feel good about the engagement. Now how exactly it will unfold, I think it should be more long-lasting, but we'll -- it's very difficult to predict. So we'll keep you posted, but so far, so good.

Yes. This is John. Just -- I mean, we've been around a long time in terms of chasing Moore's Law for the longest time. And we've built sales models that generally adapt to aligning price with value while preserving the durability of our recurring revenue model. I think what you can count on us to do is that we won't undermine customer predictability that subscriptions will remain the anchor in terms of our primary engagement with our customers. And then we won't take unbounded outcome risk either. Outcomes will be scoped and measurable. And we'll value -- we'll price on value metrics. Customers can control things like jobs and runs and compute and throughput and things like that. But -- so it will be very, very deliberate and thoughtful in terms of how we grow as we always are.

And our next question comes from the line of Gianmarco Conti with Deutsche Bank.

Congrats on a great quarter. I have a long question. Sorry to go back on ChipStack, but could we start by giving some detail about how can we bridge the gap between ChipStack, which we know is about RTL automation and where it evolves versus Cerebrus, which is about implementation with regards to NAND. I guess my question is about whether there could be some cannibalization in the future. And staying on the AI theme, -- could we have some information about given where model development is happening in AI, whether you're seeing more competition, particularly from the startups. I know that present there and whether that's kind of coming up the pitch clients. And finally, just to pile up, are there any harder constraints when you run more agents given that you're going to require more compute, especially at higher design scales?

Yes. Sorry, there's some noise on the line. So I think I got the gist of the question, but I may not have gotten all the points. So sorry, I apologize in advance. I think your question is also about the front-end agent versus Cerebrus and also start-ups, if I -- so first of all, I think Cerebrus super critical. I mean -- so I think there will be several kind of AI Agentic flows that will be needed. Now we highlighted ChipStack because it's kind of new, and it's a new category of RTL design and verification. But there are at least several agents that we are actively developing. Cerebrus, we also extended the Cerebrus to full flow.

So there has to be a front-end design agent like Cerebrus. There's a back-end agent for physical implementation because that takes a lot of time right now, and there's a lot of demand for making the implementation more efficient. And there's similar principles apply in Cerebrus AI Studio. We do more exploration and the customer gets better results as a result of that. But there will be a lot of activity we will highlight in the future on the back end, on the physical design.

So there's digital design and verification is one area. physical design in another area. Analog, of course, is ripe for. Finally, we have new technology to see if we can automate more and more of analog and migration flows. And then on packaging and system design. So we highlight ChipStack because we're super excited about it, but that doesn't mean that all the other -- there are 4 or 5 big agentic flows that we are developing.

On the start-ups, we always watch all the start-ups. We have a history of also acquiring them if they are good, but more in the earlier stages like we did with ChipStack. I think that was the best AI start-up out there. And we are very confident in our own R&D. We have like 10,000 people, the best R&D team in computational software. Half of them have advanced degrees. We have 3,000 people with customer support engineers. We're regularly meeting with customers -- with big customers in a given week, we'll have multiple R&D meetings with their R&D.

So we keep track of what the customer wants. We have massive investment in R&D. And typically, I think the start-ups are successful in areas we don't focus in or if you want to enter in new areas. But in terms of AI, we are completely focused. And we always use start-up as an accelerant if need to, but we will have massive investment in this space in all the major domains that our customers want.

And our next question comes from the line of Ruben Roy with Stifel.

Anirudh, you answered bits and pieces of what I'm about to ask, but I was hoping to put together a question on SD&A and just to understand sort of the longer-term strategy. It seems like some companies, enterprises, industrials otherwise are maybe thinking about pulling some simulation workloads in-house or partnering with the AI infrastructure ecosystem. We've seen Synopsys and NVIDIA talk about targeting Omniverse digital twins for that type of thing. How should investors think about your strategy? Is it sort of a neutral strategy and you'll work with accelerated compute providers, et cetera, and their tools? Or are you trying to build sort of an ecosystem that's Cadence specific? I'm just trying to understand kind of longer-term strategy and thinking around SD&A.

Yes. Thank you for the question. So in SD&A, like there are 2 critical areas for us. So one is 3D-IC and all the innovation that's happening, both at the packet level analysis. And then the other is physical AI, physical simulation like for planes and cars and robots and drones. And that's one of the big reasons to acquire BETA and then Hexagon. But we are focused on building the core engines, okay? And the core engines will work with the accelerated compute, like we have -- we have done GPU joint work with Jensen in India for years. And we were the first to port all our soft solvers to kind of accelerate compute platform because the physical simulation word just is -- a lot of the simulation and physical like cars and planes and robots kind of CFD and structural simulation.

And I've said this before, is naturally without getting too technical, is naturally matrix multiply, okay? And GPUs and NVIDIA is exceptional in that because AI at its core is matrix multiply. So it's a good fit. And then we work with Omniverse and all. But that is not in -- Omniverse is a great platform, but when they actually run Omniverse, they will run our tools through that. So this is another way to go to market. And then also directly with customers.

So we are neutral to that, but Omniverse is a great platform to deploy our products and NVIDIA has highlighted that with several of our customers. But our goal is to build the basic -- we are an engineering software company. We build the basic solvers that can solve the most difficult problems, combine them with AI, combine them with compute and deploy it to all platforms. So I feel good about our position that way.

And our next question comes from the line of Andrew DeGasperi with BNP Paribas.

I just had a question. You mentioned several times in the prepared remarks about taking share across the board. And I was just curious, is this kind of a change relative to previous quarters? And is it focused in any particular area? And are you surprised by this relative to what you've seen in the past?

Yes. I think our competitive position has improved. So we are noticing that and calling that out. And definitely in hardware, given the uniqueness of our platforms in IP. And I mean a lot of it, you can see it in the results as well. Our growth is much higher than the market. So IP is doing well. Hardware is doing well, EDA, 3D-IC, and we are holding, of course, our traditionally good position in analog and gaining in digital and verification. So I feel very good about -- we are technology-centric, R&D-centric company first. And I think all those investments are paying off with customers adopting more of our flows.

And our next question comes from the line of Kelsey Chia with Citi.

Congrats on the great results. I'd like to dive a little on China. So John, you mentioned that you contemplated a more prudent guidance from China. China revenue grew 18% last year, outpacing corporate average and also well above your initial guidance heading into 2025. How should we think about the sustainability of this strength? And also, what are the assumptions you have embedded in that guidance?

Yes. So look, as we said earlier that the assumptions embedded in the guidance is that we saw 12% of revenue coming from China in 2024 and 13% in 2025, and we expect it will be in a similar range, 12% to 13% for 2026. But what we've seen in China is design activity remains very, very strong, and we're seeing strong bookings growth in the region. But visibility is -- visibility in the pipeline is near term in the first half of the year. So the second half of the year, there's probably more prudence in the second half of this year's guide for China than there would be in the first half because we have more visibility in the first half. Anything to add on design activity in China?

Design activity is good in China. And I think it has stabilized. I mean we had mentioned this last year also, second half had stabilized. And I think it continues to be strong. I mean, China is all the trends that are in the U.S. are also in China, a lot of AI chips, a lot of physical AI is even stronger with cars and autonomous driving, EVs. So it's good to see China doing well.

And our next question comes from the line of Joshua Tilton with Wolfe Research.

I will echo my congratulations on a strong quarter. I kind of have a high-level one. I know a lot of times we focus on like what the 3-year CAGR has been. And I think on this call, Andrew mentioned that semis companies now represent or still represent, I think, from my understanding, about 55% of the business. So my question is, how do we think about growth over the next 3 years as the mix of semis and systems levels out and what feels like the mix of upfront and recurring levels out at what I'm assuming is kind of more sustainable levels than the shifts you've seen over the last few years?

Yes. I think we are super excited about the system companies doing more silicon. And there have been some questions in the past. And like I had said before, I think this is irreversible and accelerating trend, okay? And of course, we gave several examples this time. And especially because of AI, the system companies will do a lot. And then with physical AI, they will do even more. Now that number, 55-45, first of all, moves very, very slowly because the semi companies are doing well, too. I mean we are growing at a record pace, but both of them are growing.

So semi companies, okay, what NVIDIA has done, of course, is phenomenal. What is happening with Broadcom is phenomenal. And then Qualcomm, MediaTek, there are so many semi companies are doing phenomenally well. So the ratio, I think more and more system companies will contribute more, but it doesn't move as fast as you would think, which is a good thing because the semi companies are also growing rapidly. And of course, semi companies will have an essential role in the build-out of AI, which is driving all this growth. So that's what I would like to say.

Yes. And Josh, the -- I think I mentioned before, we expect the recurring revenue mix to remain around 80% in fiscal '26, and that's consistent with 2025. And when we say that we have a prudent guide for 2026, I think there's as much upside in our recurring revenue side of the business as there is in the upfront side. Strategically, we like the balance. Recurring provides durability, upfront reflects areas where customer demand is accelerating, and we have differentiated assets. But we're seeing strength right across the board. And I think that's why Anirudh is talking about share gains.

And our final question comes from the line of Nay Soe Naing with Berenberg.

Maybe one for John. I mean you mentioned about leveraging AI internally. And I was wondering how we should think about that in our models how should we think about your incremental margins going forward? I think with your '26 guide, what you're implying is incremental margins of about 51%, which is slightly below the rate that you've been trending in the last recent or last few years as well. So I just wanted to triangulate with the internal AI leverage and how you're guiding for margin for '26 and how we should think about margin a bit longer term in the age of AI?

Yes. Thanks for the question. I think if you have a look at what we achieved in 2025, we achieved incremental margin of 59%, I think. And I think that points to the fact that there's no near-term ceiling on operating leverage for the company. I mean the company has performed at about 45% operating margin. So there's a lot of upside to that incremental margin of 59% that we achieved in 2025. Now generally, we're more prudent with our guide at the start of the year, and we try to build from there. But -- so I think if you compare the right compare for the 51% that's in the current guide is probably against what we would guide for incremental margin at the start of each year. But -- and I think it's one of the strongest guides that we've ever had.

And then in relation to your commentary about AI and our use of that internally, that's absolutely right. That's what Anirudh is talking about for years now that it's designed for AI and AI for design internally at Cadence, we learn a huge amount from our own internal group in terms of how AI is used. But -- and if you like, I mean, we've -- we've built a great business around emulating hardware and a lot of our AI usage is like emulating engineering flows that -- and we take advantage of those, and they're helping us to get more value out of the R&D investments that we're making. But we expect to do the same as our customers in that when you have access to more engineering capability and being able to do things faster and leverage AI, we'll probably do more R&D and it will be more people, more AI, not less people.

And I will now turn the call back to Anirudh Devgan for closing remarks.

Thank you all for joining us this afternoon. It's an exciting time for Cadence as we begin 2026 with product leadership and strong business momentum. Our continued execution of the intelligent system design strategy, customer-first mindset and our high-performance culture are driving accelerated growth. Great Place to Work and Fortune Magazine recognized Cadence as one of the Fortune's 100 Best Companies to Work for in 2025, ranking it #11. And on behalf of our employees and our Board of Directors, we thank our customers, partners and investors for their continued trust and confidence in Cadence.

And ladies and gentlemen, thank you for participating in today's Cadence Fourth Quarter and Fiscal Year 2025 Earnings Conference Call. This concludes today's call, and you may now disconnect.

All right. Good morning, London. I want to say how is everyone doing, but it's probably not appropriate.

Let's start off with the safe harbor. Today's discussion will contain forward-looking statements, including Cadence's outlook on future business and operating results due to risks and uncertainties. Actual results may differ materially from those projected or implied in today's discussion. So with that out of the way, it's now my pleasure to introduce Anirudh Devgan here, CEO of Cadence Design. Anirudh, welcome to London.

It's good to be here.

Always great to see you as well.

You guys are so fashionable here, I was saying. Everybody is so well dressed.

It's -- for those listening in, everyone stood up and did a twirl at that point. So well done London.

Let's maybe get down to things here. For those who may be new to the story, let's level set everyone. Would you mind just giving us a brief overview of Cadence and where it sits within the semis and systems ecosystem here?

Yes. I think for those who are not familiar, basically, we make products, mostly software products. Some IP and hardware or full stack products to basically design chips and electronic systems. So almost any chip design in the world today uses some form of Cadence products and about 45% of our customers are now system companies, like phone companies, car companies, hyperscalers and 55% are semi companies, semiconductor companies, so they are also increasingly becoming full stack company on the semi side. And then we are -- I mean, one key thing about our software, which is very unique is we are involved in the build-out of AI. There are multiple phases of AI. I have talked about for years, but when all these companies design chips, they use our software, which is not true for most software.

And then on the other side, we -- so that is design of AI. And then we can put AI in our design software to improve our own products and then make them -- I would think about 5x to 10x more efficient, in a 10% to 20% better performance, power performance in the area. So there are 2 ways we are benefiting from AI. And we are lucky to work with all the big players, all the MAX 7, and about or like the 60%, 70% of revenue is coming from about 60 companies, and these are all the rooms who in the tech world. So that's like a brief summary of Cadence.

That's pretty good. I mean maybe just playing on the AI drivers and themes. Maybe could you break down for us what are the actual underlying structural trends as it relates to the 3 businesses you have, EDA systems, and IP. And how does that change as AI-driven demand really hits the road over the next few years?

I mean, AI, of course, is changing a lot of things, but especially semiconductors and systems. So the projection is that semi revenue will cross $1 trillion in a few years, and then system revenue already is like $3 trillion, growing faster. So all these things, the amount of design that is happening is immense, right? So we are seeing that, of course, in infrastructure AI. So I always believe there are -- I believe this for several years now. So that there are 3 phases of AI. So there is infrastructure phase, which is what we are in now. So that's data centers, of course, semiconductors, LLM, so it's more horizontal technologies. And then I think it will transition to more vertical monetization.

So vertical phase, I think one of the biggest phases would be physical AI, so that's cars, drones, robots. And that's already started, but I think it will pick up more. And then science is AI, which is applying AI to real science, like drug discovery or material science, things like that. So in these 3 phases, I want to make sure Cadence is very well positioned. But the amount of design activity is immense, right, all these companies, and we are working with all of them. Now how it applies to the -- it helps all the 3. So actually, at this point, all the 3 businesses are doing well. They're growing pretty well. And we also, for those who are not familiar, always looking at growth and margin, so our operating margin. This year is about 44.5% or something and our revenue growth last year is 14%. So that's a rule of what, 58%? So I'm pretty sure, I think we'll cross 60% in the next -- near future, and that's our goal. So we want to obviously have growth, but profitability at the same time.

And all this, we also keep the eye on SBC, so stock-based comp. So that's about 8.5% or something. So because a lot of people will want to get stock instead of cash. So real margin is, of course, operating margin minus SBC. So that also we have kind of controlled it. And it will increase a little bit, but not a whole lot. So overall, I think the company is in great financial shape and should improve going forward.

So good leverage in the model here. I wanted to maybe talk about specific business segment next in IP and we saw some of the, let's say, misstep perhaps with a rival recently in this space, some of it customer driven, some of it regional aspect. But can you maybe help us understand what are the differences here between yourselves and other players in this market as far as end markets you focus on? And then also, what do you see as the sustainable growth rate perhaps in IP going forward?

Yes, so we have about $5.2 billion, $5.3 billion this year. So IP is about 15% roughly, and systems is about 15%, and EDA, which is our core business, is about 70% rough numbers. And like I said, all 3 are doing well. Now IP is good, but it's always slightly less profitable than software. So IP means like we design certain things. And premade and sell like DDR, critical IPs, PCI, things like that. So I always was careful how much to invest in IP over the years because first, we wanted to make sure that we are very good in EDA, which we are. We have the broadest portfolio in EDA, we have very good customer traction.

And over the last few years, I have invested more in IP for multiple reasons. One, I think we have a much better team now and our -- so because these are protocol-based IPs, like DDR or HBM. So the functionality is kind of like basic, it's given and how customers choose IPs based on PPA, power performance and area. So our PPA, especially for TSMC nodes, has improved significantly in the last few years, primarily because we have a much stronger team, R&D is much better in IP than before. So that's one reason we are doing well. And we focus on, more on what I would call HPC IP and more at advanced nodes. So basically, TSMC is advanced node and there are 5 main IPs there. So there is UCI, which is -- I mean, sorry for all the technical lingo. UCI is chip-to-chip, HBM memory, DDR memory, PCIe and SerDes. So at this point, we have all the 5 key IPs at the most advanced nodes. So that is doing well. And we don't do like a lot of older nodes or a lot of kind of consumer kind of IP.

And then part of our IP business is also Tensilica which is like a core. This is like ARM cells cores. Tensilica is #2 in that kind of CPU and DSP IP. And the good thing with Tensilica is that, I mean, it's growing, and it will be more important in physical AI, but it's almost software-like margins. So our IP business is first is Tensilica, which is like software. Second, on the design IP, it's more HPC AI focused and PPA has improved significantly. And then now there are a lot of new foundries apart from TSMC, like Intel, Samsung, Rapidus, so they also need IP. So I feel at this point, IP will do well. I mean, we have done well for 2 years already. So first year, I didn't talk about it because 1 year doesn't make a trend. Then this year also, we are doing well. And I think '26 should be good. So I expect IP business to grow faster than Cadence average, which it should, given that the margin is slightly lower. I mean margin is not that bad, but it's not as good as EDA.

So I feel good about IP. And we do always some strategic M&A in IP from time to time. So we try to build out the portfolio. Like we bought the artisan business from ARM. We've got secure IC, we bought HBM from Rambus. So yes, I think IP, and the customers want more Cadence IP.

Okay. So a full portfolio, focus on PPA, leading-edge focus as well, maybe sort of double-digit growth is what we're looking at here for the business?

Yes, that should -- now our focus always is win with the winners. So we always focus on the top first. And there, portfolio has to be big enough, but they always buy best-in-class. There is no -- there is very little bundling at the very top because I mean they have enough money and resources, and they're looking for best-in-class. So if we are able to succeed at the top, you can always scale it down. So I think IP business, the other thing, sometimes we can publicly talk about the customers. Sometimes we can't. But IP business is doing very well at the winners at the top companies.

Okay. I wanted to touch on M&A. But maybe before we go there, maybe if we could touch on China. And I think you've been pretty clear that China is a region where you're seeing growth. It's not slowing. You're not seeing any sort of issues. Is that still the case as we turn into '26? And how would you characterize the opportunity set there, let's say, for the next 2, 3 years?

Yes. I think China is back to normal is what I would say. And when we guided -- because we are always prudent in our guide, okay. When we guided earlier this year, we said China will be flat, okay? And now it turned out that China is growing this year, which is good. It's always good to surprise positively. So the reason I said it will be flat is when I went to China last November, they said, oh, '25 will be a very difficult year. It will be the worst year for U.S.-China relationship.

So I think China has been always been well prepared in this kind of '25. And then they say, "Oh, by '26 by end of '25 or beginning of '26, there will be a deal. And they just want to make sure they are not -- after the deal, they are not overly reliant on U.S. So they are like 3 steps ahead of the game is kind of interesting. So we were very prudent in our -- because we didn't know exactly what will happen in '25, but we said like, why we should be more prudent. And that's what happened. I mean -- and there were a lot of other things that happened in terms of, we were banned for 6 weeks and all that. But I think even in the ban, the customers are fairly calm, I think. So overall, we had some issue of like some revenue moved from Q2 to Q3. But if you step back, China will be in 11%, 12% of revenue, which is down from like 16%, 17% a few years ago, but still, it will grow from last year.

And going forward, right now, I think the situation is somewhat stable. The customers are designing a lot of things. I mean infrastructure, AI, of course, all these big companies are Alibaba and all those are designing chips. And also physically, they are huge. I mean there are like 5 big car companies. They're all our customers, designing chips. And then there are all these consumer companies like Xiaomi, which is also doing cars, phones. Lenovo, they're all doing chip design. So it seems stable at the moment.

Pretty good. Again, I think I said I want to touch on M&A and maybe 2 parts to this question on Hexagon in particular. It looks like a good deal. Where is your integration priorities here with this business? And what are the main milestones we should be focused on? And then maybe secondary to that, is there a sort of genuine revenue synergies you can talk to today that maybe benefits your position in physical AI applications?

Oh yes, I'm very excited about Hexagon and working with Ola, who's the Hexagon Chairman, and because they're -- I think it's like a diamond in the rough, the simulation business of Hexagon. Because Hexagon wants to focus on other things. And it was like a one -- only one simulation asset, whereas we can integrate it much better in the Cadence portfolio. And we are always very careful about M&A. Because anyway, organically, we're going to go well, and that's the most profitable way. But from time to time, we will do M&A if it makes sense. So the reason for Hexagon was, is primarily for physical AI.

So the 15% of our system business is anyway growing like 20% plus for 5, 6 years. I started all this in 2018. That was not -- that time was not clear. People said like, what is the system simulation and EDA, what is SDA and EDA. But we knew what the customers were doing both on the system side doing silicon, and silicon companies doing systems. So now it's like obvious. But one thing that is -- but '25 is not '18, right? So things are changing in the system business. And the growth part of the system business, I mean, we don't need all of the -- we have enough portfolio, but the exciting part of the system business to me are 2 parts. And that's what we want to focus on. And I think we will cross like $1 billion run rate in systems reasonably soon, assuming M&A closes and all that.

So one part of system business, which is high growth is 3D-IC which is close to the chip, right? And we have a very good position with Allegro, because 3D-IC is another word for a system in a package. Now there's 3.5 DIC, did you know? That's possible, right? It's 3.5 DICs, right, which is a combination of 2.5 and 3D. But anyway, I think that one focus will be 3D-IC, which is Allegro, packaging, clarity, thermal, electron, all that, okay? And then the second -- which is going to be high growth anyway because all the segments in semiconductors will go towards 3D-IC or 3.5 DIC. And then the other part will be physical AI. So I talked for a long time about the 3-layer cake. I don't know, people say like, are you like a bakery or something?

So if you don't -- the reason talk about 3-layer cake is because unless you are like a 2-year old, when you eat cake, you eat all the layers together or consume all the layers together. So what are the layers of the 3 layers? So there AI, of course, stop with a layer, which a lot of people forget is principal ground truth, right? How transistors work, how molecules work, and bottom layer is silicon or domain-specific silicon. And then there are 3 phases of that infrastructure, physical sciences. So in my mind, this is like the 3 by 3 metrics, 3 horizontal technologies and 3 vertical applications. So now I mean, we can talk about this a lot. Like people who graduated just like a few years ago, they said, "Well, all I need is AI, right? What do I need anything else? I can fit a model for everything in life. And people who graduated 30, 40 years ago, they said like, what's all this [indiscernible] ." I need ground truth. I need to know how things really work, not a model, okay? I said, "Well, I don't want to take any sides. You need both. And actually, in our algorithms, always there was fundamental algorithms and data-driven algorithms, okay?

Of course, they're not much more powerful with AI. So anyway, we need all these 3 things, okay. Now with physical AI, all the 3 layers of the cake will get transformed, okay? So the bottom layer, which is silicon is, of course, different. Just look at Tesla or BYD. The chips are much lower power. I mean they are all custom chips for physical AI, whether it's for cars or robots or drones. There will be more mixed signal, and there will be lower power. So anyway, we are very well positioned for that. We were -- we didn't need Hexagon for that. We were already very well positioned. All the big auto companies, semi companies are big Cadence customers and then all the newer ones we are working with, like I mentioned. So -- but the other 2 will also change, okay? And I know this for a while anyway, but what the AI model, what will fundamentally change in the AI model, and there is more and more talk about this is, the AI model will move from a text model like an LLM model or a word model, WORD to a world model. Okay, my kids say, my word and world sound the same. So this is the one with the L, WORLD model or the physical model. And all these companies are working on it.

So what happens in a physical world model, like in LLM model, you can train -- if you train the transformer, you have all the data on the Internet because you have all the text on the Internet. But on a physical world model for robotics or cars, the data is not available. So you have to gather it like you have to hold this bottle. And so either you put sensors and do that. So that's very slow or you have to put simulation in the loop to generate data. So first, the model will change to a World model with an L. And then you can put the middle layer, the principal simulation in the loop and accelerate with AI. And the main technology you need, there is what is called multibody dynamics or simulation of robots and card. So Hexagon has the best multibody dynamic simulator, #1 in the market.

So then that's the reason to acquire Hexagon so that we can be as relevant in the physical AI as we are in the infrastructure.

Got you, yes. Maybe if we stay on that point there because I think that's quite interesting. If we are -- it almost sounds like you're saying we're utilizing transformers to make models. But as we move to context awareness in the physical world, there's 2 possibilities. We can have simulation readiness in the loop or we can have a real-time sensor appreciation. But it's a software solution is what you're adding?

Yes, both of them yes, we'll have both. You'll have the answers anyway on the silicon side. The silicon will change. But the inference will change because -- and the thing is that the physical AI will reinforce infrastructure AI because even like Tesla or BYD, if they run the inference on the car, or the robot. Of course, the silicon is different. The inference is different, but the model is trained on the data center, right. So it helps the -- it might pay further.

But the thing will be that we just want to be -- make sure we are completely ready for this phase as we are. Like in the infrastructure phase, we are working with all the leading players. And this one, we needed some pieces, especially I think simulation in the loop will become even more critical. So that's why. So Hexagon is a great asset for that. And then we can integrate with the beta. Beta is the other acquisition we made, which is also doing very well to make a full flow for this kind of physical AI.

Got you. So good integration with there. The other question I wanted to ask, maybe I'll open up the floor after. We are getting used now to collaborations for, NVIDIA is having with the ecosystem broadly. And 2, in particular, that are relatively close to home, NVIDIA's partnership, first of all, with Intel and what opportunity that might bring for you guys? And maybe there was a nonexclusive deal done or collaboration with Synopsys and what your sort of views were there? And should we focus on that nonexclusivity in that deal when thinking about yourselves?

Well, we worked with NVIDIA and Jensen for years. So actually, and we even released -- some of these recent news is like porting their software to GPUs and all. But I've been doing that for years. So if you look at even Cadence announcement from 2 years ago, it covers most of these topics, okay? And we do it in EDA, we do it in SDA, we do it in biodrug-discovery.

So we are glad to have NVIDIA as a great partner. And our business with NVIDIA is growing. And actually had a joint statement with Jensen ready for last week, but we didn't release it because we didn't think we needed to release it. But our partnership with NVIDIA is fabulous, and we cherish that. But in terms of investment, we would like to get business from NVIDIA and investment from you guys. That's why I'm here. That's how we also want to be -- even though NVIDIA is an amazing company, I think like I said, there will be a lot of -- as AI evolves, there will be a lot of other great company. So we want to make sure we are -- the benefit of EDA's horizontal technology. So we want to work with all the great companies. And Intel, we'll -- I mean we have, of course, my predecessor is there. We have a lot of discussions ongoing with Intel to see how they progress, especially as they focus on [indiscernible]

Precisely. Maybe with that, we'll open it to the floor and see if anyone's got burning questions. Not at this point.

Maybe just one with a minute to go. Clearly, with -- there's a little bit of margin pressure as you transition to annual subscriptions and some of the SD&A business. Help us understand what is that? Why are we doing it? What sort of margin pressure should we expect?

No, this -- I mean, this is -- we always do things for the long run, okay? So like we have a recurring business in EDA, which is great. I think what happened in SDA is that the accounting treatment is different. Then EDA by nature is recurring. SDA by nature because of the accounting rules is upfront. At least some of the business we are buying upfront or we bought -- so we can't change that. But what we can change, which is a very good business practice is annual subscription because then we report annually.

What you don't want is like multiyear deals and you take upfront, that's not healthy for the business long term. So if we get Hexagon, I think they were already doing the subscription, conversion to annual subscription, like 60%, 70% of the way we'll take it all the way. But that, in the end, you recover all that revenue and more in the future. So there might be some 1-year impact. But still it's not going to be that significant given our scale. And the other thing is our incremental margin on our organic business is close to 60%. And our goal, of course, is a 50% plus incremental margin, which we have done for almost 8 years now. So even this year, it should be pretty good.

Now last year, it was slightly less incremental margin because of beta acquisition but then we made it up this year. So if you combine '24 and '25, our incremental margin is 53% or 54% because it was slightly less last year and it's more this year. So maybe '26, '27 will follow a similar path. But in any case, our incremental margin will be more than our operating margin, no matter what with M&A. And whatever transition we do for the long run, we will recover that in '27. So I think we'll not do any M&A that destroy this fundamental financial model that our margin goes up, of course, revenue should keep improving. And then we buy half of our -- use half of our cash flow to buy back stock, primarily to make sure there is no dilution.

So if we do like 8%, 9% SBC, we buy back more than that. So that is not going to be changed based on M&A. So that is intact. But there may be some movement like 1 year, slightly less incremental margin other years slightly more, but it will be more than what our operating margin is.

Sounds clear. Anirudh, thanks very much. Clocks taking our time. Thank you.

Hi, everybody. My name is [ Natalia Winkler ]. I'm a semiconductor analyst here at UBS. I'm very excited to have Anirudh Devgan with us, CEO of Cadence Design Systems. Before we start, I want to read a quick forward-looking statements.

Today's discussion will contain forward-looking statements, including Cadence's outlook on future business and operating results due to risks and uncertainties. Actual results may differ materially from those projections or implied in the discussion today.

So maybe we can kind of start with the big picture, right? And when we think about Cadence, what's Cadence's role in this very fast developing semiconductor ecosystem?

Yes. First of all, thanks for your interest. Basically, Cadence, we provide product, mostly software and some IP and hardware products to design chips and electronic systems. So what we like to say is almost any chip design in the world today uses some form of Cadence products.

And so that's true for, of course, the traditional semiconductor company. But also about 45% of our business is coming from system companies, like car companies or phone companies or hyperscalers. So that's the mix of our customers. And of course, there's a lot of design activity now for AI and other things.

And maybe you can speak about the design activity for AI. Like are you guys seeing that more on the data center side more in the long tail of edge applications and how -- where that customer base is expanding for you?

I mean, definitely on the infrastructure side and data center. I mean, we always -- I've talked for a while that I see like three phases of AI. The first phase, which we are in is infrastructure. So that's all the data center and even some like laptops and all but mostly, of course, data center. The second phase being like physical AI. So there's cars, drones and robots and third phase being science AI which is more like biology and materials.

So I think we are still in the first phase, most of it. And I think that first phase still has a long ways to go. But we are also investing in the other two phases.

Well, an excellent because that was kind of my next question. If we think 5 and 10 years out, like how should we map those phases into kind of your opportunity [ to that ]?

And of course, these are difficult to predict exactly, but I think the infrastructure phase, of course, going gangbusters right now. And the projections for that are very optimistic in the next 3 to 5 years that the amount of compute and AI usage will be exponential. So I think we still see a lot of demand in the infrastructure phase.

And then the physical AI phase, in my mind, of course, already design activity is starting, but to reach critical mass is maybe 3 to 7 years. I mean some Waymo and Tesla is already doing a lot of self-driving. But I think within a few years, it should become a lot more mainstream. And we're already seeing design activity in preparation for the physical AI phase. So if the infrastructure phase is from now to at least 5 years, then the physical AI phase, I think, is 3 to 7 years from now. in terms of reaching like -- and then sciences, even though we're doing a lot of science work and drug discovery and things, I think it will take some time. So that I put like more 5 to 10 years from now. So these three phases. But the investment is there. Most of it is in the first phase and then proportionately less in the second and third phase.

Excellent. And when we think about maybe a little bit more near term, so you recently increased your calendar 2025 revenue growth expectations from 12% to 14%. And specifically we talked about very strong backlog, right? And I'm wondering if you could talk about -- a little bit more about which of the segments you're seeing kind of most of the strength in the near term as of this year and the next year as well?

Yes. We had record backlog. We reported end of Q3. And we also indicated that, I think Q4 -- I mean, of course, we haven't finished it yet, but all the indications are that we should end up with another record backlog end of Q4. And we are always focused on -- I mean, for those who are familiar with us, you already know, but we are not that -- we always focus on revenue growth plus margin. Our job is to make money for investors. So our margin this year is roughly 44%, a little more than -- like 44.5%, and revenue growth is 14%, okay? So that's a rule of 58%. And we want to keep building on that. We have increased margin every year for the last 5, 10 years, and we can keep doing that going forward.

And the revenue growth Yes. I mean we always want to make sure that we have profitable sustainable revenue growth. And I think last 5 years, our CAGR is about 14%. And then we'll see how things go in the future.

Excellent. And maybe we can talk about the EDA side of business. And specifically, as we think about different AI applications, like how would that be changing EDA business model over time?

Yes, yes. So the AI first has two implications because one question is, okay, how does AI affect software itself, what we sell. But for us, one thing to remember is we are also -- there are only very few software companies that are helping build AI. So EDA like especially Cadence, whether it's partners like NVIDIA or Google or all the big [ MA7 ] companies, they are using our software to build AI, right, whether it's CPUs or GPUs or all the custom chips. So part of our business is growing because there's a lot of design activity for AI, okay?

And then the other part is applying AI to our own products. So we can make our products much more efficient. So I think there's at least 10x productivity improvement we can deliver by applying AI to our products. So if you look at our products in the last 20 years, we already improved productivity by 100x, by a lot of other methods, more classical mathematics, but AI can help us next 5 years, improve by 10x okay?

And so the question always is, okay, what does that mean in terms of usage. And so one thing to remember is for our customers, the workload is exponential. So if you look at the chip design today versus chip design in 2030. Right now, the chips like the biggest chips are $100 billion or $200 billion transistors. In 2030, there will be like 1 trillion transistors at least 10x bigger plus they have 3D-IC, all the software. So the design complexity is going to be 30, 40x more than now, which is very different from the worry of AI disrupting software is assuming that the workload is constant or only growing up by GDP or something. But if the workload is going up by 30, 40x, all our customers want to use our AI tools because there is no way they're going to hire 30x more engineers from now to 2030. I think they will hire more engineers, maybe 2, 3x more, and the remaining 10x gap has to be made up by software.

So all our customers, if you look at -- we are part of R&D, right? We are engineering software. So if you look at percentage of R&D spend on Cadence and EDA has gone up from 7%, 8% to about 11%. So R&D is going and then a percentage of R&D allocated to us is going. When I talk to all the big CEOs of our customers, they wanted to continue to do that trend. They would rather spend on automation and compute to improve the design efficiency.

So our goal is, given the workload is exponential is also to get to provide value to our customers, so we become more essential to their R&D operation.

But there is a lot of ways in with AI. I mean I can give a lot of examples of how AI can improve like the tools can get 5 to 10x better A lot of times, the PPA, power performance in the area can be 10% to 20% better because AI is doing a much better job of optimizing the design than a human can do. And 10%, 20% is huge for power in area.

And I guess when we talk about EDA business, I wonder how does that really translate in pricing? And I guess the concern would be in a world of Agentic AI if some of the Agentic features can kind of reduce -- potentially reduce the number of seats of EDA software that you need. Like is that at all a threat or really the pricing per license and the number of licenses people will actually need for offsetting that potential headwind?

So if you look at our license usage, it's almost exponential. Of course, pricing improves a little bit over volume, but the number of license growth is -- and the reason for that is typically when something is faster, you do more of it.

And even with AI agents, so the way I always -- I mean, I said this for a long time to really do a good AI solution, you need multiple factors. So there is the AI itself, but you also need the base tools, the ground truth, like how the transistor operates or the classical kind of EDA tools and then the compute that it runs on.

So this -- and this is what you're seeing with agents now. The agents will do the AI, but they will also call a lot of tools, which they are already good at doing. Like if you're doing placement, that's a solve problem in mathematics. You don't need to run it with AI. Optimization, maybe you run AI and then you call these tools. So typically, when we deploy our tools like Cerebrus that give huge benefit. We have five big AI platforms. They will use a lot more of our base tools. So the actual number of usage of the tools is only going up. I mean one of it is with AI. The other is, of course, the chips get bigger and yes.

And maybe on that point, if we could talk about how the hardware business is performing given the traditional kind of refresh cycle and really how we should think about the synergies of the hardware and software businesses going forward.

Yes. So part of our business is we sell a hardware system. I mean we call it hardware, but it's hardware software together, which is like an emulator. So it will -- like it will verify the design like 1,000x faster than you can do on a regular silicon. And we sell like a rack system with hardware and software.

And almost all the big chips, all the big AI chips use palladium, which is a hardware platform to build these things. And so the benefit of that is that you can basically emulate the chip before it is fully done or comes back from TSMC.

So like about 2 years ahead, you have a model in palladium and then you can run software and do full software bring up everything like that. So Palladium and these hardware systems became basically essential to design of all modern chips. And then we -- in our verification suite, we will also sell software that goes with palladium. So that's one of the big advantages that Cadence has and the reason we are doing well in the ecosystem, especially the AI ecosystem and a lot of the other big like mobile and communication, but especially in AI because the chips are so big is the strength of our Palladium system.

And palladium, we build ourselves. We designed the chip ourselves. It's fully integrated. We have a 10-year lead in terms of how to build these things. And then it pulls in the software as well.

Excellent. So pivoting a little bit more into the IP side. You guys have obviously seen very strong momentum in the leading edge IP. And arguably, in contrast to some of the peers, right? I think like Synopsys, for example. So wondering how you guys are seeing the dynamics of the IP portfolio going forward, really and where is kind of the most growth coming from going forward?

Yes. IP is performing well. So we have five segments we report. I mean, five main areas. I mean three of them are EDA and then IP and systems. So -- and so right now, all five are doing pretty well, okay? And EDA has been a traditional strength of Cadence okay. And then a few years ago, we invested more in IP and systems because our customers are becoming more system companies. And IP, we got later into it, but we are always careful about margin, not just revenue growth. But I think now with AI and all, I think in the way the IP, there are five key AI IPs that we are investing in, things like chip-to-chip interconnect, HBM memory, DDR, SerDes, these kind of things. And I think we are well positioned with those.

And also a number of foundries is increasing because there are more advanced node foundries. The combination of our portfolio and then more demand for AI systems and foundry. I see good growth for the IP business going forward.

And should we think about the IP business from the standpoint of kind of license type of revenue or really there's increasing opportunity for royalties as well?

So we already have IPs that have royalty and that's very profitable business. So part of our IP business is Tensilica, which is used in a lot of kind of edge applications and edge AI application. So Tensilica is, I think, the #2 kind of platform after ARM, which is like a license core with royalty. And that's almost like software margins, which is very good for our IP business.

And then design IP, which is like these protocol based like SerDes and DDR. Those are more like usage licenses, with some royalty but mostly usage. But overall, I'm happy with the mix. I'm happy with the profitability. The profitability of IP is still lower than because EDA software business is, of course, we have 90% gross margin plus. But still, the growth is higher. So we always -- I always evaluate each business on this rule of mix of revenue and margin. So even IP is lower, but I think it can grow higher than Cadence average. So at this point, we are investing in that.

And I guess, if I think about the growth in IP business going forward, is it -- how incremental would be that royalty business you already have to the growth rate or really, the bulk of the growth rate is affected from the license kind of type of engagements?

Yes. I think it's both, but the Tensilica part -- the design IP, which is less royalty is growing faster than the silicon but Tensilica part is still significant. But the growth, to answer your question is more on the design IP because of all these chips being designed, which are more kind of AI HPC. Now as it moves more to physical AI, maybe there will be more Tensilica growth in the future. But right now, it's more of these big data centers, which are design IP related.

Excellent. I was hoping to talk about the Hexagon acquisition and how you -- maybe you can talk about how you see the synergies and especially specifically, given the track record of acquisitions that you guys have, how you think the integration process will go there.

Yes. We are always measured in acquisitions. We always say organic is delicious. So we are an R&D-driven company. We want to -- because that's the most profitable way to grow anyway. But from time to time, we will do M&A, especially if the opportunity is good. But that's always the second preference for us. So the question is why did we do Hexagon is basically for physical AI, okay, basically for physical AI.

So we have a system business, which is growing well last 5 years. So half of the system business is on -- focus on 3D-IC. That's a big trend for all these AI systems, which is multiple chips in a package and all the analysis that goes with it. So in systems, that's a huge trend, and there will be a trend for next 5, 10 years.

The other thing I'm always optimistic about is physical AI, okay? And in physical AI, I think everything is going to change. So if you look at the 3-layer keg again, which is AI and then principal simulation of the ground truth and the silicon. So when you go to physical AI, all three are different. Of course, the silicon is different because it's a power constraint. So you look at the Tesla car has this AI 4 chip or AI 5 chip is very different than a data center. Same thing with BYD and all the other, Rivian all these companies.

So the silicon will be different. But silicon, we are already well positioned, and there will be more mixed signal and all in cadence strength. But also the AI model is going to be different. So I mean, all this talk recently anyway off a world model, WORLD, like a physical model rather than an LLM model. And the thing is in LLM models, all the data is available already on the Internet to train the model, but in a physical model for a robot or a drone, the data is not available. And the data is not easy to get because they have to put all kinds of sensors on people. And so in there, the simulation becomes critical.

So Hexagon D&E business has the leading multi-body stimulator. It's a robotic simulator in the market. So that will really be critical for this physical AI models. So that's why we had a good discussion with Hexagon and they wanted to -- actually, they are building their own robot. They wanted to focus in a different way. And all the software businesses, which they call [ D&E ], we acquired, and we can integrate in our flow. And so then in system business, we'll have one half focused on 3D-IC, one half focused on physical AI and both are big growth drivers.

And the physical AI opportunity effectively opens a new customer base for you, right, kind of the emerging physical AI?

Absolutely. And there are some traditional customers there too, like cars. That's a big business already. And of course, it's going to a lot of change with self-driving and electrification. But a lot of the business that Beta CAE, which is the acquisition we did and Hexagon is already in automotive. But then there could be newer things like drones and robots. So all three will be critical, yes.

And have you guys sized sort of that physical opportunity specifically in the fields you will be playing with the total kind of total addressable market?

Yes. It's difficult to say. I think it can be huge, but I don't have any I think it will be -- the main thing I want to make sure is we are already well positioned in infrastructure, AI. I want to make sure that as this new thing happens, we are also well positioned physically -- it's difficult for me to point out exactly, but I think it will be significant. Yes.

Excellent. So a key competitor in our space Synopsis, they recently did an acquisition of Ansys, right? And I think one of the applications there was also the physical and the digital twin capability Ansys had I'm wondering how that has changed compared to landscape for you.

We are doing this from 2018. So I don't know if you go back and look, I am the one who -- because one when I was supposed to take over as CEO, one question from the Board was, okay, EDA is a good business, but what's the future of EDA? And this very different time at that time, 2018, this was before all the AI and. And I always believe that silicon and system have to merge. And you're seeing that, of course, now it's obvious, right, whether it's NVIDIA or Qualcomm or Broadcom and all the hyperscalers. So we have been investing from 2018 in this. And our system business has grown like, I don't know, 25% a year for the last 5, 6 years. So we have a pretty good portfolio, and we are focused on the high-growth part of the system business. like I mentioned, 3D-IC and physical AI. So it's a good customer. We're growing well. Customers are happy with our solutions and we go from there. We are already competing with them separately. Together is not -- we just want to focus on what we can deliver to our customers.

Great. So I guess coming back to more financial side. So from the regional standpoint, I think in the recent quarters, you've seen significant strength in China business. I'm just kind of curious what's driving the outperformance for gains versus the peers there. And how does China actually fit in kind of -- actually, maybe this is more a long-term question in your transition.

China is a good business. I think China, if you step back, has come down over the years. And of course, semiconductor companies have a lot of China exposure. We -- from an EDA or a software standpoint, we used to have, I think, 16%, 17% used to be China a few years ago. Now it's like 11%, 12%. Still good business for us in China has a lot of design activity, as you know, both in the infrastructure and physical.

I think this year was a weird year for obvious reasons, a lot of geopolitical. So when we started the year, we were pretty conservative. That's our culture anyway. We'd rather print the numbers than then project something we are not sure about. So we were pretty conservative in our China assumption because we knew that there will be a lot of uncertainty. But we are doing better than we thought which is good. And I think China business should grow this year.

But the behavior of the customers is fairly normal to me, so it looks like. And we had some issue in Q2, Q3 I mean there was some -- because of some of the restrictions, some of our business moved from Q2 to Q3, but now all those things are resolved. So the shape of the curve in China is a little different for quarter-by-quarter. But if you step back and look at the full year, I think we will end up around 12% -- 11%, 12%, something like that. So I feel that the -- I mean, of course, it's very difficult to predict the geopolitical, but seems stable for now and the design activity in China is back to normal and they're investing a lot anyway in silicon and systems in cars, robots, there are a lot of companies in China.

Excellent. And so maybe a little bit on your margins. So your non-GAAP operating margins is 44%. How should we think about the trajectory from here? And specifically, in light of the acquisition, I think you mentioned that may have been somewhat under-invested. How should we think about the impact from that?

Yes. We manage the overall margin anyway for Cadence. I mean MSC is still will be important, but it's part of -- a small part of Cadence. So what our margin is about 44% or a little more this year. And what we always look at is incremental margins. Like if we add $100 million of revenue, what is the margin on that. So for the last several years, I don't know, 8 years running, our incremental margin is 50% or better. okay?

And that's what -- so there's still a lot of room for improving margin from 44%. Actually, our organic incremental margin is close to 60%. Okay. Now if we do some M&A, then it comes down to maybe low 50s. But M&A, we will do if it makes sense and sometimes it does make sense. So yes, there will be some effect on margin from M&A. But overall, we'll try to make sure that the margin still improves for the company over time. And we are always shooting for 50% plus incremental margin.

And that applies to short term as well as long term, right, that commentary?

Now sometimes, there's like in a particular year, you closed it depending on when you close, there could be some Like, for example, last year, our margin was slightly lower incremental margin, but this year was really high. So if you average the 2 years out, our incremental margin is like 53%, 54%. And so it may happen in '26. Our incremental margin is a little lower, but '27 it may accelerate. But overall, our goal is still we can drive margin improvement and also make the team more efficient, of course, use AI internally. So again, revenue growth and margin, both -- there's still room to go. Yes.

And when we think about the physical AI opportunity, how should we think about the kind of margins in that potential business going forward several years out.

That should be good. We don't...

Look, there's no structural change to either gross margin or operating margin compared to...

No, no, no. This is mostly software no. And also the physical AI, of course, see the main -- a lot of businesses still infrastructure, AI and all the AI build out. The good thing with physical AI is that it also reinforces infrastructure AI. Just as an example, like Tesla, of course, they run the model on the car, but they train it on the data center and same thing with other things. So it will be additive to the current trend, and it will reinforce the data center side. So no, we'll make sure margins are good anyway. And we have done this for like 8, 9 years, if you look at our margin trend.

Excellent. And maybe in the last minute or 2 here, how should we think about the capital allocation priorities for Cadence after the deal closes with MSC and...

We, like I said, most of it is organic investment. And of course, we generate a lot of cash. We also -- there's no change there. We will take 50% of our cash flow and we buy back our stock, that we have done that also for 7, 8 years. And the reason for that is that we are always looking at SBC. So we also track margin minus SBC. Stock-based comp is a very important thing for us. So it's about 8%, 9% right now, which is still better than the peers. Because to me, 44% margin doesn't mean anything if your SBC is so high because all our employees would rather get stock rather than -- so we're also very careful on SBC.

Now it's going up a little bit, but overall, still much better than everybody else. And then the goal of buying 50% back is we want to make sure that there is no dilution. So we're actually buying more than we issue in SBC.

And then the remaining cash, we'll see if we do some opportunistic M&A or -- but it does not change our model, which we have done. Good thing Cadence is a predictable business. We are integrator of value compounder of value. So this should be the same. And this kind of M&A doesn't change our financial model.

Excellent. I think this is it. Thank you very much.

Yes. Thanks a lot.

Perfect. So we'll go ahead and get started. I'm Joe Quatrochi, the semi and semi-cap analyst here at Wells Fargo. Excited to have the Cadence Design Systems team here, John Wall, CFO; as well as Matt from the IR team. Thanks for joining us.

Thanks for having us.

First, I think I've been asked to read the safe harbor. So let me get through that real quick. Today's discussion will contain forward-looking statements, including Cadence's outlook on future business and operating results due to risks and uncertainties. Actual results may differ materially from those projected or implied in today's discussion.

Perfect. With that, maybe before we get like just into kind of more pointy questions, John, curious, like can you talk about just what you think is underappreciated or maybe not understood by investors from the Cadence story?

Yes. Great question, Joe. Essentially, we're -- I guess, Cadence is so central to the whole AI infrastructure stack, and it's an engineering software company. I think we get bundled in with a bunch of -- well, we get bundled in with semi sometimes. We get bundled in with software companies, but it's engineering software and the workload is growing exponentially. So therefore, when you're central to AI and everything, it just generates so much work for us, and there's a whole flywheel of I suppose -- with the amount of work that we're doing with customers, it just gets more and more work.

And if you have a look at what we've done over the last 10 years, double-digit, low teen kind of revenue growth with constantly improving operating leverage kind of speaks to the strength of the model. So it's just that essential nature and working with those biggest companies that are driving the whole AI world, it kind of informs our R&D road map as well. But -- so it's quite a good ecosystem.

Okay. You started this year thinking total revenue grow somewhere like 12%. Now your latest guide is about 14%. Can you just talk about like what's been the biggest upside driver as you think about like relative to the beginning of the year, what's been the biggest surprise that's been positive?

Right. I suppose that AI workloads are getting so complex that the growth in the license count is taking off on the EDA -- core EDA side. But we're seeing strength across the board across all our businesses because Cadence used to be like an EDA company, but now you've got that core EDA business and IP business and the System Design and Analysis business layered on top, and they've all become quite sizable. And of course, they all work seamlessly together. And I think we're just seeing strength across the board. But nothing is getting any less complex, which is probably good for us.

No. Yes. You think about -- you talked about, I think, last quarter, exiting the year with record backlog. And looking into '26, can you kind of help us understand -- you touched on it a little bit, but like what is driving that record backlog? Where are you seeing like the area like the strongest orders from customers? And like what are those things that they're really focused on just trying to accelerate the road maps?

Yes, sure. I mean that's great. I mean we have tremendous momentum right now with strength across all lines of business and multiyear contracts that we ended Q3 with record backlog. We back tested the last 10 years. I think there's only 1 year where Q4 bookings didn't exceed Q4 revenue, and we've had such a strong start to Q4. It looks like we'll exit the year with another record backlog, which kind of bodes well for planning for next year and everything.

Yes, I mean there's just so much strength across all lines of business, lots of design activity. The complexity is kind of through the roof between AI, high-performance computing that all of the automotive companies are -- we're kind of pulling in new customers, I think, across the systems landscape as well.

And like when you think about '26 and relative to prior years in guiding and I say you're going to wait until February

[Technical Difficulty]

I'll try to speak up. I think like, I guess, as you think about '26 and you think about visibility, right? And this year, you've seen really strong demand from our systems and hardware refresh cycle. How does that -- how do you think about like your visibility in '26 relative to prior years in that dynamic?

Yes. Visibility is probably as good, if not better than what we've had before. Like in the past, we've always seen strong visibility on the [indiscernible] particularly on the software side [indiscernible] so it's quite [indiscernible].

[Technical Difficulty]

Okay, that's helpful. Maybe you can hear me. I don't know. So maybe on the core -- let's stick with the core EDA side for a second. How are you thinking about when you kind of parse out the business like AI growth versus non-AI customers, what is that -- what are the growth algorithm, I guess, look like? And like how do we think about the split of that business?

Right. So I guess most of you have [indiscernible] and you have that AI layer on top that [indiscernible].

So I guess like think about AI as kind of broadening your customer base? Or is this additional spend by existing customers? It sounds like maybe a little bit of both but -- yes.

[indiscernible] so that kind of deepens the relationship, not just as a customer but as a partner. And then it goes to a bunch of extra customers across the board, right across systems.

Okay. Maybe I think in the past couple of quarters, I think what's been interesting to me is you've highlighted particularly like strength from foundries. Can you kind of maybe expand on like what is driving that? Is it the foundry itself? Is it the foundries customers? Like just kind of help us understand that dynamic.

Yes. Well, they're key ecosystem partners since everyone wants to design silicon these days. But -- and then I don't think it's fair for TSMC [indiscernible] using the silicon. So they probably don't have the cycles or the bandwidth to cover everything that the world requires that -- which I think is great for the Samsung's, and Intel and GlobalFoundries and of the world.

The -- yes, I mean, there's so much going on there. And of course, they're key partners for us because the more closely we work with them, you kind of build out the flows for future customers and then you get -- people tend to adopt their flows. So it's -- yes, it's a key relationship for us. What we've seen is -- I mean, we're very, very strong at TSMC, but we've been increasingly engaged in places like Samsung and Intel. And more recently, I think you've seen announcements on what we've been doing with Samsung. And it's kind of -- it's a new relationship for us really because we were never that strong at Samsung or Intel before.

That's helpful. Maybe spend a second like on non-AI semis, right? Like we've seen -- obviously, it's gone through kind of a pretty difficult cycle. Maybe things are kind of bottoming out, starting to maybe bounce a little bit off the bottom. Have you seen like a change in like their EDA -- whether it be like discussions or just kind of their planning in terms of if the market is kind of their market or their financials have kind of found some stability? Like how does that translate into kind of thinking about EDA spend?

Yes, we're seeing them come back a little bit. I mean the last couple of years have been lean years for many of those non-AI semis. But they seem to be gradually doing more and more. But this year, we're seeing bigger engagements and I think they've probably found their base now at this point.

Yes. Do you think that accelerates into next year?

I mean it's hard to tell, but everything seems to be accelerating. There's so much complexity in design that the tools become more and more essential to what our customers are doing.

Okay. That's helpful. Maybe back on AI, you guys have kind of been explicit in saying we're not ready to kind of quantify things in terms of the benefit of AI, what we think it could drive for EDA. Like are we getting closer to that? Or how -- where should we kind of think about where we're at in that?

Yes. It's -- I mean it's hard to bifurcate it because I mean, you've got so much revenue coming from the traditional kind of EDA workloads. But these AI tools that you can layer on top and the ability that it gives engineers to use multiple licenses does pull through quite a lot of additional revenue for us. Now it's mainly recurring revenue or ratable revenue. So it kind of shows up a bit more slowly, but we're definitely seeing that pick up faster than we originally anticipated.

Okay. That's helpful. Maybe switch gears a little bit. The hardware cycle has been a really strong year, record year again. I think you talked about last quarter thinking that you could see further growth next year. Like what inning do you think we're in, in that refresh cycle? Like how do we think about like that relative to prior cycles?

That's -- certainly it still feels like early innings because everything has become so much more complex, which means the hardware emulation becomes even more critical and more important. And there strategic purchases for our customers now, not just the need for design, but it's really, really important for them to get silicon right first time and do what they can with -- in terms of emulating the system because it speeds up their time to market essentially. The -- and with the large installed base we have, I don't think there's anything out there to touch it. We're still seeing demand. Customers are asking for our Z2s if we can't get them to Z3. So it's -- they're very, very popular.

I think you guys try to make some investments like in the supply chain to ramp up that. Like where are we at in terms of...

We're kind of a constant kind of rate of expanding that production line. We like to keep a strong backlog of orders. But -- so we're constantly increasing the amount of hardware we produce.

The revenue in each year and each quarter is probably throttled by the volume that we can supply and we're still building a backlog of orders. We tried to aim for somewhere between 8 and 22 weeks of lead time. We're somewhere in the middle of that right now. And kind of throttle the -- we kind of throttle everything from a revenue perspective based on production. It's kind of the sweet spot for pricing and maintaining visibility into the backlog.

Okay. No, that's helpful. So I mean, when you think about, I guess, like you're talking about 6 months visibility of hardware. I mean, it sounds like -- I mean, it's a bit -- maybe a bit more than that right now.

Well, the BU itself, like the business group, they would tell me they have a lot more visibility. I don't like -- we're very prudent with the way we guide. I don't like including things in the guidance to see them in the pipeline. But -- and then at that will kind of -- we'll factor the typical kind of percentage closure we'd have our conversion rate for those opportunities. But they can see now from the large installed base and the kind of pattern of behavior from how customers purchase and replace older systems that they feel that they have more predictability there. I'm just cautious. I don't want any inventory issues and we'd rather slow and steady and we keep the -- what's working is working for us.

Sure. How does that refresh work in terms of customers kind of do a trade-in thing? Or do they actually keep some of those like Z2s and things like used kind of workloads.

Yes. So what tends to happen like you take Z2 and Z3 that what you're trading is like server room space generally. On a Z3 with the same footprint, you probably do twice the capacity. So if you can swap it out, you don't -- if you have a limited finite space in the room when you're swapping out one system for another, you're getting double the capacity for the same kind of footprint.

They tend to pull a lot more power, though. But -- so often, there's some power changes that need to be made for -- in terms of the building infrastructure itself. And then we have this company that we bought a few years ago called Future Facilities, that people use that software to basically create a digital twin of their server rooms and decide to optimize what they can fit in and the power requirements for the room before they actually make the changes.

Interesting. That's interesting. Maybe shift gears a little bit to SD&A. In September, you guys announced the acquisition of the Hexagon, Design and Engineering business. Maybe can you talk about just where this business fits in the Cadence portfolio, what you're most excited about?

Yes, sure. I mean we're building out the Systems Design and Analysis portfolio. And Anirudh's focus is always on solving the customers' biggest issues that -- and MSC, which was a company that Hexagon bought and call it their Design and Engineering group, that's what we're pulling out of Hexagon.

We think it's been underinvested, but we do think there's more we can do with it. They have specific solutions there that we think are very, very important to the physical AI space. And just the breadth of what we offer now in System Design and Analysis, we're basically trying to repeat what we did with BETA. BETA was a tremendous acquisition for us because not only did it bring top technology and top talent to Cadence, but it pulled through so much more other business from their customers, and it kind of opened up a whole bunch of doors for us in the automotive space. We think similarly with MSC, it's an opportunity to kind of land and expand.

Their customer base, I think you talked about the -- like just in the press release and last quarter, I think that there's not a lot of overlap? Or it sounds like there's a lot of new incremental potential customers there as well.

There will be, yes. Yes. So I mean the issue for us at Cadence in terms of starting is that -- so with Synopsys buying ANSYS, you're not just buying ANSYS, but you're buying the whole shopping mall, right? That they have the whole infrastructure and the sales infrastructure. With us starting from an EDA company and you're adding tools through innovation and then we're buying smaller companies that we basically have to build our own shopping mall.

So that's where we saw the value of beta, a marketplace where those customers naturally came that was a conduit to selling more of our other offerings. We think they'll be the same with MSC. And we're kind of building out the whole Cadence licensing platform to be able to do that so that we can add more and then kind of have one kind of unified storefront, I guess, under the Cadence brand.

Okay. You kind of touched on that a little bit, but like can you walk us through like the integration process? Like just how do we think about like the model implications? I think the deal is supposed to close like first quarter-ish next year. What does that kind of look like in like time lines and things of what you normally would do like in closing an acquisition of that size?

Yes. I mean -- I guess with all acquisitions, and we're getting better at them, the more we do. But -- so there's there are integration teams throughout Cadence. But essentially, we'd expect something like that to be dilutive in the first year when we bring it in that -- that we pick up these things, normally, we'll identify kind of revenue synergies and cost synergies that -- but it will take us 12 to 15 months to extract those. You saw that with BETA. Now we normally aim for 50% incremental margin. And in 2024, we missed that 50% incremental margin target. Now what I told the team and the Board at the time was it wasn't so much that we failed to hit the 50%. It was just that we ran out of time. But because we're bringing in BETA so late in the year, it's kind of a headwind in the short term. And I told that we'll prove that in -- we prove that we just ran out of time in 2025 because once we extract the synergies, you'll see that incremental margins in '25 will be so much stronger.

And if you look at '24 and '25 together, and actually, if you do look at '24 and '25 together now, you're probably getting incremental margin of about 53%, which is kind of 53%, 54% is what we've been averaging. We aim for 50% plus. We generally get a bit more because we're normally prudent with the guide. But the -- I think it's similar with MSC. The first year -- depending on when it closes, but I mean, I think they were doing about $280 million of revenue stand-alone, but they would do multiyear business. I think on an annualized basis that if you flip them all to annual contracts immediately, it's probably a run rate of about $200 million a year. But -- and we're kind of modeling a cost basis on that. We would think within 12, 15 months, not only will they be in tiptop shape for -- as part of the Cadence portfolio, but they'll probably be adding to our incremental margin by that point.

Okay. Yes, so I mean in terms of like cost, there's probably not a lot of costs you can really take out or -- there's a lot.

There's a lot. Because I mean it's not just -- so we'll probably spend more in R&D than we spend. But we can take out a lot of the infrastructure cost, the G&A costs and...

Okay. So that would come with it, I guess.

Yes, that's right. And there's a lot of synergies we get from the sales and marketing costs when you add in all the other portfolio of products that we have on SD&A.

Okay. Okay. That's helpful. So I mean, I guess like net-net, we think about like -- you're not obviously not guiding, but like you think about like the leverage or incremental margin structure for '26, maybe looks a little bit more like '24?

I think the next 2 years will probably have the same kind of profile as '24 and '25, depending on how early MSC closes. If the earlier closes the more time we have to go fix things up. Also, we're getting better at it. I mean you saw -- one of the things I suppose always feel stupid looking back, it's -- we're always like -- I'm at Cadence now over 28 years, and we operate with continuous improvement in my knowledge.

So you're always doing things trying to do things better and better. And I shudder to think like what -- how bad we must have been in the past. But because when I look at BETA now, one of the silly things that it did was we kind of weighed -- when we were finding synergies, some of those synergies were coming on the Cadence side. We didn't have to wait for beta to close to go get those. So learning from that experience for this one with MSC, we've already taken some actions on the Cadence side. And you saw that already come through in margins in Q3. It will benefit Q4, but it kind of clears the playing field for when MSC does.

Yes. Hit the ground running.

Yes. Yes, exactly. So I think we'll get faster at converting these. But to the level of profitability we would want to get. And the pull-through opportunity, we expect, I think, could be significant as well.

Yes. Okay. That's helpful. You've grown like the SD&A business like organically as well as inorganically. I mean you -- when this acquisition closes, you look at like just kind of the portfolio you have, like are there still holes that you need to fill? Or do you feel like you've got like the right kind of portfolio for where the market is at today?

Well, there's always gaps, right, because the market moves as well, but the gaps are small that -- I mean, if you ask Anirudh and I will spend lots of time dealing with general managers of each of the groups coming asking us for more investments because they want to invest in something organically or plug some gap. But generally, it's stuff that we can address organically with more investment. There's not a lot out there from an M&A perspective, the occasional small tuck-in. MSC is the biggest thing we've probably ever done. But -- but I would imagine that our concentration will be on that in the next year or 2. But anything else we do will be small.

Small, okay. Maybe shift gears a little bit to Design IP. It's been an area of more focus for you guys over the last few years. Can you just kind of talk about like that journey some and just kind of where you've come from and where you're still going?

Yes, sure. I mean we were allergic to IP probably a decade ago that we just thought it wasn't the most rational business that people were offloading their own IP work, I guess. And you could print any revenue growth you wanted if you didn't care about profitability or cash flow. But, yes -- and that wasn't -- that's not our style. We're farmers, not hunters. So we're always thinking of the long term. And so we kind of backed away because it wasn't the right use of capital for us that we're always trying to, I guess, optimize the value of the scarce resource, which we think is engineering talent.

But if you look at the operating income per employee at Cadence and adjusted for share-based comp or because you got to take share-based comp out because some of them get paid in shares. The Cadence and Synopsys, probably 2016, 2017 time frame, we're generating maybe $45,000 to $50,000 per employee in terms of operating income. If you look at where we are now, we're up to about just over [ $140,000 ], and I think Synopsys at [ $70,000 ] and that's been the benefit of applying or allocating that talent to the most productive use, I guess, for the company. But what we've seen over the last number of years, we also have all these data metrics.

So Anirudh and I love the whole moneyball idea. So we were kind of trying to get metrics on everybody and identify who are the best engineers, keep them and that move the other ones on. And like I said, always trying to optimize for the value that we can create and allocate capital to the right areas.

Over the last few years, we found IP has become much more rational that customers are asking us to play there. But of course, though, it doesn't make sense for us to go back and build an IP portfolio for older process nodes that -- so you're building it for -- so the market will gradually come to us more and more because we're building out our IP portfolio to where the market is actually going.

So where do you draw that line of like the 28-nanometer, is it 10?

Well -- I mean they let the customers draw the line for us. I mean things like the ARM Artisan libraries is customers coming to us and saying, okay, look, if you're not going to build from scratch, how about you take this and they're trying to do matchmaking between ourselves and ARM because they think that, well, if you've taken in just a little bit more investment now you can broaden your portfolio. But -- but it's thoughtful -- it's done in a thoughtful -- very thoughtful long-term partnership way with our customers. And our customers get it that I think they knew that they were taking advantage of companies that were willing to do IP business for kind of very low-margin business.

But I mean, at one point, I mean, I freaked out at one point because I remember seeing them -- a contract come across my desk and it was like a bluebird, one of these $20 million bookings. And I'm like, wow, what's the $20 million IP booking that -- and of course, the sales got resolved for it. I called a friend of mine, the CFO of the company that was given us the $20 million booking. I'm like, guys, you're trusting us with this. And he's like, well, it was going to cost us $24 million. If you guys can do it for $20 million. And I thought, well, why do we think we could do it for $20 million when they have years of experience of doing it that -- and it was going to cost them $24 million. And of course, people are dangerous with spreadsheets and particularly if there's commission involved.

But -- so I remember sitting down with Lip-Bu and Anirudh going, this does not make sense. It scares me. I can't sleep at night if we're doing stuff like this. So I'd rather let's redirect the business elsewhere. So that's why we shied away from it. Now I think -- and of course, customers knew what they were doing. But now it's gotten to the point where they really want us to play. And they know that for us to play, they have to be more rational on pricing.

Okay. I mean do you think like a piece of that is just also just the -- we talked about earlier, the increased complexity and just...

Absolutely -- because everything has to -- I mean, it's so complex now, and it's not getting any less complex that -- and everything has to work together seamlessly. So it's the portfolio right across System Design and Analysis, EDA and IP that makes the whole partnership work with our customers, particularly the bigger ones.

I guess, like as you think about that portfolio, right, I mean, you guys have maybe started adding some more things like in foundational IP. Do you feel like that is -- you have a good portfolio there. I mean it sounds like you're kind of going with customers on that journey in terms of progression of nodes and technology. But like how do you feel about like just the foundational library that you have today?

Yes, I think it's very broad now. The very -- it's like I say, mainly customer-driven that we engage with customers for them to help guide us in terms of what the road map is. And I met with our Head of IP yesterday. He's delighted with the way everything is going. They have their own gaps where they want to build something themselves, but he thinks he has the right critical mass now and the talent to be able to do it. I don't -- Matt, is there anything you want to say about it?

Yes. No, I was just going to mention that we're focusing in on those advanced areas, AI, high-performance computing, those 5 areas really being on the design [ PDR, ] PCIe [ PDIe, ] high bandwidth. I mean those are extremely important areas that our customers want us to get better on that.

That's perfect. And like can you remind us, I mean, obviously, your closest peer has had some issues with their IP business in terms of they talked about some maybe more idiosyncratic issues, but also like go-to-market or changing the way that maybe there's some pricing mechanisms and things. Like can you remind us kind of like how you go to market with that and those different pricing mechanisms that are involved in the IP business?

Yes. So you're talking about Synopsys?

Yes.

A great company, right? And we know lots of people over there. We think they're great people as well. They have a slightly different approach. I mean we're very much like I say, farmers, not hunters, that take a long-term partnership view of everything. I get the sense, like just from talking to our customers that Synopsys maybe over the last few years have become more transactional rather than...

Customized.

Yes, it's kind of customer field. It's about bookings or whatever. With us, normally, it's like Anirudh is always focused on what's the problem we're trying to solve and let's do it together and what's the plan. And it tends to be a multiyear approach, and then we're throwing whatever we have in the tool bag at it.

Yes, I think what we're seeing is that, of course, they had a big head start in IP. We've started playing now in IP. It's more -- and customers asking us to play, probably didn't want to be captive to Synopsys or ARM, but I'm wanting more choice. And I think as the businesses move down the process, there'll probably be more competition from us. So that's probably what you're seeing.

Okay. Okay. Can you remind us just last question around Design IP, like the profitability profile of the -- I mean you talked about being very focused on like maintaining margins and they're fitting in your framework. But can you remind us kind of like where does the profitability of that business sit relative to kind of the corporate average?

I mean it's very, very strong. I mean it's getting stronger every year, again, because it's rational. It's tremendous for incremental margins. The importance to us in terms of -- I mean, we're getting a lot of incremental margin because scale, company scales really well. I mean when I look at organically now, we were aiming for over 50% incremental margin. We're probably hitting 60% organically. But -- and then with the M&A that we're doing, because it tends to be dilutive in the first year or so that kind of pulls us back towards the low 50s overall as a combined unit.

But if you look at Cadence. We normally -- we budget for double-digit revenue growth, and we typically achieve low teens. But when you budget -- and I guess I was lucky, I was talking to the Board yesterday about like we look over the last 3 to 5 years, we've probably been averaging around 14% revenue growth. Maybe 2% of that is coming inorganically, 12% organically.

But as we get these new acquisitions in and we clean them up essentially and get them into our kind of workflows, they adopt the profitability profile of the organic businesses that -- and it gives us a lot of bandwidth to go and kind of buy the small tuck-ins and continue to invest in R&D. I would say -- I mean, Anirudh will tell you that we're innovators and innovators lead, imitators follow, that it's really important. We continue to invest very heavily in R&D and help our customers go where they want -- where they need to go.

Okay. That's perfect. We just got a little bit of time left. A quick kind of have to ask about China, right? I think it's been a wild year for business in China. But maybe talk about just kind of the demand trajectory that you're seeing. Have things been back to normal? Are you seeing the customers kind of still being a little bit apprehensive in terms of potential further restrictions or restrictions going back in place?

I suppose the -- I could have used China for the surprise question because it's -- actually, we're surprised with the resilience of China customers. But it does feel like back to normal. There is a strong desire when those temporary restrictions were lifted in early July, there was a strong desire from all of our China customer base that, look, if we have hardware backlog, can we have it now?

Yes, sure.

And so we prioritized all of China hardware backlog. Like when I look at hardware backlog at the end of Q3 and probably end of Q4 now, there'll be much less of a percentage of China in it than there normally is because we're kind of prioritizing that at the moment. I would think -- if you step back and you look at the way the business is going in China, it will probably be a smaller and smaller percentage of our overall business. But I think because they don't have access to the...

Leading technology.

They don't have -- we don't see the same AI pull-through there as we're seeing in other regions. So probably slightly less than average growth is what I'd expect over the next 3 to 5 years. But it's still -- I mean, it's still a great region for us and a source of strong kind of revenue growth and cash flow for us.

Okay. Perfect. I think we're out of time. Thanks for joining.

Thanks for having us. Cheers. Thanks.

Ladies and gentlemen, good afternoon. My name is Abby, and I'll be your conference operator today. At this time, I would like to welcome everyone to the Cadence Third Quarter 2025 Earnings Conference Call.

[Operator Instructions]

Thank you. And I will now turn the call over to Richard Gu, Vice President of Investor Relations for Cadence. Please go ahead.

Thank you, operator. I would like to welcome everyone to our third quarter of 2025 earnings conference call. I'm joined today by Anirudh Devgan, President and Chief Executive Officer; and John Wall, Senior Vice President and Chief Financial Officer.

The webcast of this call and a copy of today's prepared remarks will be available on our website, cadence.com. Today's discussion will contain forward-looking statements, including our outlook on future business and operating results due to risks and uncertainties. Actual results may differ materially from those projected or implied in today's discussion. For information on factors that could cause actual results to differ, please refer to our SEC filings including our most recent Forms 10-K and 10-Q, CFO commentary and today's earnings release. All forward-looking statements during this call are based on estimates and information available to us as of today, and we disclaim any obligation to update them. In addition, all financial measures discussed on this call are non-GAAP unless otherwise specified. The non-GAAP measures should not be considered in isolation from or as a substitute for GAAP results. Reconciliations of GAAP to non-GAAP measures are included in today's earnings release.

[Operator Instructions]

Now I'll turn the call over to Anirudh.

Thank you, Richard. Good afternoon, everyone, and thank you for joining us today. Cadence delivered excellent results for the third quarter of 2025. With strong operational and financial performance across all product categories and geographies as we continue the disciplined execution of our strategy. Bookings exceeded our expectations with backlog growing to over $7 billion, underscoring our continued technology leadership and reaffirming Cadence as a trusted partner, enabling customer success.

Given the ongoing strength of our business, we are raising our full year outlook to approximately 14% revenue growth and 18% EPS growth. John will provide more details on our financials shortly. The accelerating AI megatrend is fueling an unprecedented wave of design activity across industries ranging from hyperscaler infrastructure to fast-growing physical AIR of autonomous driving, drones and robotics to the emerging domain of sciences AI. As AI drives exponential design complexity and new system architectures, cadence is uniquely positioned to capture this generational opportunity with our differentiated and comprehensive portfolio spanning EDA, IP, 3D-IC, PCB and system analysis. The Cadence.ai portfolio embodies our strategy of design for AI and AI for design, empowering customers to build out the global AI infrastructure, while we infuse AI into our own products, to deliver breakthrough automation and productivity.

With deep partnerships across AI innovators, foundries and system leaders and a comprehensive chip to systems portfolio, Cadence is driving transformative PPA and productivity gains, positioning us well for sustained growth in the AI era. In Q3, we meaningfully expanded our partnership with Samsung through a wide-ranging proliferation of our core EDA software as well our system software across PCB, advanced packaging and system analysis. We also deepened our long standard partnership with a leading semiconductor company in Q3. Through a board proliferation of our core EDA, IP and systems portfolio and are closely collaborating on next-generation agentic AI EDA solutions.

We expanded our long-standing partnership with TSMC to power next-gen AI flows supporting TSMC's N2 and A16 technologies. Our Integrity 3D-IC solution provides comprehensive support for the latest TSMC 3D fabric die-stacking configurations and our design in ready IP including HBM 4 and LPDDR 6 on N3P enabled next-generation AI infrastructure. At TSMC's OIP conference, Broadcom, highlighted Integrity 3D-IC full flow deployment success for hyperscaler high-capacity ASICs. Our IP business maintained strong momentum in Q3 driven by global accelerating IP demand and increasing customer proliferation, of our expanding IP portfolio.

Our profitable, scalable IP strategy focused on AI, HPC and automotive verticals positions us well for continued growth. Increasing complexity of interconnect protocols driven by AI and chiplet architectures along with new foundry opportunities are providing strong tailwinds to our IP business. Bookings were strong and tracked ahead of our expectations. Our design IP portfolio secured several competitive wins at top AI and memory customers.

For instance, we won a highly competitive engagement at a marquee memory company that embraced our HBM 4 and DDR5 IP for its new AI design. The recently completed acquisition of the ARM Artisan Foundation IP further augments our design IP portfolio with standard cell libraries, memory compilers, and IOs optimized for advanced node at the leading foundries. Our Tensilica audio and vision DSPs and Neo-AI accelerator, NPUs scored multiple design wins with leading customers in U.S. and Asia for mobile, automotive and data center verticals. Our core EDA business delivered strong results, driven by growing adoption of our AI-driven design and verification solutions.

In digital, Cadence Cerebrus AI Studio, the industry's first agentic-AI, multi-blog, multi-user design platform continues to deliver unparalleled PPA and productivity benefits. Samsung U.S. taped out a SF 2 design using Cadence Cerebrus AI studio to achieve a 4x productivity improvement. In another instance, Samsung used Cadence Certus, Tempus and Innovus to rapidly close and sign off a multibillion instance AI design on SF IV with 22% power reduction and first-pass silicon success. Our Virtuoso studio and Spector platforms saw strong momentum. With their AI-driven features and workflows, gaining rapid traction as the customers leverage the automated design migration and optimization capabilities. Our hardware verification platforms have become the de facto choice for AI designs offering industry-leading performance, capacity and scalability.

Hardware had a record Q3 with several significant expansions especially at AI and HPC customers. We deepened our overall collaboration with OpenAI as they expanded their commitment to our Palladium emulation platform in Q3. Where ACM [indiscernible] AI saw growing adoption as it delivered dramatic debulk productivity, test bench efficiency and accelerated coverage closure. NVIDIA, Samsung and Qualcomm all presented [indiscernible] AI success stories at Cadence Live India, highlighting 5x to 10x improvement in verification throughput.

Our system design and analysis business achieved another solid quarter, driven by expanding set of innovative solutions and growing adoption across a broadening customer base. In Q3, we significantly expanded our cadence reality, digital twin platform library, with NVIDIA DGX Superpod model and DGX GB-200 systems to accelerate AI data center deployment and operations. Three major memory providers significantly increased their clarity and security usage as they transition to a full Cadence flow for advanced IC packaging, displacing competitive solutions. Beta CAE continued its momentum with multiple competitive displacement, underscoring its accuracy and performance advantages including a significant competitive win at a large Tier 1 automotive company in China.

In Q3, Infineon Technologies standardize its PCB design workflow on the Cadence AI-driven Allegro X platform for their future designs. Last month, we signed a definite agreement to acquire Hexagon's T&E business, including its MSC software business to bring industry-leading structural analysis and multi-body dynamics technologies to Cadence. Complementing our multiphysics portfolio, this will accelerate our expansion in SDA and put us at the forefront in unlocking new opportunities across automotive, aerospace, industrial and the rapidly emerging world of physical AI.

In summary, I'm pleased with our Q2 results and the strong momentum across our businesses. The AI era offers massive market opportunities and through the co-optimization of our entire portfolio with AI and accelerated computing, Cadence is uniquely positioned to be the trusted partner to deliver AI-centric transformational solutions across multiple industries. Now I will turn it over to John to provide more details on the Q2 results and our updated 2025 outlook.

Thanks, Anirudh, and good afternoon, everyone. I'm pleased to report that Cadence delivered strong results for the third quarter of 2025 with broad-based momentum across all our businesses. We exceeded our guidance for Q3 revenue, operating margin and EPS and are raising the full year outlook across these key metrics. With the updated outlook and at the midpoint, we now expect our 2025 revenue to grow approximately 14% year-over-year, on track to achieve double-digit growth across all our product categories for the year. Third quarter bookings were strong, resulting in a backlog of $7 billion.

Here are some of the financial highlights from the third quarter, starting with the P&L. Total revenue was $1.339 billion. GAAP operating margin was 31.8% and non-GAAP operating margin was 47.6%, and GAAP EPS was $1.05, with non-GAAP EPS, $1.93.

Next, turning to the balance sheet and cash flow. Cash balance at quarter end was $2.753 billion, while the principal value of debt outstanding was [ $2.500 billion ]. Operating cash flow was $311 million. DSOs were 55 days, and we used $200 million to repurchase Cadence shares.

Before I provide our updated outlook, I'd like to highlight that it contains the usual assumption that export control regulations that exist today remain substantially similar for the remainder of the year. With that in mind, for Q4, we now expect revenue in the range of $1.405 billion to $1.435 billion. GAAP operating margin in the range of 32.5% to 33.5%. Non-GAAP operating margin in the range of 44.5% to 45.5%. GAAP EPS in the range of $1.17 to $1.23 and non-GAAP EPS in the range of $1.88 to $1.94.

As a result, our updated outlook for 2025 is revenue in the range of $5.262 billion and $5.292 billion. GAAP operating margin in the range of 27.9% to 28.9%. Non-GAAP operating margin in the range of 43.9% to 44.9%. GAAP EPS in the range of $3.80 to $3.86. Non-GAAP EPS in the range of $7.02 to $7.08. Operating cash flow in the range of $1.65 billion to $1.75 billion, and we expect to use at least 50% of our annual free cash flow to repurchase Cadence shares.

As usual, we published a CFO commentary document on our Investor Relations website, which includes our outlook for additional items as well as further analysis and GAAP to non-GAAP reconciliations. In conclusion, I'm pleased with our Q3 results. Following 2025 as we continue to deepen strategic partnerships across the ecosystem. As always, I'd like to close by thanking our customers, partners and our employees for their continued support. And with that, operator, we will now take questions.

[Operator Instructions]

And our first question comes from the line of Vivek Arya with Bank of America Securities.

Your IP business is now, I think, tracking to over 20% growth for the second year. Anirudh, I was just hoping you would give us some sense for what's driving this growth? Because your competitor expressed a lot of concerns about their IP business, whether it is in China or [indiscernible] or just IT visibility in general, and I think they were talking about a new business model. So how do we square that and the growth you are seeing? How sustainable is this growth? And what is your visibility in your IP business?

Yes. Thanks, Vivek, for the question. I'm actually quite pleased with the performance of our IP business. And we don't look at any 1 quarter, but even if you look how we performed last year, of course, this quarter was exceptional. But overall, how we performed this year and what we see backlog and activity going into next year, overall IP business is performing quite well and there are multiple reasons for it. First, our IP business is different. I think it's much more profitable even though the profitability is less than our EDA business, but I think it's more profitable than general IP business because we also have Tensilica, which is almost like software like profitability. But a lot of the growth is coming in design IP and the reason for that is our IP business is focused on AI and HPC at the most advanced nodes.

Since we got started later in the IP business, we focused it -- where the future is going, which is AI, HPC and chiplet-based architecture. So a lot of the -- like SerDes and PCIe and HBM 4 IPs. And that part of the market is doing well actually across the world.

And then the second reason is, as you know, there is more and more foundries entering especially at advanced nodes. And we have a long-standing partnership with TSMC, but also Samsung, Intel and now Rapidus. So there are at least 4 major foundries now at leading nodes. So that's, I think, a second reason for our IP business to be well positioned. And as the performance of our IP business has improved, the PPA -- our PPA is competitively better in design IP and a lot of customers want to shift over to cadence. So the customer demand, I think, is the third reason as our IP business strengthened that we are seeing strength in the IP business. So I think for these 3 main reasons, I'm pretty optimistic about the IP business.

And going to next year, we're not getting to next year, but just to give you indication, I would be surprised if our IP business does not grow better than Cadence average, which it should, given the profitability profile. We want that to happen. If the profitability is slightly lower than EDA, then the growth should be higher than Cadence average. So overall, I think that would make like 3 years trend. And overall, I'm pleased by our IP performance.

And our next question comes from the line of Jason Celino with KeyBanc Capital Markets.

Great. Last quarter, I think you mentioned the second half having good renewal opportunity with some of your large customers. With the uptick in backlog, I imagine some of that strength was from some of these renewals. But as we think about Q4, do you still have renewals on the docket?

Yes. Thanks for the question. I'll let John comment on the timing of the renewals. But overall, I do think that our performance in Q3 is much -- is better than we expected. And the primary reason and this is true in all geographies. But I think the primary reason is that the AI infrastructure build-out, as you know, is accelerating, okay? And we are essential to the design and build out of the AI infrastructure. Of course, we -- I have said publicly, there are 3 big phases of AI in my mind. AI infrastructure being the first one, physical AI being the second one and Sciences AI there being the third one. But most of our focus on investment is, of course, on the first one. And as you see in the last 6 months, it is accelerating. And also the -- we are privileged to work with all the MAG-7s and also investment in internal chip design is accelerating along with, of course, the big merchant silicon companies like NVIDIA and Broadcom and AMD. So I think that is coming through in our booking activity in Q3. And so far, we see that strong demand continuing in the future.

Yes, Jason, I would just like to add that the mix as well as healthy across EDA, IP hardware and SDA. And the core EDA and IP backlog is weighted towards multiyear recurring arrangements, and that supports durable double-digit growth.

And our next question comes from the line of Joe Vruwink with Baird.

Great. I guess I'm struck by the number of times the word acceleration has already been used on the call so far. And I guess the third quarter bookings much stronger than we were expecting, and it would support a future acceleration. I know it's atypical to kind of get 2026 comments, but Anirudh already defer the IT business. I'm just wondering if you can maybe start to frame expectations for next year based on what you have in hand and it certainly seems like things are setting up well. Do you have the type of visibility at this point. So maybe comment on this.

Yes. I think what I would like to say is that we always look at our business in terms of how well our products are doing, okay, and we report like 5 lines of businesses, as you know. And I would say at this point, all 5 lines of business are performing very well. And you can see that in this year, I think we will grow double digits in all 5 lines of business. And also, we are performing well in all geographies. So in terms of products and geographies, which is our main focus. Are we aligned with the leading companies? Are we trusted partner of the market-shaping companies. So if you look at products, geographies and customer alignment, I think we are well positioned.

Of course, as you know, as we enter a new year, we are always prudent in our outlook, and we will give you an update about next year when we come to January, February time frame. But I think Cadence is very well positioned -- in a better position than it has been, I think, compared to last several years, and we look forward to working with our customers in the future.

Yes, Joe, we won't guide FY '26 today. But exiting FY '25 with probably record backlog and broad-based momentum from deepening strategic and trusted partnerships across the ecosystem positions us well for next year. You can expect our framework will remain disciplined. We typically aim for double-digit top line ambition, continued operating leverage and balanced capital allocation. And that's all underpinned by secular AI demand across chip to systems.

And our next question comes from the line of Lee Simpson with Morgan Stanley.

Great. Congratulations on another great quarter. I just wanted to ask around about China, really. The -- it looks as though you're up about 53% year-on-year doing well in the mix up to 18%. That feels more than just a sort of return of business post the restrictions on the BIS letter last quarter, it feels though there's genuine momentum there. So I wonder if you can talk me through what is driving this? Is it IP? Is it hardware? Is it [indiscernible] EDA? What are the vectors here.

Thanks for the question, Lee. Yes, I mean, we saw broad-based strength and China design activity remains very strong. The region returned to business as usual for us in the second half that with the lifting of the export regulations that changed for EDA in early July, but Q3 really was only slightly better than we expected, and we now expect China to be up year-over-year for fiscal '25. Anirudh, do you want to add anything to what's happening in China?

Yes, Lee, that's a good question on China. I mean, overall, I would say the behavior in China from what I can tell is back to normal. Of course, there was a disruption in Q2 for obvious reasons, given the policy in Q2, but the behavior that we are seeing is back to normal in Q3. And a lot of it was driven by like us prioritizing hardware deliveries that we could not do in Q2 into Q3. But overall design activity is strong in China across -- semiconductors are essentials to every country in China continues to invest in semis. But overall, I would say the -- our strength is broad-based, not particularly tied to any 1 geography and there was some makeup from Q2 to Q3. Now it's difficult to predict the future, but what I see, I don't see any unusual activity in China, like question maybe like is there any pull-in from future quarters. We don't see that in terms of what we see. And we see overall broad-based trend in other geographies as well.

And our next question comes from the line of Siti Panigrahi with Mizuho.

Congratulations on the strong execution. Anirudh, I want to ask you about on your system design, mainly that simulation analysis market. Help us understand your strategy. You made acquisition last year, BETA CAE and this year again you've announced MSC software. Help us understand how you're going to position yourself against your competitor in that market? This is definitely a growing market. I would appreciate any color on that.

Yes, Siti, thanks for that question. I mean I'm pretty pleased with the overall performance of SD&A. And I mean just to remind everybody, Cadence is the one started this whole thing in 2017, 2018. Now it is considered obvious that silicon and systems are going to come together. I mean we have been talking about this for a very long time. Now I think what the acquisition that we did this quarter is more forward-looking in the sense that, like I mentioned, these 3 horizon technologies, horizon One being infrastructure, horizon 2 being physical AI, horizon 3 being sciences AI. And that's how we are focused.

Most of our investments in Horizon 1, but of course, like maybe 70%, 80% is Horizon 1, about 20% Horizon 2 and few percent horizon 3, but horizon 2 of cars, drones and robots can be a very, very big market in the future. And what happens is AI is going to change also for Horizon 2.

As you see, there's a lot of reports that the word is going to move from LLM based AI to a word model-based AI, in which robots you have to -- it's no longer the text data that trains the robot. It is the physical movement and all that. And one of the key challenges in training robots or cars is that there is not enough data that is available. When you train an LLM model, basically, the data is available on the Internet and as well -- language data is available.

Whereas training a robot, the data is not available, okay? So the data either has to be generated manually, like they put sensors on a human and the person picks up the object, that could be data. But that's a very slow form of getting data. The best way to generate data for a word model is through simulation. And this is what we have talked about also for a very long time of the 3-layer cake.

So then the fundamental simulation of multibody dynamics becomes essential in horizon 2 physical AI and Hexagon had a leading simulator for multibody dynamics, along with structured simulation, which helps in all kinds of electronics and automotive. So I think I'm pretty optimistic that this can position us well for the second horizon, which is physical AI. And so what that will do for our SD&A business, the way I look at it, our SDA business once we complete this acquisition, we'll have 2 strong pillars. And it will actually -- the run rate should cross $1 billion in 2026 if the acquisition closes and one pillar will be driven by 3D-IC and chiplets. Allegro is in our SD&A business. Allegro is the de facto standard for package design in the world. And so if you take a Allegro, combined Sigrity and clarity and Celsius, our kind of electromagnetics and electrothermal tools. That's one key area of this merger of silicon and system. And we will be very, very strong in that -- in our partnership with TSMC, our partnership with all the leading AI players like NVIDIA positions us very well with Allegro and 3D-IC. So that will be roughly 1/2 of our SD&A business because there's going to be a lot of growth in this chiplet-based architecture.

And the second part will be this physical AI structural analysis and the combination of beta, which was the leader in pre-post processing with Hexagon which has a lot of solvers like multi-body dynamics, structural. And then we acquired a great new CFD solver from Stanford a couple of years ago. So if you put all those solvers together with beta, that will be roughly half of our SD&A business and really well positioned for the physical area.

So if you put it all together, the benefit of Hexagon is that it will give us 2 strong pillars in SD&A in the areas that are going to grow the most in the future. One is 3D-IC and HPC, the other is physical AI and connected technologies.

Our next question comes from the line of Jim Schneider with Goldman Sachs.

I was wondering if you could maybe frame for us some of the tailwinds you expect you might see over the next couple of years as a result of inclusion of AI features into your products on the core EDA side. Maybe talk about any kind of productivity metrics you can give us in terms of time to market or developer productivity and how that might translate into either revenue or adoption rates of that technology and features.

Absolutely. Great question. As we have said before, there are 2 parts to our AI strategy, which is we call design for AI and then AI for design. Okay. I think the first part is the build-out of the AI ecosystem, whether it's infrastructure or physical AI. And that we are very well positioned with all the leading players, of the MAX 7 companies -- and now I think your question is on the second one, which is, of course, applying AI to design. So even this time, we highlighted several examples. So we have at least 5 major platforms and some of the big examples are, for example, [ SME AI ] , which is using AI to accelerate verification. Verification is almost an exponential task in chip design. And we are seeing with SMA, 5 to 10x improvement in logic simulation efficiency and coverage, which is one of the mostly heavily used tools in verification. And even in cadence like Samsung and Qualcomm and NVIDIA highlighted this. So these are demonstrated benefits at customer sites being highlighted by the customer themselves.

The other area is in physical design the back-end physical design with Cerebrus AI studio. Again, we had Samsung Cox improvement in productivity and also 22% improvement in PPA. By the way, this is huge numbers because when you go from like 5- to 3-nanometer 3-nanometer to 2-nanometer, typically, a node migration, which the industry is spending like billions and billions of dollars will give like 10% to 20% PPA improvement. And if we can get that with better optimization with better AI, that's a huge value for our customers.

So the good news is that I think the adoption of AI tools is almost taken as a de facto. All the big customers are adopting our AI tools. And I said even before that the monetization of that takes some time. It always takes 2 contract cycles. And I think we should be able to do that or slightly better. So -- but the productivity is huge by applying AI to EDA. And the reason I think it is different in EDA than other things is, first of all, there are multiple reasons. One is we have done automation for 30 years. The chip design process is highly automated. About 80%, 90% of it is already automated.

So we have a lot of history of automation and then AI is the next 10x that automation that can happen. I mean we have probably improved chip design 100 in the last 20 years. And AI can give the next 10x. And the other thing that is different in chip design versus other industries, I believe, is because the workload is exponential. The chips in 5 years from now will be like 5x, 10x bigger the complexity will be 20, 30x more given software and chiplet. So AI productivity is needed just to keep up. So our workload is exponential is very different than a workload is not exponential. So the customers are expecting us to deliver more productivity and are accepting of deploying that in their designs.

And our next question comes from the line of Harlan L. Sur with JPMorgan.

Great job on the quarterly execution, as always. On the third generation upgrade cycle on your emulation and prototyping platforms, you're about 5 quarters into the upgrade cycle to record revenues in Q3. If I rewind back to your second-generation launch, right, the team drove 3 years of record revenues post launch. You still have the same drivers in place, right, design software complexity increasing exponentially, the cadence of new chip program introductions accelerating addition of new customers like Open AI, as you mentioned on the call today and proliferation of all of these challenges into new markets like automotive and software-defined vehicle. Given the lead times for your proteom and palladium systems. I assume you're already booking into next year. What's the demand curve look like? And do you anticipate continued momentum in growth in 2026 for the hardware platform?

Yes, Harlan, as always, you're always very perceptive in the overall trends in the market. Yes, hardware is doing phenomenally well and I expect the trend to continue. So will 26 be better than 25%. That's what we would think. Now how much better? We are always prudent in that because hardware, you don't have like a full year visibility like we would have in the software business. So when we go into any given year, we only have a 6-month visibility.

So we are always prudent in our hardware guide. And then if the business comes in as expected, just like this year, we can improve our guide for the rest of the year. But that's on the -- that's more on the guiding discipline, which we want to be -- we want to derisk our guide for our investors. Now in terms of fundamental technology trends and market trends, I mean this is this is a great setup for hardware because first of all, we are the only company that builds our own systems.

We build our own chips at TSMC there are full radical chips. You should see these things. But these rags have 144 liquid cool ships connected by InfiniBand and Optical and the customers will connect like 16 racks together that can emulate like 1 trillion transistor designs. I mean, there is no other platform that can compete with that. And also, the demand for hardware is increasing not just because of their more AI designs. But as we go from 3-nanometer to 2-nanometer to 1.4 to 1, which will take next 7, 10 years, the size of the chips only increases. And so there is more and more demand for hardware. So overall, competitively and market trend wise, I think we are well positioned in hardware. But of course, for any given year, we are prudent in the guide. John, I don't know if you want to add?

Yes, yes, yes, Harlan, what I'd add there is demand remains very strong, particularly across AI, HPC and auto markets. we've been scaling manufacturing capacity and trying to improve lead times. We've also had hardware gross margins become more healthy. We remain focused on throughput to meet the elevated need from AI designs. And if you look at our financials this quarter, you'll see that we've been building inventory to try and meet the demand in the -- that's reflected in the pipeline for the next 6 months.

And our next question comes from the line of Jay Vleeschhouwer with Griffin Securities.

I know you gave several examples of customer activity, customer engagements and so forth. And I would like to ask you about the recent announcement of the joint work that NVIDIA and Intel are going to be doing. Would it be fair to presume that combined GPU and CPU work would necessarily lift up demand and capacity requirements for multiple types of EDA tools. Also IP, probably hardware as well. So there would be a general uplift as a result of that combined work, but at the same time, would it also necessitate your increasing your investments, for example, in AEs as you did when you had that breakthrough with Intel several years ago.

Jay, that's a good observation in terms of CPU, GPU together. By the way, I've said this for almost 15, 20 years that the CPU GPU need to work together because EDI is a very well-optimized workload. And it is computational software, mathematical software, which is very similar to AI. And what happened in the history of EDA is that -- of course, there are a lot of SMD tasks like which can be done in a GPU kind of machine, but there are also a lot of conditional tasks, which need to be done on a CPU kind of machine. So we always wanted both CPU and GPU and we also wanted CPU and GPU to be close to each other. And actually, to NVIDIA's credit and Jensen's credit of Grace Hopper and then Grace Blackwell.

I mean, they are 1 of the first people to track to kind of wash this trend. And now if you look at all the major designs from other companies, too, there is a combination of CPU and GPU together. And that's the reason for the last several years, we are already working on porting our workload to CPU plus GPU.

And a perfect example was when we announced Millennium earlier in the year. So we are moving not just system analysis workloads, which are more GPU friendly, but also EDA workload, which are critical for axillary EDA and 3D-IC to CPU, GPU combination. So what I would like to say is I'm actually very pleased to see that the whole industry now is going towards this combination of CPU plus GPU whether you look at Apple chips or AMD chips and of course, NVIDIA, amazing platform. And this partnership with NVIDIA and Intel is good for us in terms of it gives us a new kind of x86 plus GPU and also, we have a long-standing partnership with NVIDIA. And then as Intel does more work with NVIDIA is also good for our overall discussions with Intel, which I think are proceeding well. And I think Intel has to invest both its ecosystem for foundry and also its own products. And I think [indiscernible] knows that, and it's good to see the investment on both sides.

Just to be clear, aside from the porting that you have to do internally for your own tools, you are presuming that of demand that this customer activity would necessarily increase the consumption of EDA.

The customer activity should -- I mean, I think first of all, if the EDA tools get better because of CPU GPU system being optimized. Typically, the customers will adopt. We are always looking at ways to improve our our tools. And this gives another vehicle to improve the performance of our tools. So that's good for all customers. And then I think in this particular partnership, there are specific design activity that needs to be done without getting into too much detail, NV-based IP and -- so yes, we are working with the particular companies on designed to make this design happen just like we would work with any of the leading designs. So yes, there is a specific customer activity connected to NVIDIA and Intel. And in general, there is customer benefit if our tools are optimized better on this platform.

And our next question comes from the line of Gianmarco Conti with Deutsche Bank.

Congrats on another great quarter. Maybe just going back towards China, especially given the amazing quarter you guys have had, of course, part of it was recouped from Q2. But how should we think about a sustainable growth rate in the region beyond what was [ released ] last quarter. And potentially, if you could give some color on if there's any real risk from yet another ban in the region. Obviously, there was some news flow going on, and I think investors want to be a bit wary about like what was real in terms of potential risk to EDA or what is sort of like a broader macro level impact? Any commentary, that would be great.

Yes. I think China, like I said, the design activity seems back to normal to me. And I think we mentioned -- of course, when we started the year, we were very prudent because I said before, when I went to China last year, I mean, they were expecting very tough kind of macro environment, geopolitical environment, which turned out to be true in '25. So we were very prudent in our guide of China in the beginning of the year, which turned out to be correct. Now I think at this point, like John also mentioned last time and this time, we expect China to grow. How much it grows will depend, we'll have a better idea. It's very difficult to predict. We'll have better idea at end of the year. But I do expect China to grow this year.

And then it's good to see -- I mean it's very difficult to predict the geopolitical environment, and I definitely don't want to do that. But it's good to see that there is a lot of discussions between the kind of presidents and through big economies. So any stability there and certainty is good for our business. So we look forward to that. But I do expect that design activity is strong and if there is no unforeseen development and the environment is stable, it should help our business. And I just want to remind you that our strength in Q3 is helped by performance in China, but it's very broad-based, given like all the reasons you mentioned the build-out of the AI infrastructure, the emerging design of physical AI, the overall AI megatrend. So we are pleased -- so we are not indexed to any particular country. But it's good to see that the environment is improving in China.

Yes. And Gian -- I'd like to remind you that our Q4 and full year outlook assumes today's export regime remains substantially similar. And we always incorporate prudence for regulatory variability and we'll continue to comply rigorously with -- while supporting customers globally. And as Anirudh says, we're seeing strength right across all businesses and across all geographies.

And our next question comes from the line of Joe Quatrochi with Wells Fargo.

I was wondering if you could just maybe help us understand like the OpEx dynamics. I think 3Q was a bit better than expected, but 4Q is a bit worse than expected. Is that related to just the Artisan deal timing of closing that? Or just any sort of help there would be helpful.

Sure. Yes. But -- yes, I mean it's really just the timing of some hardware delivery shifting between Q3 and Q4. But overall, the year is slightly ahead of what we were expecting, and we're pleased by the broad-based execution and strong demand across all product categories. Core EDA software is performing very well. Hardware continues to be strong. We're continuing to make progress in SDA and we've continued IP momentum and healthy renewals set up for Q4.

[indiscernible] the OpEx?

Sorry, can you repeat the question.

The question was on the OpEx side, like the OpEx timing?

Yes. So on the OpEx side, we did a small restructure that benefited Q3. The hardware gross margins were very healthy in Q3. And then it's offset a little in Q4 by some new expenses we're picking up from new acquisitions.

And our next question comes from the line of Charles Shi with Needham.

Anirudh, congrats on the nice results and John, similarly here. The question, I look at the legal growth rate for the overall company for the last 3 years, it has been maintaining around that 40%-ish plus/minus range. Truly remarkable. Feels like you didn't really step up a bit at all. But when I look under the hood, the lots of moving parts, right, like let's compare last year versus this year.

Last year, China was bad. Hardware was kind of decelerating, I think that was largely due to hardware transition into the [indiscernible]. I mean, I'm looking at the upfront revenue as to inform you about your hardware growth. But this year, both things have kind of turned out much more network positive, like your upfront revenue is probably going to grow somewhere closer to 50%. China looks like at least it's going to grow above the corporate average. So wonder when we look at -- think about next year, do you think both halfway and China can maintain the current momentum, maybe especially on software based on the observation of the V2X2 cycle, I believe that was somewhere in between '21 and '24. When you go into like a third tier-ish, the growth rate -- in the V2X2 cycle, it kind of decelerated a little bit. So my question is, is this time can be a little bit different in terms of the hardware growth rate going forward? And could any fear of your -- from your customers regarding hardware transition to, let's say, V4X4 in the maybe the next 1 to 2 years, causing some of the deceleration of compare revenue? I know this is a long question, but I think that this is the most important one when we think about the Cadence outperformance going into next year.

Thanks for the question, Charles. We're trying to unpack it. So I think -- I wouldn't focus too much on any 1 quarter or even any 1 half in terms of results. If you recall, last year, the shape of the revenue curve was kind of back-end loaded. Q3-over-Q3 comps can be a bit skewed, particularly as well with China, given that we had that temporary restriction in China from me to the early July. But generally, when you're talking about hardware, demand is very, very strong. But -- and we're seeing a secular trend in hardware demand for many years now because the growth in complexity continues unabated that we're seeing a very strong pipeline for the next 6 months, and we're ramping up on inventory for some large orders that we have to fill in the next couple of quarters. But -- so we're seeing lots of momentum, and we expect to -- I mean typically -- if I go back, I think the last 5, 6 years, and is typical of Cadence, Q4 bookings would exceed Q4 revenue.

So we just finished with $7 billion of backlog at the end of Q3, which is a new record for us. Given renewal timing in Q4 and the visibility we have, we'd expect to end '25 at a fresh high. And with that mix being so healthy across all of the different businesses, I think it bodes well for next year.

So maybe a quick follow-up, so Anirudh, from your perspective, the current hardware V3X3 enough to support 1 trillion transistors, but with the AI really like moving really fast, do you foresee like when you have -- when you probably need to like do another halfway refresh? And is there any light you can shed on this?

Charles, yes, I am very confident in the hardware position. We talked about palladium. We're the only company that designs our own chips and also protium with FPGA systems, and that's also doing well with a dynamic deal. And like John said, we do see good demand. Now I just want to remind you that when we guide we always are prudent given hardware is not as predictable as software. But it is almost -- even though we reported kind of upfront revenue, but what has happened is that all these big customers are almost buying every year. It's not that they're buying -- so the buying behavior is different than 4, 5 years ago because they're doing so much design at all the really big customers, it has almost become like annual kind of subscription, even though financially, it is reported, of course, at upfront.

So now will the hardware trend continue? I mean, right now, I don't see any reason that it won't and so I think '26 will be stronger than '25. How much stronger, we will have a better idea. Now in terms of our next generation, we are always investing in R&D. We have a huge investment in R&D, as you know, 35% of our revenue is invested in R&D, but if you look at the expense side, almost 65% of our expense is invested in R&D and about 25% is invested in application engineering.

So more than 90% of our investment and headcount is in engineering, customer support and R&D. And that's true for hardware. So we are -- we don't want to get into all the details, but you can assume we are well on our way designing the next generation of hardware systems and they will come in time. One thing -- good thing is about our current systems already support 1 trillion transistor design, and that is supposed to happen in 2030, but before 2030, we will have a next generation of hardware, which will support it for the next 5 years. So I think I'm pretty confident in our hardware road map, and the demand itself, I think because Harlem, you know all this area well, I mean, AI, the chips are only getting bigger. And also, what's happening is like even with like Blackwell, it's not just 1 chip now. We have multiple chips and then grace together. So the customers are also not emulating just 1 chip, which is growing 2x every node, they're emulating systems of chips like Grace and Blackwell together or if you have chiplet with architectures. So the demand for hardware may move faster than just more or technology scaling because of this 3D-IC, but again, we will see that we are well positioned. We'll see how it progresses. But systemically, there is no issue in demand for hardware and our competitive position.

Charles, there was a lot in your question, I think you referred to upfront recurring revenue as well. I mean we continue to frame '25 around 80-20 recurring to upfront on a rolling 4-quarter basis. And I think as you mentioned in your question, the variability quarter-to-quarter is driven mainly by strong upfront businesses like hardware and IP and the timing of China ratable revenue earlier in the year that with core EDA growing so well, we're comfortable that 80-20 is probably the right kind of mix of business for the forseeable future.

And our next question comes from the line of Gary Mobley with Loop Capital.

Thanks so much for squeezing me in. And let me extend my congratulations. I really just had a clarification or a question to get to a clarification. So if I recall correctly, given the timing of the export control repeal, which I believe is July your China backlog was not in your June quarter ending backlog. But I presume now that it is. So given that $600 million revenue or $600 million delta in your backlog, how much of that was a function of the inclusion of China backlog versus the prior quarter.

Let me take a crack at it and then I think right, our backlog grew from $6.4 billion to $7 billion. So there's a growth of $600 million. So I think about -- I would say about 25% of that, about $150 million is catch-up from Q2 to Q3. And the rest growth growth strength across our business.

That's exactly right.

And our next question comes from the line of Clarke Jeffries with Piper Sandler.

Anirudh, I appreciate the comments on the mechanics of the strength in the IP business and specifically, the demand for design IP you're seeing for AI projects. I wanted to follow up with just how the wallet opportunity is changing with those AI projects. Specifically, do you see any potential for growing pains or lower profitability to serve the industry as they make more customer bespoke technologies with chiplet or custom memory designs incorporated into those AI and HPC designed has cadence changed its investment plan or selling motion to serve that more custom nature required by the industry? Or is that even needed at all?

Yes. Great question. I mean, this is a big trend, right, design of custom silicon. I mean we have talked about it for years. System companies doing silicon. And as you know, about 45% of our business is coming from system companies, and 55% is coming from semi companies. And so -- so with this, especially with AI, there is acceleration of custom silicon, and I think 1 different from 6 months ago or 1 year ago to now is when I look at these big system companies, they are more and more committed to custom silicon. And of course, we have a great partnership with NVIDIA and NVIDIA is going to do phenomenally well. But so will custom silicon and we can see from Broadcom results and we also work very closely with Broadcom and the customers themselves. So -- and there's opportunities because the demand is so high in terms of -- if you look at all these big customers, they're projecting AI compute demand to grow like 2x every year for next several years. So I think there is growth for everyone involved in that. And the benefit of doing custom silicon, at least for the inference part, can be so high that they are willing to invest in EDA internal chip design.

So I think the financial and the customization benefit for our customers. And these are, of course, the biggest companies in the world is significant doing custom silicon. You can look at all the big ones like Google and Meta and all the others like Microsoft, Amazon, Tesla. So I think there's going to be acceleration of that. And as they do more internal design, of course, they need to invest in EDA and IP and hardware. So I think the trend is healthy there. profitability questions are similar. We want to have discipline in our pricing. So our profitability is similar, but the benefit to our system companies is high as they do their own chips.

And our next question comes from the line of Ruben Roy with Stifel.

Anirudh, I had a quick question, I hope on a comment you made during your prepared remarks about collaborating with a customer on next-generation agentic AI solutions. I'm wondering, is that something that you're seeing across a wide swath of your end customers? And if so, just wondering if you could walk through maybe some of the implications of that, whether it's how some of those collaborative efforts on that type of solution might be monetized longer term? And how you're thinking about Agentic-AI overall relative to specific it almost sounds like custom solutions by customer versus a broader Agentic AI solution set that cadence might offer to the broader ecosystem.

It's a great question. We could talk for a while on this one. And -- and we are privileged to have the partnership with several companies on AI. I mean not just the design of AI, but AI for design in our solutions and especially on agentic AI because this is a new emerging area. We have like 5 major AI platforms. But what is unique about agentic AI, of course, is all the gen AI stuff. And if you look at even 1 of the biggest applications of AI is kind of vibe coding or software development. But if you look at it, part of the chip design is also coding. We have automated, like I mentioned earlier, 90% of the workflow for chip design. But 1 part of workload, which is not automated is the customers still have to write RTL. RTL is like a -- is like a language, registered [ trans ] language that describes the chip. And this happens in the very beginning part of the chip design process. So that process is still manual. But the algorithm that is helping vibe coding or C++ coding for general software development, kind of these agenting methods can also help for RTL development, okay?

And it can provide a lot of benefit to this 10% of the workflow that is not automated. So therefore, we have a massive investment in Agentic AI, which you will see as we announce more products going forward. And we already have several partnership in there, and we are highlighting 1 of them. And the way we are going to market there is this is longer is through [indiscernible]. I've talked about JEDI before. So Jedi is joint enterprise data and AI platform. So it does have some standardized component. The database is standard, all the models are available. AI models has interfaced to all our AI tools. So part of Jedi is standard across all customers, and we work with foundries and all to kind of train our models.

Now part of it could be customer specific, okay? And in that case, the data is held at the customer side. And that's why we architected [indiscernible] from the very beginning to be both on-prem and cloud-based because sometimes the customers want it cloud-based, but sometimes if they want data to be localized, they want it on-prem. So that's why for years, we have invested in this kind of unique platform, JEDI that allows us not just to build unique solutions like RTL development and verification plan development, but also deploy it either in a general way or more specialized to a particular big customer. But I'm pretty optimistic in how identic AI can automate the remaining kind of part that was manual and again, focus our customers to do higher-level tasks and remove some of the mundane task of RTL coding, verification plan generation, things like that.

And our final question comes from the line of Joshua Tilton with Wolfe Research.

Thank you so much guys for sneaking me in here and congrats on a very strong quarter. Given the time, I'm just going to actually ask a pretty direct clarification question. John, I think it's pretty much for you. In the event that you do see some impacts in the China region, given the ongoing tariff negotiations this coming quarter, do you feel or can you help us understand how you kind of handicap the updated guidance for some -- for some, if any, potential negativity in the region?

Josh, I mean that's a great question. I'd love to be able to tell the future. The -- I mean, as always, we incorporate prudence for all kinds of regulatory variability and we base our guidance, assuming that today's export regime remains substantially similar going forward through the end of 2025, but it's very, very hard to predict what's going to happen. But by all reports that we've hired that we believe that geopolitical tensions are lower than people expect.

Helpful. Congrats again on a good quarter.

And I will now turn the call back to Anirudh Devgan for closing remarks.

Thank you all for joining us this afternoon. It's an exciting time for Cadence with strong business momentum and growing opportunities with semiconductor and system customers. With a world-class employee base, we continue delivering to our innovation road map and working hard to delight our customers and partners. On behalf of our Board of Directors, we thank our customers, partners and investors for their continued trust and confidence in Cadence.

And ladies and gentlemen, thank you for participating in today's Cadence Third Quarter 2025 Earnings Conference Call. This concludes today's call, and you may now disconnect.