Penguin Solutions Stock

Well, hello, everybody, and welcome to the Penguin Solutions fireside chat at the Goldman Sachs Communacopia + Technology Conference. I have the privilege of hosting Kash Shaikh, CEO of Penguin Solutions. My name is Kat Murphy, and I cover Penguin and IT hardware here at Goldman Sachs. We'll have about 35 minutes for today's discussion, inclusive of Q&A towards the end.

So maybe to go ahead and get started, September marks 7 months since you joined Penguin as CEO, joined at an interesting time as the company transitioned from a holdco into an AI solutions-oriented business. For the purposes of this audience, can you help us understand your current business mix and where Penguin is really positioned to go after this investment cycle around AI?

Thanks, Kat. So in the last 7 months, we have focused on 3 main areas. First, prioritizing our data center AI infrastructure business and our integrated memory business. These 2 businesses have very high demand because of the super cycle of the infrastructure, and they represent our AI-driven businesses.

Second, we have increased our investments in product innovation. And then third, we have accelerated go-to-market execution with a clear focus on 2 main segments, neocloud segment and the enterprise segment for our AI infrastructure business.

We also introduced what we call our AI factory platform. This platform combines differentiated products like an OEM offering as well as a system integrator like end-to-end services design, build, deploy and manage. This allows us to be the single partner for our customers building the factories, AI factories as well as operating the factories on their behalf. And the results show the progress.

In Q3, which was our last earnings that we announced publicly, our AI-driven businesses represented 74% of company net sales, and they grew 104% year-over-year. Company also had the record net sales at company level. And based on our guidance that we provided in the last earnings forecast, we are expecting yet another record quarter at company level, and it will be our first over $500 million net sales for the company.

Great. You also provided a preliminary outlook on earnings last quarter for fiscal '27 and guided to 30% year-over-year growth at the consolidated level, which implies an acceleration sequentially. Can you talk about what is informing your confidence in that sequential acceleration? I know there's moving parts within the business. And where in the business, across those 3 opportunities that you outlined, you see the most incremental upside opportunity?

So as we mentioned in our last earnings, our confidence comes from, first of all, the strength of the market we play in. We have 2 tailwinds, very high demand data center AI infrastructure market, very high demand memory market. And we have a very unique product market fit for both of these product lines.

And then combine that with the visibility that we have with our bookings, backlog as well as the pipeline that is really driving the outlook for FY '27. And in terms of the upside over the outlook, it really depends on how fast we can convert some of the large AI factory opportunities that we have in our AI infrastructure business.

You've reported historically the company in 3 different segments, but I want to focus on advanced compute and then the integrated memory opportunity. First, on the advanced compute side, can you break down the lineup of hardware, software and services within your portfolio and how you're going after both the neocloud and enterprise opportunities that you talked to with those products?

Yes. So within our advanced computing, we primarily focus on our AI infrastructure solutions business. This is where we offer the AI factory platform. And this AI flagged factory platform is a very unique combination of products.

These products include ClusterWareAI, which is an AI factory operating system. Just like an operating system like Windows will provide a platform to combine all of the resources between memory as well as applications and CPUs. Our AI factory ClusterWare operating system provides a similar consolidation of the cluster as well as what we have done with our ClusterWareAI recently in the last 6 months, we have also created agentic experiences.

As an example, the operators can now use the natural language to find out the health of the GPUs as well as the utilization of the GPUs. So that is one of the key differentiators from the product perspective in our platform.

We have also invested in our MemoryAI product line. This is a product line where we are leveraging our unique insights from decades of experience in the memory architecture as well as the AI infrastructure data center build-outs. And this MemoryAI product line provides benefits, especially for the inference workloads as AI moves from training to inference powering agentic AI. This is where the context sizes of the messages are very long. And these require memory architectures that can help accelerate the LLM responses and performance. So that's the next product line.

We have also introduced ComputeAI product line within this portfolio. What it is, is essentially is a set of GPUs and accelerators from NVIDIA as well as from AMD. So we can provide the products to our customers that help them manage the factories as well as build out the factories.

Then we combine them with our end-to-end services. So we get involved in much earlier in the cycle with our customers. So we come in, we design the factories on their behalf. And as we are designing the factory, we recommend our products, partner products. And then we also build out and do the system integration for them. And then we sign contracts with them 3 to 5 years to manage those factories.

So this is this unique combination of offering like an OEM product company, combine that with an end-to-end system integration offering that allows the customer to work with us as a single build partner as well as a single operate partner for the AI factories, primarily focusing on neocloud build-outs as well as large enterprises.

That's very helpful. So you talked about AI factories, MemoryAI, ComputeAI services, that's kind of the appropriate way to summarize them. And you're going after different customer verticals. For the neoclouds, which of those 4 products resonate most? And for the enterprise, is there a different kind of go-to-market or sales motion that you're trying to engage with your customers on?

So it depends on the journey of the customer, right, whether it is a neocloud customer or an enterprise customer. In some cases, we start small as in they may be looking for GPUs, and we provide them GPUs with our ComputeAI product line, whether they are enterprise or neocloud, and then we expand with our services or other products. So that's our land and expand strategy across both enterprises and neocloud providers.

However, in some cases, the land is pretty significant, what we are seeing, especially in the last 6 months, what we have seen is the acceleration of the neocloud providers because both the hyperscalers that are giving them the business as well as the business they get from the enterprises, that creates a very unique opportunity for Penguin Solutions where these neocloud providers are relatively newer providers.

So they are setting up the entire data center. So they're able to go out and, let's say, lease the data center and the power. Then we go in and we do everything for them, as in design it, procure the hardware for them and the software, build it out and then we are signing contract to help them manage the factory.

So they can focus on getting the offtakers as in their customers, and we are the builder and the operator for their AI factories. So it really depends on where they are in the journey. In some cases, it's a smaller land with expand, in some cases, a pretty significant land, and then they continue to expand with us as they are building out new data centers.

Something that's also unique to Penguin is your agnostic approach to other third-party hardware within some of these ecosystems that you're helping set up. Can you talk about how that is a point of differentiation and why and what in Penguin's portfolio, maybe this goes to the ClusterWare software operating system layer, allows you to bring forward a best-in-breed solution rather than a sole vendor solution?

So we believe in, first of all, an approach that is much more focused on meeting the objectives of our customers. Let's say, it is a neocloud customer. Neocloud customers typically sign the SLAs with their offtakers or their customers. And based on those SLAs, we get engaged, as I mentioned, much earlier in the cycle with them.

And as we are designing those AI infrastructure for the factories, we build the architecture in a way that it meets their requirements, whether it is our products or it may be the products of our partners such as Dell and along with the ClusterWare.

And the advantage of the ClusterWare, especially with our engagement with the customers, which starts much earlier than the product companies, we have the discussion and conversation at the architecture layer, as in you have these SLAs for you to meet these SLAs, this is the kind of architecture we represent, and also the fact that our ClusterWare is hardware agnostic.

So whether it is NVIDIA, whether it is AMD products, we can provide them the support of this operating system that is helping them manage their GPUs irrespective of the vendor and helping them achieve their objective without necessarily either just dropping the hardware on their doorstep or necessarily forcing them to use a single vendor, which is usually not the case because, as I said, depending on their SLAs, they have different requirements, and we meet the customer where we believe we meet their requirements versus just forcing our product.

Maybe to help illustrate all of the opportunities that you have and the ways you're participating in these types of engagements, could you talk through an example? I know publicly you've talked about a Tier 1 financial customer that you won jointly with Dell. Maybe talk more about what that sales motion looks like, where you're involved versus where Dell is involved, and how you're capturing value across that design, build, deploy and manage cycle?

Yes. So this large -- and it's actually a good example of some of the things I mentioned. So this large bank put out an RFP, and their driver was they were consuming the AI infrastructure from a neocloud provider. However, they had some new applications that they are developing, which they believe that will help them create new revenue streams.

So they put out an RFP. And in that RFP, obviously, we competed against the product companies, product OEMs. We competed against system integrators. And their requirement was they were building this on-premise AI factory the first time, and they needed someone who not only can provide them the product, but help them manage this factory because these factories are pretty complicated and it requires deeper understanding of the products and architectures to be able to deliver on the requirements.

And we were able to win this RFP. And then when I sat down with the CIO to understand why us versus -- obviously, we had very large product companies on the other hand and very large system integration company. It came down to 3 areas at the high level.

First of all, they were looking for something that is much more comprehensive, as in not necessarily just give me this product, this product and this product. They had their objectives, which is I have new revenue opportunities and I need someone to help me with the factory. But my goal is to develop these applications that will help me with revenue generation opportunities.

So the fact that we provide the whole design, build and deploy to meet their objectives was an advantage so that they are not piecemealing all of these products to be able to stitch together their first on-premise factory, which is a pretty significant size. So that was number one.

Number two was our ClusterWareAI product because, again, that ClusterWare makes it easier to manage and automate the monitoring and management of the AI factories. And the third one, interestingly, was our MemoryAI product line.

And they were the first customer who actually bought it, considered it now that we have -- now we have other customers, but they wanted to make sure they were considering the applications that are much more context-rich inference applications, and they wanted to make sure that they have technology that can enable the performance and faster response for the LLMs. So those were the 3 differentiators for us in terms of winning that project.

That's very helpful. I want to touch both on the ClusterWare piece as well as the MemoryAI and maybe transition to the memory side of the story as well. But first on ClusterWare, how are you using ClusterWare as part of your sales motion? Is it something -- maybe said differently, is this something that is included as part of the pitch as to how you can manage an estate across kind of the end-to-end nature of it?

Or is it something eventually you could package and sell separately? What are the alternatives to ClusterWare? Like what are AI factories who aren't using ClusterWare doing as an alternative? Anything to just contextualize the moat that you may have built with that platform?

Yes. So the current -- first of all, the current motion of ClusterWare is primarily with our services. So as we get involved, whether it is design or management services, ClusterWare is the software our managed services team is using.

However, our goal in ClusterWare is to make it stand-alone and especially with the agentic experiences that we are creating, these agentic experiences will allow users who are not as much familiar with our product can easily use it, as I mentioned, with natural language interface that we have provided.

So short term, primarily a differentiator for us to win larger projects with our services business. Longer term, our goal is to make it available to the customers, whether these are neocloud customers and enterprise customers.

And that will also help us more with the foot in the door land and expand strategy where we may not have the entire project, but we start with the software. And then as they realize the benefits of the software, we can offer other products and other services.

And in terms of the alternatives, that's an interesting question because there isn't an equivalent of ClusterWare because most of the alternative software available, they are very vendor-specific. And the challenge in the vendor-specific software is, again, you have to piecemeal different pieces of infrastructure when you are managing the factory -- you are building the factory or managing the factory. So the challenge becomes increased time, manual configuration, and may not result into the best utilization of the infrastructure.

That's very helpful. Let's talk more about the MemoryAI compute platform or MemoryAI platform as well as the offering you have within the broader integrated memory portfolio. What exactly are you selling within integrated memory? And how is MemoryAI an extension of that? And what are your core competencies that are starting to drive differentiation in the way that you serve customers as inference demand increases?

So in our integrated memory business, we build memory modules for OEMs. These are large OEMs. You can think of large networking companies, large compute companies. And our moat there is essentially, at the high level, engineering expertise to design those memory modules because this is just not just memory, it is a memory card that goes -- plugs into as an interface within the OEM products.

So we have engineering capabilities. We have validation capabilities. We build them, we validate them. And then we have manufacturing capabilities. We have factories where we manufacture those cards. And then we have the supply chain capabilities because we manage this business at scale.

What we have done recently with this business at the high level, first of all, we are primarily prioritizing data center customers because the memory demand is high everywhere, but a lot of this demand, in some cases, is a function of just supply and demand. We believe the demand in the data center is driven by AI.

And we believe the demand in the data center will be more durable than the cyclical memory cycles we have seen, which is why we are very disciplined and focused on the data center. So that's one of the things we are changing and it is helping us both in the short term, and we believe it will help us with the durability of the business.

The second thing is focusing on new AI infrastructure opportunities. Looking at the use cases such as the inference I mentioned with the adoption of agentic AI requires more memory in terms of having the capability to store the context so that the GPUs don't have to compute it all over again as the LLMs are doing the inference for an application. And CXL is one of the standard compute express link. So we have been the early adopter of CXL.

And using CXL, we create 2 products. One of the products is our memory expansion cards. So these are just the CXL-based memory expansion cards based on that standard that we provide for the OEMs as a part of our OEM go-to-market with the integrated memory business. Then we took that CXL capability and created an appliance, which is MemoryAI KV Cache appliance. So you can think of an appliance which has both the hardware and the software to be able to provide the memory capabilities for the LLMs to store the context.

So let's say, if the LLM is writing a book. One of the options is every time it is writing a new sentence, it's doing the whole computation again. The challenge becomes inefficient use of GPUs and high bandwidth memory, or you can have a memory bank like our MemoryAI KV Cache appliance, where you can store the context of the book you have written so far. So next time when you are writing the next sentence, you don't have to compute all over again as an example. That's really the benefit.

So what it really does is providing you the faster response for the LLM. And it is also saving the computation power for the GPUs. So what is the net effect? You may not need GPUs for all the conversation, so your spend can go down. And then you will have faster responses for the LLMs, which means better economics and faster responses.

That product is part of our AI infrastructure data center design, build and manage because it is an appliance that is connected to the GPUs versus the CXL cards are the part of the go-to-market for the integrated memory business.

You've talked about the focus on neoclouds and enterprises customers, but these CXL memory cards seem like they would also resonate with the hyperscale type customer, which would be part of this prioritization of data center. Is that also an opportunity either from a shipment perspective, technology licensing perspective, can you go after the CXL hyperscale part of the market?

There is definitely an opportunity and especially with the focus -- the newer focus of our integrated memory business on the data center in general. So we are working with data center customers, including hyperscalers that are considering our CXL, especially in the environment where, let's say, DDR5 is a bottleneck.

So they are looking at all the possibilities to be able to get the maximum throughput as well as the performance for their GPUs. And then as I mentioned, the MemoryAI can be both hyperscalers, can also be neocloud providers and the enterprise use case with the Tier 1 potential.

Something unique to Penguin's model as well as in integrated memory is that memory is largely a cost that you get to pass through. And there's also some opportunity in instances to build inventory and earn a spread in markets like this where prices are up significantly.

Can you talk to some of the dynamics that are influencing the growth in revenue and the margin expansion opportunity within memory, maybe breaking up what is coming from market price increases versus incremental drivers of demand?

So overall, our growth in the integrated memory business is a combination of both. So prices are going up, which are driving the revenue higher. However, what is more encouraging is the volumes. The volumes are increasing across the board, especially as we discussed for the data center product.

So while, let's say, our integrated memory business in Q3 grew 111% year-over-year, we had as much backlog remaining in the bookings that are bookings that we have received that we are going to be able to ship in the future, and that's a function of the volume growth. And the prices, obviously, at some point, may stabilize. What we see is the increase in volume will still drive higher revenue opportunities for us, especially in the data center.

Can you talk to how Penguin is navigating through some of the supply scarcity? Obviously, you're building a backlog, but demand and availability are both informing that. How are you working with suppliers? And is there anything unique about the types of OEMs that you're serving in integrated memory that may insulate you from some of the impacts of memory scarcity in the broader market?

So there is no denying that there are supply challenges in the overall market. However, some of the things that are helping us and our strategic advantages for us, we have a set of OEM customers in the integrated memory business that have supply agreements with the memory suppliers, whether it is Micron or SK hynix or Samsung. And that represents about 50% of our business, depending on the particular quarter.

Then we have this other segment of the customers where we procure the memory. And this is based on our long-term relationship with the memory suppliers as well as we have some contracts with the suppliers to be able to acquire the memory because part of the business is not just the memory, we do the design as well as validation and then the supply chain for those parts.

But in the dynamics of, let's say, when we are procuring the memory on the behalf of the customers who don't have the supply agreements, we have more pricing power, right? Because we're doing 2 things for them. We are procuring the memory for them in a supply-constrained environment in addition to creating more value with our design and manufacturing of the board.

So in general, we feel confident that we will continue to drive the growth in this business. While there are challenges with the supply, we have arrangements that are helping us with the continued growth of the business.

Great. Maybe I'll ask one more, and then we'll open it up to see if there's any questions. But looking beyond the innovations around CXL, which is still in the early days of adoption, you've also talked about developing new architectures in photonic memory.

What are some of the projects and work that you're doing there? And when do you expect to see the benefits of photonic memory appliances or memory cards start to impact the opportunity within this segment?

So our work in photonics is based upon our early investments in Celestial AI. Celestial AI was a photonic start-up, which was recently acquired about 6 months ago by Marvell for $5.5 billion. So in addition to getting the proceeds from our investment, we continue to work with them on developing an appliance that will allow very high-speed connectivity between GPUs and the HBM, high-bandwidth memory.

Because right now, HBMs are primarily directly attached and connected within the GPUs. That kind of interface, as in a photonic interface, will allow the expansion of the high-bandwidth memory beyond the direct attached memory to the GPUs. And as we know, there is nothing faster than the light. So that gives us the opportunity to really create value of sort of a shared memory platform for accelerated applications.

The availability of that product will be second half calendar year 2027, even if we have active R&D work going on with Celestial AI. But it represents the long-term opportunity, building upon our CXL memory expansion cards, MemoryAI KV Cache appliance based on CXL, this becomes a natural evolution for us, again, focusing on the data center with our integrated memory business and focusing on the next-generation architectures and represents a long-term growth opportunity for the company.

Okay. Any questions in the audience? I have a couple more here, too. Let's talk about the 2 segments together and the unique advantage of having both the expertise on the memory side as well as this very services-oriented platform on the compute side.

Why is that interconnected nature of the 2 segments increasingly important for Penguin? And how does it differentiate Penguin against -- in the example of the Tier 1 financial win against some of the other projects that you bid against?

So at the high level, first of all, it's a very unique position in the market, especially as AI transitions from training to inference, memory becomes really strategic for the AI -- next phase of AI. So we believe that allows us to continue to innovate at the intersection of AI infrastructure and memory, continue to create innovations that can help us lead the second phase of AI with inference taking off.

And at the same time, we plan to continue to invest and differentiate in our AI factory platform. The ClusterWareAI that I mentioned, we plan to continue to invest in making it much more agentic and much more easier to use for the customers to build and manage their AI factories.

So the continued evolution, continued investment in our AI factory platform, also using our unique position at the intersection of memory and AI infrastructure to create new solutions and at the same time, continue to create operational efficiencies, so we have the operating leverage and deliver the profits to the business and our shareholders.

We've talked about all of the ways in which you are enabling AI adoption for hyperscalers, neoclouds and enterprises, but Penguin is an enterprise itself. What applications of agentic AI are you seeing internally? And how is that helping some of your initiatives around things like investing in ClusterWare, go-to-market as you expand your customer type, things of that nature?

So within our company, we are using AI across all of our functions. And one of the use cases that we have realized that is -- where we are seeing the productivity and the return, it's a pretty clear ROI, is the code generation. So we are using agentic AI to develop the code and deliver much faster code for our ClusterWareAI. That's one of the use cases.

And interestingly, what we see, especially with the large enterprise customers, that is one of the use cases they are considering when they are moving or expanding beyond neocloud, consuming AI infrastructure from the neocloud to building their own AI factories.

And what we see with these enterprise customers, the driver is really economics. So what happens at scale for a code generation application, even if they deliver a pretty clear, higher productivity, the cost becomes a challenge because of they have to pay for GPU as a service, which is a subscription, and they typically are using frontier models, which are also quite expensive at scale.

So when they are moving to on-premise, they are obviously spending a lot more upfront with the CapEx, building the factory. But then they don't have to worry about the subscription of the AI GPU as a service, as well as they are considering open-weight models increasingly so they can run these high-volume persistent inference applications such as code generation at scale on-premise much more effectively with higher security, data retention and privacy along with better economics that they would get in the cloud operating model.

Great. We have about a minute left here, but any final thoughts would be great to know as you look to your first full year as CEO and beyond, where your time is going to be most focused from a strategic priorities perspective?

So going back to the 2 high-demand businesses, we continue -- we will continue to focus on capturing our fair market share, especially as AI transitions and expand beyond training to inference in the data center, AI infrastructure business as well as integrated memory business. At the same time, continued investment in innovation, whether it is the software, MemoryAI appliances or the compute appliances. And last but not least, continue to drive operational efficiencies with using AI across the company to deliver and continue to improve our operating leverage.

And in the end, we believe, while we have started our journey and the business is growing and we have created value for our shareholders, all of the things that are happening with the infrastructure super cycle of $7 trillion that will be -- or more than $7 trillion that will be invested in the next 3 years, we believe Penguin Solutions is very uniquely positioned to capture large share of that investment and create long-term value creation opportunity for our investors.

Great. Thank you very much for being here, Kash.

Thank you. Appreciate it.

Welcome to the Penguin Solutions Third Quarter Fiscal 2026 Earnings Call. I will now hand the conference over to Suzanne Schmidt with Investor Relations.

Thank you, operator. Good afternoon, and thank you for joining us on today's earnings conference call and webcast to discuss Penguin Solutions Third Quarter Fiscal 2026 results. On the call today are Kash Shaikh, Chief Executive Officer; and Nate Olmstead, Chief Financial Officer.

You can find the accompanying slide presentation and press release for this call on the Investor Relations section of our website. We encourage you to go to the site throughout the quarter for the most current information on the company. I would also like to remind everyone to read the note on the use of forward-looking statements that is included in the press release and the earnings call presentation. Please note that during this conference call, the company will make projections and forward-looking statements, including, but not limited to, statements relating to statements about the company's growth trajectory, financial outlook and preliminary expectations for future fiscal periods, business plans and strategy, product development and innovation, market demand and shifts, supply chain conditions and cost assumptions, leadership transitions, strategic agreements and existing and potential collaborations.

Forward-looking statements are based on current beliefs and assumptions and are not guarantees of future performance and are subject to risks and uncertainties, including, without limitation, the risks and uncertainties reflected in the press release and the earnings call presentation filed today as well as in the company's most recent annual and quarterly reports. The forward-looking statements are representative only as of the date they are made and except as required by applicable law, we assume no responsibility to publicly update or revise any forward-looking statements.

We will also discuss both GAAP and non-GAAP financial measures. Non-GAAP measures should not be considered in isolation from, as a substitute for or superior to our GAAP results. We encourage you to consider all measures when analyzing our performance. A reconciliation of the GAAP to non-GAAP measures is included in today's press release and the accompanying slide presentation.

And with that, let me now turn the call over to Kash Shaikh, CEO. Kash?

Good afternoon, and thank you for joining our third quarter fiscal 2026 earnings call. Penguin Solutions delivered an exceptional quarter with record results that reflect disciplined execution and strong customer traction for our AI factory platform strategy. At the company level, we delivered record net sales and significantly higher-than-anticipated EPS. Building on our excellent third quarter we are again raising our full year outlook for both net sales and EPS.

All of our business units performed well. Growth was outstanding in our AI-driven businesses. In Q3, our AI-driven businesses represented 74% of total company net sales. and grew 104% year-over-year. These businesses consist of integrated memory and non-hyperscale AI infrastructure solutions. In these businesses, AI-driven demand continued to outpace net sales growth, contributing to a growing backlog. This gives us confidence that the AI opportunity is expanding as enterprises increasingly adopt agent workloads powered by inference at scale.

Our results this quarter demonstrate that Penguin solution is evolving into a leading AI factory platform company. We believe we are still in the early stages of a significant long-term profitable growth opportunity. Before discussing our performance in more detail, I want to address our finance leadership transition. As announced on June 1, Nate Olmstead will step down as Chief Financial Officer on July 8, to pursue an opportunity in a different industry. Nate has been an important partner during a meaningful period of transformation for Penguin Solutions. His leadership helped strengthen our financial and operational foundation, and I want to thank him for his contributions and wish him well. I am also pleased that Aaron Johnson, our Vice President of Finance and Accounting, will serve as interim CFO effective July 9. And Aaron brings more than 16 years of public company experience in the technology sector and deep understanding of our business. We have also initiated a search for a permanent CFO.

This transition does not change our operating priorities, financial discipline or focus on execution. Today, I'll cover 4 key topics before reviewing our third quarter performance. First, the market environment as AI moves into production scale inference and agenetic AI. Second, how our AI factory platform is positioned to address this opportunity. Third, the customer and partner momentum that validates our strategy. And fourth, how we are executing across product innovation, operations and go to market. Since our April earnings call, the demand environment has strengthened as AI continues to transition from early front and response experimentation to production scale inference and agenetic AI. A simple way to think about the shift is that early AI answered questions while a genetic AI performs work.

In the early phase of AI adoption, many workloads were transactional. A user asked a question, AI generated an answer and the session ended, Agenetic AI is different. Agents are persistent, context-rich and task oriented. They can operate continuously across workflows, applications and data sources moving AI from an adviser to an operator. For example, in software development, each engineer can now have multiple agents that write test, review, document and monitor code. In business operations, agents can monitor customer activity, analyze supply chain risk, that pair follow-ups and update workflows.

As entrance workload scale and Agentic AI deployments accelerate infrastructure requirement for production, AI environment continue to expand across full AI data center technology stack. Demand is increasing not only for GPUs and accelerator attached high bandwidth memory or APM, but also for general purpose compute including CPU as well as memory, storage and networking. Every GPU deployment depends on a surrounding layer of general-purpose compute and memory to feed data coordinate workflows, manage contact and connect agents to enterprise applications.

We believe this trend is strengthening demand for our memory solutions. These persistent context-rich workloads require more memory capacity, faster access to context and better orchestration. As AI moves to inference at scale, the industry is increasingly recognizing that memory, not compute alone is becoming one of the primary bottlenecks for a large context AI and trends performance. We designed our memory appliances specifically to address this bottleneck and improve both inference performance and token economics. We believe that for enterprises, sovereign EI initiatives and new cloud customers, AI factories are becoming essential for optimizing time to first token performance and token economics for inference.

These developments reinforce the strategic premise behind Penguin solutions. Production scale AI infrastructure required a full stack AI factory platform approach. Our platform connects AR infrastructure, memory operation software and operational execution to help customers deploy and operate at scale. Our second topic today is how Penguin is becoming a leading AI factory platform company at the intersection of memory and AI infrastructure.

Our platform combines 5 core elements with our partner ecosystem. First, ClusterWare AI operating system software for AI factories. Second, memory AI and integrated memory solutions; third, advanced computing systems under our compute AI brand. Fourth, [indiscernible] factory architectures and fifth, end-to-end design, build, deploy and managed services. We believe our platform differentiation comes from combining these proprietary products and services into an integrated offering. -- customers increasingly need more than just AI hardware procurement.

They need architecture that performs in production at scale, an accelerated deployment path and an operating model that helps them generate profitable revenue or deliver operational efficiencies in production time to production deployment is directly tied to time to revenue and our platform is designed to compress both customers need the ROI and superior token economics that our Penguin AI factory platform can provide.

Turning to our third topic, customer and partner momentum. With our land and expand go-to-market strategy, we continue to add new customer logos and deepen engagement with existing customers. In Q3 '26, we added 4 new AI infrastructure customer logos. New logo acquisitions are important in part because they often lead to repeat business. For example, across 4 trailing quarters from Q3 '25 to Q2 26, we added 13 new AI infrastructure logos and 7 of those customers have already increased their business with us. New customer engagements can involve longer sales cycles, but they also support deeper customer relationships repeat business and more durable long-term growth. DRAM is a strong example of the power of our partner model and technical differentiation. Penguin designed and deployed an optimized inference environment built with Dell infrastructure and NVIDIA technology to support enterprise voice, AI workloads.

In Q3, we expanded our engagement with TAM to support additional production capacity as its business scales. And [indiscernible] also acquired our ClusterWare product and additional services. In Q3, we also expanded our previously disclosed engagement with a Tier 1 financial institution. This expansion includes our Compute Express Link, or CXL powered memory KV gas server, ClusterWareAI, software and services with Dell providing AI compute. Deployments such as Hain in South Korea, which provides local new cloud CPU as a service continue to demonstrate our ability to support AI infrastructure requirements across a range of customer environments.

So does our recent Q3 customer win with a leading quantitative trading firm. Another important proof point is Spectra, a new sovereign deployment. This supercomputer system was built and deployed through our collaboration with Sandia National Laboratories and next silicon. For Penguin, sector demonstrates our ability to design and deploy complex system in a demanding national security environment. In our memory business across trailing 4 quarters from Q3 '25 to Q2 '26, we added 16 new logos and 5 of those customers subsequently increased their business with us. We also saw continued expansion with a generative AI customer that is purchasing our CXL memory expansion cards to support inference workload solutions. These relationships validate our land-and-expand strategy.

Once our customer experiences the value performance and ROI of our AI factory platform, we have an opportunity to deepen and scale our relationships. Finally, our partner ecosystem continued to recognize Penguin's capabilities. We were recently named an NVIDIA AI factory specialized partner, a recognition of our expertise in designing, building and deploying and managing full stack and video-based AI factory infrastructure. We were also recognized this quarter as the 2026 Dell Technologies, global alliances, Americas AI Partner of the Year.

Finally, let me share some insights into our product operations and go-to-market execution. In our product portfolio, we continue to invest in the areas where we see potential for durable differentiation, including our memory AI products our ClusterWareAI operating system software and our origin AI reference architectures. Memory AI is focused on addressing the memory bottleneck that emerges as AI inference workloads become contextually larger, more concurrent and more latency sensitive. Our CXL-based memory AI KV cash server is designed to keep a large context influence closer to the workload, allowing large language models to respond faster without needing to reprocess entire data sets for every front.

Memory IKB Cash can dramatically improve AI factory efficiency by enabling superior token economics for inference and Agentic AI workloads, delivering first, up to 2x higher insurance performance. Second, up to 8x lower time to fordtoken latency. And third, expanded memory capacity beyond GPU HPM by leveraging CXL memory that is approximately 4x to 5x more cost effective than the Tier 1 financial services customer we added in Q2 purchased at [indiscernible] gas servers in Q3 for their on-prem AI factory, which is initially focused on inference and agent AI for core generation using open wave MLMs.

BC AII sited cogeneration as a common use case for inference-focused on-premise AI factories where the right architecture can deliver superior token economics. Our early investment in Celestial I a pioneer in photonic fabric technology reflects our long-standing focus on memory architecture innovation. We continue to advance our memory AI photonic memory appliance or TMA, through our continued partnership with Celestial IT now part of Marvel. This appliance is designed to extend memory capacity and bandwidth for large context AI inference environments.

Together, our CXL memory expansion cards, CXL-based memory AI, KVs server appliance and memory TMA appliance reinforce our innovation agenda and long-term vision for AI memory infrastructure. ClusterWare addresses another critical need, managing and orchestrating complex AI infrastructure across heterogeneous environments. ClusterWareAI is a hardware vendor agnostic operating system for AI infrastructure in the data centers. It acts as a unified control plane across GPUs a CPU, memory and networking, allowing the infrastructure to operate as a signal cluster while reducing manual configuration and supporting faster time to production.

We recently introduced a new AI factory operations, Egypt, which provides data center administrators with a conversational natural language interface for operational insights. This is the first in a planned family of agents designed to simplify and automate cluster operations and increased administrator productivity with a human in the loop approach. Origin brings these technologies into validated reference architectures that can reduce deployment complexity and risk. Faster deployment can mean faster time to revenue for our customers. We recently introduced Origin AI inference solutions. Together, these solutions are designed to leverage Penguin solutions 4 billion hours of GPU run time experience.

They also draw on more than 30 years of expertise delivering advanced memory and AI infrastructure solutions. As our deployments continue to scale across enterprise, sovereign AI and new cloud customers, our focus is on increasing repeatability. This includes architectures, software, services and operations. Over time, this can strengthen both customer outcomes and platform economics. Within Penguin, we are applying a genetic AI across our own operations with a focus on measurable business outcomes that support our profitable growth plan.

This includes AI-assisted software development, often referred to as coating and AI-powered productivity tools designed to accelerate our product cycles. Improve time to market and make our supply chain more scalable and responses. We are also managing supply chain, working capital and growth investments with discipline as we scale to meet AI-driven demand. Turning to our third quarter performance. We delivered a record quarterly net sales of 479 million, up 48% year-over-year and 40% sequentially. Non-GAAP operating income was a third quarter record at $64 million, up 67% from the year ago quarter, demonstrating the strong operating leverage in our model as we scale.

AI-driven demand continued to drive growth in integrated memory and in the nonhyperscale AI infrastructure business within advanced computing. Together, these 2 businesses represented 74% of total company net sales and grew 104% year-over-year. Advanced computing net sales totaled $138 million, representing 29% of total company and growing 4% year-over-year, demand and customer engagement of our AI infrastructure business within advanced computing remains very strong across enterprise, sovereign AI and neo Cloud customers.

Our non-hyperscale AI infrastructure business is scaling rapidly, with third quarter net sales up 81% year-over-year. In Q3, this non-hyperscale AI infrastructure business represented 58% of total advanced computing net sales. versus 33% in the third quarter of last year. The shift to inference at scale with [indiscernible] aligns directly with Penguin's AI factory platform strategy and the differentiated role we play at the intersection of memory and compute infrastructure. Our memory AI appliance continues to gain traction. It generated both revenue and new bookings this quarter. The pipeline also continues to strengthen across enterprise, sovereign AI and new cloud customers.

Memory business net sales were outstanding at $275 million, up more than 111% year-over-year, supported by both higher volume and pricing. Our strong results reflect a fundamental shift genetic AI is driving sustained structural demand for memory. Reinforcing our view that this AR-driven demand is more durable than a traditional cyclical memory of turn. We are entering our memory business on data center market, where demand is driven by agent, we are growing our engagements across networking, hyperscale infrastructure and enterprise computing customers for the data center products. We exited the third quarter with a very strong backlog.

This reflects robust data center-focused AI demand. This demand continues to outpace net sales growth. giving us a strong backlog entering the fourth quarter. Just as important, we are broadening that growth across a wide set of memory customers. This is creating a more durable customer base. This momentum reinforces the expanding reach of our data center-focused memory solutions. It also highlights the strength of the opportunity ahead. Our CX memory expansion cards continue to gain traction, generating both revenue and new bookings this quarter, while our customer pipeline continues to strengthen.

Our LED business also performed well this quarter with Q3 bank sales totaling $66 million, up 7% year-over-year. Our strategic priorities remain centered on memory and AI infrastructure within advanced computing. At the same time, LED continues to be managed with discipline. It generates positive cash flow and continues to execute against its innovation road map. On June 1, we announced an improved financial outlook. At that time, we expected full year fiscal 2026 net sales and diluted EPS to be at the high end of our previously issued ranges. Today, based on our third quarter results and supported by strong AI driven demand, we are further increasing our fiscal 2026 outlook for both net sales and diluted EPS. Nate will provide more details in a moment.

Looking ahead, we intend to continue balancing growth investments with operational discipline. Our priorities are clear: first, invest in differentiated software, AI memory and compute infrastructure solutions. Second, execute with discipline and speed. Third, add new logos and deepen existing customer relationships; fourth, grow our partner ecosystem and diversify our customer base. Together, these actions are designed to support more consistent and predictable, profitable growth. In closing, AI is entering a production phase. This phase is reshaping both architecture and economics of data centers as Agentic AI makes influence more persistent and demanding, several capabilities are becoming more essential. These include memory orchestration full stack integration and operational discipline.

We believe we are strategically positioned at the intersection of memory and AI infrastructure to lead the next phase of inference at scale powering its genetic AI workloads. We are still in the early innings of a very large growth opportunity. Thank you to our employees, customers and partners. Our teams are energized by the opportunity ahead. And we appreciate their continued focus on execution and customer delivery.

With that, I will turn the call over Nate for a closer look at our financial results and outlook.

Thanks, Kash. I will focus my remarks on our non-GAAP results, which are reconciled to GAAP in our earnings release tables and in the investor materials available on our website. With that, let me now turn to our third quarter performance. In the quarter, both net sales and profits were significantly higher than expected, driven primarily by accelerating AI driven demand for our memory products and continued adoption of AI infrastructure solutions which drove a 40% sequential increase in our overall net sales and 42% sequential increase in our non-GAAP operating income. For Q3 FY '26, total Penguin Solutions net sales were a record $479 million, up 48% year-over-year.

Non-GAAP gross margin came in at 28.1% and above our expectations and down 3.6 percentage points versus Q3 last year. Non-GAAP operating margin was 13.4%, up 1.5 percentage points versus last year and non-GAAP diluted earnings per share were $0.84, up 79% year-over-year and up 62% versus last quarter. In the third quarter of fiscal 2026. Our overall product net sales were $414 million, representing 87% of total company net sales and growing 60% versus the prior year. Services net sales totaled $65 million or 13% of total net sales and were down 1% versus the prior year.

Net sales by business segment were as follows: In advanced computing, Q3 net sales were $138 million, representing 29% of total company net sales and grew 4% year-over-year and 19% sequentially. This sales increase reflects stronger AI infrastructure sales to enterprise customers, which more than offset reduced sales to hyperscale customers. Within advanced computing, our non-hyperscale AI infrastructure business continues to scale rapidly with net sales up 81% year-over-year in the quarter. In addition to accelerating growth in this part of our business, we continue to make good progress on diversifying our net sales to new customer categories.

In Q3, the nonhyperscale AI infrastructure business represented 58% and of total advanced computing net sales versus 33% in the third quarter of last year. We continue to see strong demand from enterprise, neo cloud and sovereign AI customers and expect these customer categories to represent an increasing share of our advanced computing net sales over time. In integrated memory, Q3 net sales were $275 million representing 57% of total company net sales, up 111% year-over-year and 60% sequentially. And in optimized LED, Q3 net sales were $66 million, representing 14% of total company net sales and were up 7% versus the same quarter last year. Non-GAAP gross margin for Penguin Solutions in the third quarter was 28.1%, down 3.6 percentage points year-over-year and down 3.1 percentage points sequentially.

The primarily attributable to the ongoing wind down of our Penguin Edge business and a shift in the overall mix of sales across our business units. These factors were partially offset by AI-driven demand which supported favorable pricing in our integrated memory business during the quarter. Non-GAAP operating expenses for the third quarter were $70 million, up 9% year-over-year and up 14% versus last quarter. The increase in operating expenses sequentially is due to normal seasonality, increased investments in R&D, including for our ClusterWareAI software and memory AI solutions and higher variable compensation as a result of our strong net sales and profit performance.

Q3 non-GAAP operating income was $64 million, a third quarter record, up 67% year-over-year and up 42% sequentially. Operating margins were up 1.5 percentage points versus the prior year and 0.2 points sequentially driven by strong operating leverage. Non-GAAP diluted earnings per share for the third quarter were $0.84, up 79% versus Q3 last year and up 62% versus the prior quarter. Adjusted EBITDA for the third quarter was $68 million, up 51% year-over-year and up 34% versus the prior quarter.

Turning to the balance sheet. For working capital, our net accounts receivable totaled $704 million compared to $293 million a year ago, with the increase primarily driven by significantly higher memory sales volumes and prices. Days sales outstanding remained healthy at 53 days, up from 47 days a year ago and up from 50 days last quarter. Inventory totaled $498 million at the end of the third quarter, up from $184 million a year ago, reflecting increased memory costs, growth in our memory and AI infrastructure businesses and strategic purchases to maximize supply for anticipated future memory demand.

Days of inventory were 42 days, up from 36 days a year ago and down from 51 days last quarter, primarily due to the timing of receipts and shipments. Accounts payable were $736 million at the end of the quarter, up from $272 million a year ago, due primarily to higher memory costs, growth in our memory and AI infrastructure businesses and the timing of purchases and payments. Days payable outstanding were 62 days compared to 53 days last year and 63 days last quarter. The year-over-year and quarter-over-quarter movements were due to the timing of purchases and payments. Despite the significant growth in our net working capital, our cash conversion cycle remained within our typical range at 33 days, an improvement of 5 days compared to Q2 and up 3 days versus last year due to the timing of purchases and payments.

Consistent with past practice, days sales outstanding, base payables outstanding and inventory days are calculated on a gross sales and gross cost of goods sold basis, which were $1.22 billion and $1.09 billion, respectively, in the third quarter. As a reminder, the difference between gross and net sales is primarily related to our memory businesses, logistics services, which are accounted for on an agent basis meaning that we only recognize the net profit on logistics services as net sales. Cash, cash equivalents and short-term investments totaled $440 million at the end of the third quarter, down $295 million from Q3 last year and down $49 million sequentially.

The year-over-year decrease was primarily due to proceeds from the issuance of preferred shares in Q2 of last year offset by debt repayments for our term loan in Q4 of last year. Sequentially, the cash decrease was primarily due to investments in working capital to fund our growth and partially offset by approximately $40 million received from proceeds from the disposition of our 19% equity investment in Delia Technologies. Third quarter cash flows used for operating activities totaled $75 million compared to $97 million provided by operating activities in the prior year quarter. The decrease in cash flow in the quarter versus last year was due primarily to investments in net working capital to support the significant growth in our memory and AI infrastructure businesses. For those of you tracking capital expenditures and depreciation, capital expenditures were $3 million in the quarter and depreciation was $5 million for the quarter.

Wrapping up our cash flow activities, we spent $9 million to repurchase approximately 466,000 shares in the third quarter under our stock repurchase program. As of May 29, 2026, a an aggregate of $56 million remained available for the repurchase of our common stock under our current authorizations. And now turning to our outlook. Given our solid performance over the first 9 months and an improved Q4 outlook for our memory business, we are raising our full company net sales and non-GAAP diluted EPS outlook for the year. which at the midpoint now calls for 22% net sales growth and $2.60 of non-GAAP diluted EPS, up from our previous outlook of 12% net sales growth and $2.15 of non-GAAP diluted EPS.

As a reminder, our full year outlook assumes that we will continue to diversify our customer sales mix and does not include any advanced computing AI hardware sales to hyperscale customers. And also consistent with our assumptions from last quarter, our FY '26 financial outlook reflects the ongoing wind down of our high-margin Penguin Edge business. We expect sales from this business to essentially cease by the end of fiscal 2026. The combined effect of these two assumptions reduces our FY '26 growth outlook by approximately 14 percentage points at the total company level year-over-year and approximately 30 percentage points within advanced computing.

With that said, our full year net sales outlook reflects the following full year growth ranges by segment. For advanced computing, we are improving our full year outlook compared with our previous outlook and now expect net sales to decline 15% to 20% year-over-year. As it has previously, this outlook reflects the Penguin Edge and hyperscale hardware sales impacts mentioned earlier.

For Memory, we are increasing our full year outlook, and we now expect net sales to grow between 90% and 95% year-over-year, driven by continued AI-related demand and a favorable memory market environment. And for LED, we are also improving our full year outlook and now expect net sales to decline approximately 5% year-over-year. Our non-GAAP gross margin outlook for the full year is now 28.5%, plus or minus 0.5 percentage points. We adjusted our gross margin outlook up by 0.5 percentage point to account for favorable memory pricing which helped our Q3 gross margins.

Our Q4 outlook assumes less pricing favorability than we experienced in Q3, and we therefore expect some downward pressure on gross margins as we exit the year. Our full year expectation for total non-GAAP operating expenses has increased to $260 million, plus or minus $5 million. For the full year FY '26, we now expect a non-GAAP diluted share count of approximately 56 million shares, up from our prior outlook, primarily reflecting the expected dilutive impact from our convertible debt as a result of higher share prices. As a result of this dilutive impact, we expect our non-GAAP diluted share count in Q4 FY '26 to be approximately 62 million shares.

Our non-GAAP full year diluted earnings per share is now expected to be approximately $2.60 plus or minus $0.05, our forecasted FY '26 non-GAAP tax rate is now 20%, down from 22%, reflecting increased pretax income in jurisdictions to lower tax rates. Note, our Q3 non-GAAP tax rate was 17.3%, which reflects the year-to-date true-up of this new lower full year tax rate. While we expect to use this normalized non-GAAP tax rate throughout FY '26 and beyond, the long-term non-GAAP tax rate may be subject to changes for a variety of reasons, including the rapidly evolving global and U.S. tax environment, significant changes in our geographic earnings mix or changes to our strategy or business operations.

Given the strength of current AI-driven demand trends, we want to give some preliminary color on our initial expectations for net sales and non-GAAP EPS growth in fiscal 2027. Based on our current customer signals and our expectations about ongoing AI-driven demand, our preliminary fiscal 2027 view contemplates both total company net sales growth and non-GAAP EPS growth of approximately 30% from the midpoint of our full year FY '26 outlook. The EPS outlook factors in the higher share count beginning in Q4 '26 that I mentioned earlier. This is an initial planning view, and we expect to provide a full FY '27 outlook on our next earnings call. Our outlook for fiscal year 2026 and our preliminary view of FY '27, are based on the current environment and our current assumptions, including, among other things, assumptions relating to customer demand, the global macroeconomic environment, ongoing supply chain constraints, and supply costs, especially as they relate to our advanced computing and integrated memory businesses.

This includes extended lead times for certain components that are incorporated into our overall solutions impacting how quickly we can ramp existing and new customer projects and fulfill customer orders. Our outlook also contemplates the industry-wide higher cost for memory, which may slow customer demand for our products and solutions and may lower our gross margins in our advanced computing and memory businesses. We believe the combination of accelerating AI-related demand and expanding enterprise neocloud and sovereign AI customer base and our disciplined operating model positions Penguin well for continued profitable growth.

Please refer to the non-GAAP financial information section and the reconciliation of GAAP to non-GAAP measures tables in our earnings release and the investor materials on our website for further details. With that, operator, we are ready for Q&A.

[Operator Instructions]. Your first question comes from Katherine Murphy with Goldman Sachs.

Thank you for the question. It was great to see the integrated memory guidance raised up to 90% to 95% for the full year after the strong results of the quarter. Can you talk about how much of this is driven by the higher pricing environment versus demand for new products that you talk to the CXL cards?

And then to ask my follow-up now. You also discussed the broadening customer set for your memory products to hyperscale. Can you elaborate on what exactly you're shipping into that market? And how much of this contributes to the preliminary outlook for 30% growth in total company sales for fiscal '27?

Thank you, Katherine. So at the high level, for our memory business, the higher outlook takes into consideration both the volumes as well as the pricing. And this is for the overall business. CXL is a part of our business. It's a new product line that is growing, but it is a part of the business versus majority of the business is the data center focused products. And these are the products we create specialized memory modules for data center products for our OEM customers.

And to answer your second question, one of the customer is a hyperscaler, and we provide memory module for their data center product for this but it is a portion of the overall business in terms of having multiple customers, having mutable product, and it's one of the customers that we serve with our memory business.

Yes. Kate, let me just add to that. For the FY '27 view that we gave, it's really not a significant piece of that. This is not some multi-hundred million dollar hyperscale customer. It's just a portion of the portfolio, and it's kind of consistent with the run rate we had in FY '26.

Your next question comes from Brian Chin with Stifel.

Ask a few questions, and best wishes to you, Nate. Maybe for the first question, I appreciate, firstly, you're providing preliminary fiscal '27 guide for top line overall and EPS. I understand you may not want to hone in too much at this stage, but if we kind of look at your implied fiscal 4Q '26 maybe Memory revenue, you could approach, let's say, $300 million quarterly. If I just kind of flatline that, and it probably grows on pricing, et cetera, next year. That just could easily grow sort of mid-30%, maybe more % year-over-year to above sort of that baseline. But to say LED maybe is not super growthy in your assumption set for next year. Can you put some guardrails maybe on what the advanced computing growth could be in fiscal '27?

Yes. Brian, it is preliminary, right? We're just kind of -- we're in the middle of our own planning for next year. We have pretty good visibility, I'd say, into the first half at this point with backlog and customer conversations that we're having. But I would say for advanced computing, probably mid-teens, something like that be kind of the starting point. Of course, we're still -- we still have some of the impact year-over-year from the wind down of the Edge business and kind of the transition away from meta, but a lot of that would have been lead. So what you're starting to see next year in advanced computing is more of the growth coming through from the non-hyperscale AI infrastructure business?

Great. That's very helpful. And then I think cash and you both referenced a little bit about the dynamic of the order to revenue cycle and advanced computing extending somewhat. That's new customers at components, key components perhaps. I guess how much of a swing factor do you think that could be in fiscal '27, are you being pretty conservative about that in terms of anticipating some continued scarcity of Memory or other key components to grow those key segments next year?

I think the perspective from the overall business, if you look at our advanced computing, especially in non-hyperscale AI infrastructure business, some of the advantage we are going to have going into the next year is the bookings for, let's say, AI infrastructure business will deliver the revenues or the net sales going into the first half.

So that would be the advantage. As we discussed in the last earnings call, our bookings to revenue is for that part of the business. Infrastructure business within advanced computing is between 3 to 6 months. So that obviously provides us an advantage going into the next year. And we see, at least for the first half, same time frame in terms of bookings to revenue lag of about 3 to 6 months.

Your next question comes from Ananda Baruah with Loop Capital.

Congrats on all, and thanks for taking the question. I guess just to, if I could, I'll ask them at the same time. On Advanced Solutions and the broadening customer base, what are you seeing in terms of engagement from services capability perspective? I mean, what are the -- what are you doing increasing for across the new broader customer base, are you seeing people move towards more of Agenetic usage.

And then the second question is just as memory gets -- as the memory business gets bigger and you guys are playing hooking into the ecosystem and increasing role are you guys having to do anything different with the business as you go to market as it becomes more strategic and any broader context. And that's it for me. And Nate, great working with you again. I'm sure we'll work together again at some point.

Yes. So for our AI infrastructure business within advanced computing, services is a strategic advantage for us. So as opposed to alternatives where customers may procure hardware and then they have to stitch it together. The value we provide to our customers, we provide full system integration of both our products as well as other products that would include the full AI factories or AI data centers for those customers, including design, build, deploy and manage.

So that becomes an advantage for us. In some cases, we are also managing these factories for the customers for 3 to 5 years. Which, again, is an advantage for the customers to be able to get the ROI from solution we delivered. In terms of memory business in terms of our strategic direction, One of the things we have done in the last, I'd say, 3 months and increasingly going forward, focusing it on the data centers, right? And the data centers is where the demand is exploring, especially both obviously accelerated compute and then now the general purpose compute and memory because of agent TKIs driving the demand. So strategically, we are focusing on the data center at the high level in terms of the strategy.

Your next question comes from Rustam Kanga with Citizens.

Great. Congrats on the strong set of results and improved outlook and Nate sad to see you go, but it sounds like you're certainly going on a high note. My question is on the enhancements to the cluster where with the AI factory operations agent. So as these capabilities increasingly are able to sort of simplify foster operations and increase the administrator's productivity cash.

Could you maybe just talk about the relationship between the customer uptake and services adoption? And as these capabilities evolve from more conversational LLM towards more identic capabilities? Is there a potential for sort of further monetization on that estate portion of the business?

Yes. So the AI agent that we introduced is a first agent in the family of the agents. And the way to think about it is obviously helping our services differentiate in terms of providing a software that can create automation for both monitoring -- deployment and monitoring of the infrastructure so that further provides them the value through our services and we discuss services and advantage for our customers when we provide both product and services. And in terms of creating the Agentic experiences with ClusterWareAI, the benefit we are going to have is in cases where we have not fully deployed our solution in an existing data center or an AI factory. We can have the cluster were AI provided to the customer, so that the customer can start managing their existing data center or an AI factory and realized efficiency that they may not be getting with the traditional deployment. So that creates additional opportunity for us to increase our software revenues over time.

Your next question comes from Matthew Calitri with Needham & Company.

And Nate, I'll echo my regards. It's been great working with you and best of luck. I'm curious if you guys saw any change in trends in the memory pricing and supply environment over the last quarter? And if there's any color you can share on like how much of the quarterly performance was driven by pricing? And if customers -- if enterprise customers are starting to back at prices at all?

So the demand continues to increase. So while the prices are going up, and it may stabilize at some point, but we see increased demand. And it is, as we discussed earlier, is a combination of -- if you look at the total demand between the backlog and the revenue that we were able to ship the demand is increasing and our backlog is increasing.

Okay. That's great to hear. And then was there any change to the guidance philosophy with Nate's departure? And is there any update on where you are in the CFO search or what sort of qualities you're looking for in a successor?

So first of all, no change in our operating model or our strategy as a result of the transition. We have a very strong finance team and accounting team. And we are fortunate to have Aaron Johnson, [indiscernible], CFO, while we look for the permanent and he's been with the company and he's been with Nate and been working with me. So that provides the continuity to the business, and we are doing a formal search with an executive search firm where we are looking at both internal and external candidates. So the philosophy will be primarily how do we have the continued focus and disciplined execution as we have had in Q3 and before and how we look at our business in terms of being very mindful of delivering on our commitments. So no change in strategy and pretty smooth transition from my perspective.

Yes, I definitely expect a very smooth transition with Aaron. He's been my hand person for the last 2 years and worked very closely with me on all elements of finance. So I expect things to be very smooth.

This concludes the Q&A. We will now hand the conversation over for closing remarks to Kash.

Thank you, operator. Our third quarter results demonstrate strong execution in a market with durable AI-driven demand. We are excited about the opportunity ahead. Thank you.

This concludes today's call. You may now disconnect.

Ladies and gentlemen, thank you for joining us, and welcome to Penguin Solutions' Second Quarter Fiscal 2026 Earnings Call. [Operator Instructions] I will now hand the conference over to Suzanne Schmidt, Investor Relations. Suzanne, please go ahead.

Thank you, operator. Good afternoon, and thank you for joining us on today's earnings conference call and webcast to discuss Penguin Solutions' Second Quarter Fiscal 2026 results. On the call today are Kash Shaikh, Chief Executive Officer; and Nate Olmstead, Chief Financial Officer. .

You can find the accompanying slide presentation and press release for this call on the Investor Relations section of our website. We encourage you to go to the site throughout the quarter for the most current information on the company. I would also like to remind everyone to read the note on the use of forward-looking statements that is included in the press release and the earnings call presentation. Please note that during this conference call, the company will make projections and forward-looking statements, including, but not limited to statements about the market demand, technology shifts, industry trends and the company's growth trajectory and financial outlook, business plans and strategy, including investment plans, product development and road map, anticipated sales, orders, revenue and customer growth and diversification and existing and potential strategic agreements and collaborations.

Forward-looking statements are based on current beliefs and assumptions and are not guarantees of future performance and are subject to risks and uncertainties, including, without limitation, the risks and uncertainties reflected in the press release and the earnings call presentation filed today as well as in the company's most recent annual and quarterly reports. The forward-looking statements are representative only as of the date they are made and except as required by applicable law, we assume no responsibility to publicly update or revise any forward-looking statements. We will also discuss both GAAP and non-GAAP financial measures. Non-GAAP measures should not be considered in isolation from, as a substitute for or superior to our GAAP results.

We encourage you to consider all measures when analyzing our performance. A reconciliation of the GAAP to non-GAAP measures is included in today's press release and accompanying slide presentation. And with that, let me now turn the call over to Kash Shaikh CEO. Kash?

Good afternoon. Thank you for joining our second quarter FY '26 Earnings Call. This is my first earnings call as CEO of Penguin Solutions, and I'm excited to step into this role. I want to start by thanking Mark Adams for his leadership and for the strong foundation he built. Since joining in early February, I've spent significant time with customers, partners and our teams around the world. I've witnessed the strength of the company, both in our technology and our customer relationships.

What is clear is this: AI is moving from experimentation to production, with workloads increasingly shifting towards real-time inference. We are already seeing this translate into customer demand beyond hyperscale across enterprise, neo cloud and sovereign AI markets. We expect this transition to expand our addressable market and drive increased demand for integrated AI infrastructure, where Penguin is already winning. We see this firsthand in the breadth of our deployments from a sovereign AI factory, Haien in South Korea to enterprise voice AI with Deepgram to large-scale research systems with Georgia Tech along with a growing pipeline across all 3 market segments.

What makes this opportunity so significant is that the architecture of AI is also changing. Model training was largely compute bound, inference [indiscernible] agentic AI is memory bound and latency sensitive. We believe this is driving a re-architecture of the data center across compute, memory, interconnect and software. We also see AI driving memory demand. Not only for the high-bandwidth memory or HPM used with GPUs, more other accelerators, but also for general purpose memory.

General purpose compute wraps around every GPU build out and whether it's reinforcement learning pipeline or inference surveying. That workload runs on processors backed by significant memory content across the entire system. So while memory markets are cyclical, we believe AI is adding a more durable layer of demand for memory. As AI factory scale, I expect customers to increasingly prioritize partners that deliver with speed and precision, along with full stack AI factory platform capabilities, including compute, scalable memory systems, cluster management software, end-to-end services and a partner ecosystem to deliver a differentiated solution.

Time to deployment is now directly tied to time to first token. Against this backdrop, we are building Penguin into an AI factory platform company. Our AI factory platform is built around 6 core elements: first, Penguin ClusterWare, our AI infrastructure management software. Second, our new Penguin memory AI line of systems designed specifically for AI inference workloads. Third, Penguin advanced computing systems optimized for AI workloads. Fourth, Penguin OriginAI factory architectures, our reference designs for AI factories. And fifth and sixth end-to-end services and our partner ecosystem.

Production-grade AI factories require full stack design across compute, memory, storage, networking and software. We partner with leading AI companies, including NVIDIA and SK Telecom and partners like Dell. We also offer complete end-to-end services spanning design, build, deploy and manage services. We are strategically positioned at the intersection of AI infrastructure and memory with a long tack record in both. Few if any companies combine these capabilities and scale, we believe that together, our AI infrastructure and memory expertise position us to meet the evolving requirements of AI infrastructure as it shifts towards inference workloads.

This supports our ability to develop differentiated solutions. Given the momentum we are seeing in our AI infrastructure business and the significant market opportunity ahead of us, we are very focused in this area. We plan to invest more in our AI factory platform to accelerate our AI business growth. Specifically, in product innovation, go-to-market and customer engagement. In March at NVIDIA GTC conference, we announced 2 AI inference centric solutions aligned with this strategy.

First, the Penguin MemoryAI server. Building upon our Compute Express Link, or CXL-based memory expansion capabilities, we introduced a new line of scalable memory systems called MemoryAI. CXL is a high-speed interconnect that enables scalable, shared memory across GPUs and CPUs. We also announced the immediate availability of our new MemoryAI KV cache server. Here KV or key value cache stores inference context to accelerate large language model responses. Second, the expansion of our OriginAI factory architecture portfolio which now includes blueprints that address the larger workloads and the low latency demands of AI inference.

We also continue to expand capabilities of ClusterWare toward a unified, control plane for AI factory infrastructure, integrating the open ecosystem to deliver repeatable production scale deployments. To accelerate innovation and strengthen our leadership team, we recently appointed Ian Colle as Senior Vice President and Chief Product Officer. Ian brings more than 2 decades of experience building AI infrastructure platforms and scaling high-performance computing most recently at Amazon Web Services.

He was recently named by HPC Wire to its People to Watch 2026 List, reflecting his reputation in the industry. Now let me briefly address our second quarter performance. In Q2, we delivered net sales of $343 million. Non-GAAP gross margin was 31.2%. Non-GAAP diluted earnings per share were $0.52. These results reflect strong demand and execution in memory and continued progress in our AI HPC business. Before turning to the segments, I would like to address our updated outlook.

As Nate will describe in further detail, following our solid Q2 net sales and EPS performance, we are raising the midpoint of our full year net sales and EPS outlook. We are raising our outlook for our integrated memory business, fueled by AI-driven demand, strong execution by our team and favorable pricing dynamics. While our second half advanced computing net sales outlook is lower than our prior expectations.

We are encouraged by strong year-over-year Q2 bookings growth for non-hyperscaler AI HPC business, which included 5 new AI HPC customer wins that brings our first half total this year to 7 new AI HPC logos compared to 3 in the first half of last year. With that context, let me turn a closer look at each of the segments.

Starting with advanced computing, net sales for the quarter were $116 million, representing 34% of total company net sales and declined year-over-year. Advanced computing net sales for the second quarter reflect both the timing of large deployments and our transition away from hyperscaler concentration. They also reflect the previously disclosed wind down of our Penguin Edge business. We believe diversification of net sales and wind down of Penguin Edge will strengthen the long-term quality of the business.

As I mentioned, we are transitioning our AI infrastructure business from hyperscaler concentration toward a more diversified customer base across enterprise, Neo Cloud and sovereign AI. This transition is showing very encouraging progress, but we still have more work to do. Non-Hyperscale AI/HPC net sales grew 50% year-over-year for the first half of the year, representing over 40% of first half segment net sales, supported by strong non-hyperscale year-over-year booking growth in the quarter, including 5 new AI HPC logos across financial services, biomedical research and energy.

We expect further diversification in the second half of the fiscal year. Our AI/HPC pipeline continues to strengthen, with opportunities to acquire additional logos in the second half of the fiscal year across enterprise, neo Cloud, sovereign AI customers.

As previously discussed, these engagements typically progress over many months from prospecting to design to award followed by contracting and ultimately, system build and deployment. While the sales cycle can be long, often 12 to 18 months and can introduce quarterly net sales variability, it also supports deeper customer relationships, repeat business and a more durable long-term growth. I'm encouraged by the trajectory of the business and the signals we are seeing in the market.

Beyond the numbers, we are also seeing increased activity in specific enterprise verticals. For example, we recently announced our collaboration with Deepgram and Dell to support enterprise voice AI deployments. This win highlights the growing demand for low latency, production scale inference infrastructure in real-time applications. In this engagement, Penguin designed and deployed an optimized inference environment built on Dell PowerEdge servers and NVIDIA RTX Pro 6000 Blackwell GPUs. This solution facilitates Deepgram's speech to text, text to speech and voice agent functionalities for applications within health care and retail sectors.

This case study also demonstrates how design and integration expertise delivers differentiated value. As entrance workload scale, we expect these types of deployments to become an increasingly important driver of AI infrastructure demand. Georgia Tech's AI Makerspace developed in partnership with NVIDIA is a strong example. Our relationship with Georgia Tech continues to grow and validates Penguin's ability to help organizations move efficiently from concept to production-grade AI infrastructure.

Now turning to integrated memory. Net sales for the quarter were $172 million, representing 50% of total company net sales and grew 63% year-over-year. AI-driven demand remains strong across networking, telecommunications and computing market segments. Pricing dynamics were favorable and although supply remained tight, we continue to manage constraints effectively through our supplier relationships and disciplined procurement. Stepping back, our AI/HPC and memory segments taken together enable us to integrate compute and memory architecture in ways that need the requirements of production AI environments.

Memory architecture is becoming increasingly central to AI performance, particularly as inference workload scale. Our early investments in CXL position us well as customers evaluate more dynamic memory architectures. Furthermore, we are beginning to see this demand translate into customer deployments, including a recent substantial order for CXL cards from a generative AI company building solutions for inference workloads. This reinforces our strategic position at the intersection of memory and AI infrastructure to capitalize on the next phase of AI, focused on inference powering agentic AI workloads. These solutions are sold to enterprise AI infrastructure buyers.

The same customers we serve in our AI/HPC business. For example, we sold our CXL powered KV cache servers to a Tier 1 financial institution for their on-premise AI factory. In parallel, we continue to advance development of our photonic memory appliance or PMA, formerly referred to as OMA, which is designed to extend memory capacity and bandwidth for large-scale AI environment. We were an early investor in a photonic memory company, Celestial AI, reflecting our long-standing focus in memory architecture innovation and our early conviction in the importance of optical interconnects for next-generation AI systems.

Celestial AI was recently acquired by Marvell in a multibillion-dollar deal. Beyond the portion of proceeds we received from the acquisition as an investor, we are positioning ourselves for future growth in this market. As inference workloads expand, technologies like PMA can help address key memory scaling challenges in the next-generation AI systems. Last but not least, LED. Net sales for the quarter were $56 million. representing 16% of total company net sales and were down 7% year-over-year. The business continues to operate with focused leadership and dedicated operational discipline.

While market conditions remain mixed, we are maintaining a disciplined approach to investment and capital allocation. We are focused on optimizing portfolio value while concentrating resources on areas where we see the strongest long-term returns. In close, the demand for data center, AI infrastructure and memory is expanding rapidly. AI factories are becoming infrastructure that powers artificial intelligence across a range of industries.

As AI shifts towards inference and agentic systems and scales across large enterprise neo cloud and sovereign AI environments, we expect demand to accelerate. At the same time, memory is becoming a defining constraint and a defining opportunity. Penguin sits at the intersection of AI infrastructure and memory innovation. And we believe that is a powerful position to be in. Our focus is clear. We are prioritizing 4 areas. First, to invest in product innovation across our AI factory platform, particularly at the intersection of AI infrastructure and memory to drive profitable growth; second, to execute with speed and precision; third, to deepen customer engagement and our ecosystem to support long-term growth.

And fourth, to continue diversifying our customer base while building toward more consistent and predictable growth. We believe this focus positions us well to execute in a rapidly evolving market while continuing to build a durable and scalable business. With that, I'll turn it over to Nate.

Thanks, Kash. I will focus my remarks on our non-GAAP results, which are reconciled to GAAP in our earnings release tables and in the investor materials available on our website. With that, let me now turn to our second quarter results. In the quarter, total Penguin Solutions net sales were $343 million, down 6% year-over-year. Non-GAAP gross margin came in at 31.2%, which was up 0.4 percentage points versus Q2 last year.

Non-GAAP operating margin was 13.2%, down 0.2 percentage points versus last year and non-GAAP diluted earnings per share were $0.52, flat year-over-year. In the second quarter of fiscal 2026, our overall services net sales totaled $64 million, up 1% versus the prior year. Product net sales were $279 million in the quarter, down 8% versus the prior year.

Net sales by business segment were as follows: In advanced computing, Q2 net sales were $116 million which was 34% of total company net sales and down 42% year-over-year. This sales decline reflects both the ongoing wind down of our Penguin Edge business and hyperscale hardware sales in Q2 last year, which did not recur in Q2 this year. Drilling down deeper into our advanced computing results. Our non-hyperscale AI HPC net sales were down 35% year-over-year in the quarter, but up 50% for the first half of the year. given the project nature of the business, where sales can be lumpy from one quarter to the next, we believe looking at the multi-quarter trend is a helpful way to evaluate the growth in this portion of our business.

In addition to solid first half growth in our non-hyperscale AI/HPC business, we continue to make good progress on diversifying our net sales to new customer segments. For the first half of the year, the non-hyperscale AI/HPC business represented more than 40% of total advanced computing net sales versus approximately 20% in the first half of last year.

We expect to see our mix of net sales from enterprises, neo clouds and sovereign AI customers increased further in the second half of this fiscal year. In integrated memory, Q2 net sales were $172 million, which was 50% of total company net sales and up strongly with 63% growth year-over-year. And in optimized LED, Q2 net sales were $56 million, which was 16% of total company net sales and down 7% versus the same quarter last year. Non-GAAP gross margin for Penguin Solutions in the second quarter was 31.2%, up 0.4 percentage points year-over-year and up 1.2 percentage points sequentially, with strong margin performance in each business driven primarily by product mix in advanced computing, favorable pricing in memory and tariff recovery in LED.

We currently project lower gross margins in the second half driven by a higher mix of lower-margin AI hardware and memory sales, rising memory costs in our AI factory solutions and less tariff cost recovery in LED. Non-GAAP operating expenses for the second quarter were $62 million, down 3% year-over-year and relatively flat sequentially. We expect a modest sequential increase in operating expenses in the second half reflecting normal seasonality and increased investments in R&D, including for our ClusterWare software and MemoryAI solutions.

Q2 non-GAAP operating income was $45 million, down 8% year-over-year and up 9% versus last quarter. Operating margins were down 0.2 percentage points versus the prior year, but up 1.1 points sequentially driven by higher sequential gross margins in both memory and advanced Computing.

Non-GAAP diluted earnings per share for the second quarter were $0.52 flat versus Q2 last year and up 7% versus the prior quarter. Adjusted EBITDA for the second quarter was $50 million, down 6% year-over-year and up 11% versus the prior quarter. Turning to the balance sheet.

For working capital, our net accounts receivable totaled $371 million compared to $330 million a year ago, with the increase driven by higher memory sales volumes and variations in sales linearity across quarters. Days sales outstanding were healthy at 50 days, consistent with the prior year and down 1 day versus last quarter. Inventory totaled $322 million at the end of the second quarter, up from $200 million a year ago, reflecting increased memory costs, growth in our memory business and strategic purchases to fulfill memory and AI demand in the second half of the year.

Days of inventory was 51 days, up from 37 days a year ago and 38 days last quarter, primarily due to our strategic memory purchases and the timing of receipts and shipments. Accounts payable were $401 million at the end of the quarter, up from $238 million a year ago, due primarily to higher memory costs, growth in our memory business and the timing of purchases and payments.

Days payable outstanding was 63 days compared to 44 days last year and 55 days last quarter. The year-over-year and quarter-over-quarter movements were due to the timing of purchases and payments. Our cash conversion cycle was 38 days, an improvement of 5 days compared to Q2 last year and up 3 days versus last quarter due to the timing of purchases and payments. Consistent with past practice, day sales outstanding, days payables outstanding and inventory days are calculated on a gross sales and a gross cost of goods sold basis, which were $672 million and $578 million, respectively, in the second quarter. As a reminder, the difference between gross and net sales is primarily related to our memory businesses, logistics services, which are account for on an agent basis. meaning that we only recognize the net profit on logistics services as net sales.

Cash, cash equivalents and short-term investments totaled $489 million at the end of the second quarter, down $158 million versus Q2 last year and up $28 million sequentially. The year-over-year fluctuation was primarily due to proceeds from the issuance of preferred shares in Q2 of last year offset by debt repayments for our term loan in Q4 of last year. Sequentially, the cash increase was due to cash generated from operating activities as well as approximately $32 million received from proceeds from the disposition of our investment in Celestial AI in connection with its sale to Marvell technology. These sources of cash were partially offset by our share repurchase activity in the quarter.

We ended the quarter with $450 million of debt down $20 million versus last quarter due to the retirement of our 2026 convertible notes. In total, we closed the quarter in a net cash position and based on our current debt maturity schedule, have no further scheduled debt payments due until 2029. Second quarter cash flows provided by operating activities totaled $55 million compared to $73 million provided by operating activities in the prior year quarter. The decrease in cash flow in the quarter versus last year was due primarily to investments in net working capital to support growth for the second half of this fiscal year.

For those of you tracking capital expenditures and depreciation, capital expenditures were $2 million in the second quarter and depreciation was $5 million for the quarter. Wrapping up our cash flow activities, we spent $32 million to repurchase approximately 1.7 million shares in the second quarter under our stock repurchase program. As of February 27, 2026, an aggregate of $64.5 million remained available for the repurchase of our common stock under the current authorization.

And now turning to our outlook. Given our solid half 1 performance and an improved half 2 outlook for our memory business, we are raising our full company net sales and non-GAAP diluted EPS outlook for the year, which at the midpoint now calls for 12% net sales growth and $2.15 of non-GAAP diluted EPS, up from our previous outlook of 6% net sales growth and $2 of non-GAAP diluted EPS.

As a reminder, our full year outlook assumes that we will continue to diversify our customer sales mix and does not include any advanced computing AI hardware sales to hyperscale customers. And also consistent with our assumptions from last quarter, our FY '26 financial outlook reflects the ongoing wind down of our high-margin Penguin Edge business. We expect sales from this business to essentially cease by the end of fiscal 2026. The combined effect of these 2 assumptions in our FY '26 outlook remains approximately a 14 percentage point unfavorable year-over-year impact to our total company net sales growth and approximately a 30 percentage point unfavorable impact to advanced computing.

With that said, our full year net sales outlook reflects the following full year growth ranges by segment. For advanced computing, we now expect full year net sales to change between minus 25% and minus 15% year-over-year. While our advanced computing net sales outlook for this fiscal year is lower than our previous forecast, we are encouraged by our AI HPC bookings, including several new logos and pipeline growth. As it has previously, this outlook reflects the Penguin Edge and hyperscale hardware sales impacts mentioned earlier. For memory, we now expect net sales to grow between 65% and 75% year-over-year, driven by strong demand and a favorable pricing environment.

And for LED, we continue to expect net sales to decline between minus 15% and minus 5% year-over-year. Our non-GAAP gross margin outlook for the full year is now 28%, plus or minus 0.5 percentage points. We adjusted our gross margin outlook down by 1 percentage point to account for a higher mix of memory sales, which have a lower gross margin than our company average and higher memory costs in our AI hardware business. Our full year expectation for total non-GAAP operating expenses remains $250 million, and we have narrowed that range to plus or minus $5 million. For FY '26, we now expect a non-GAAP diluted share count of approximately 53 million shares, down from our prior outlook, primarily reflecting the impact of our recent share repurchases. Our non-GAAP full year diluted earnings per share is now expected to be approximately $2.15, plus or minus $0.15. Our forecasted FY '26 non-GAAP tax rate remains at 22%. And while we expect to use this normalized non-GAAP tax rate throughout FY '26 and beyond.

The long-term non-GAAP tax rate may be subject to changes for a variety of reasons, including the rapidly evolving global and U.S. tax environment, significant changes in our geographic earnings mix, or changes to our strategy or business operations. Our outlook for fiscal year 2026 is based on the current environment, which contemplates, among other things, the global macroeconomic environment and ongoing supply chain constraints, especially as they relate to our advanced computing and integrated memory businesses.

This includes extended lead times for certain components that are incorporated into our overall solutions, impacting how quickly we can ramp existing and new customer projects and fulfill customer orders. Our outlook also contemplates the industry-wide higher cost for memory, which may slow customer demand for our products and solutions and may lower our gross margins in our advanced computing and memory businesses.

Overall, we believe our focused execution, disciplined expense management and balance sheet strength provide a strong foundation for sustained profitable growth. We expect these qualities to support our continued progress as we pursue opportunities to enhance long-term shareholder value. Please refer to the non-GAAP financial information section and the reconciliation of GAAP to non-GAAP measures tables in our earnings release, and the investor materials on our website for further details.

With that, operator, we are ready for Q&A.

[Operator Instructions] Your first question comes from the line of Katherine Murphy from Goldman Sachs.

I'll ask about the raised memory segment outlook for 65% to 75% growth, how much of this is from increased favorable pricing versus demand for new product categories? And as a follow-up, how should we think about the impacts to the operating margin outlook for this segment and the investments that need to be made into new technologies like CXL and photonic memory appliances?

Kath, it's Nate. So on the memory outlook, listen, we're really pleased with the demand that we're seeing as well as the favorability that we see in the pricing environment. I would say for the increase that we're seeing in the second half, that's majority pricing, but demand is also very strong across telco, networking, AI-driven demand is just very strong. In fact, to get to the high end of that outlook really just refers to our ability to secure materials, which is really the only inhibitor we see right now to raising that outlook here in the second half.

[indiscernible] materials. We're using the balance sheet to strategically purchase ahead where we can, but the demand is very strong in memory. In terms of the investments, we've reflected it in the outlook. So I kept the OpEx for the year at $250 million, plus or minus $5 million. We're balancing the portfolio as we always do, to look for opportunities to accelerate our investments in innovation in AI or in the memory solutions that we've been talking about. But that's all included in the outlook.

I expect the operating margins for memory to remain pretty healthy in the back half of the year. I do expect some pressure on gross margins in AI as we see a higher mix of new hardware shipments in the second half as well as factoring in some of the higher memory input costs that we have in that business.

Sorry, your next question comes from the line of Brian Chin. We're experiencing some mild technical difficulties. My apologies. Your next question comes from the line of Brian Chin from Stifel.

Ask a few questions, and maybe first question, I guess in advanced computing, what changed that caused you to lower the midpoint of your prior guidance to the new range you've communicated? And can you describe how booked you are to that midpoint of that new range? .

So one of the main factor is the lag between our bookings and the revenue. Our revenue lags about 3 to 6 months of -- from the time of the bookings. And this is primarily driven by the timing of the deployments. In some cases, the material availability and so on and so forth. And given where we are in terms of our fiscal year, we have 5 months remaining. So going forward, most of the bookings that we are expecting may not materialize into the revenue for the second half of this fiscal, but we believe that it will have a positive impact, obviously, going into the first half of the next fiscal.

So that's one of the reasons that we are lowering the guidance for advanced computing driven by the deployments, but we are seeing strong momentum in our pipeline as well as bookings. Bookings grew very significantly in Q2 for non-hyperscale AI HPC business, which is very strategic for us, and we are encouraged to see the progress. We closed 5 new logos with the AI/HPC in Q2 and first half, that figure total to 7 new logos as compared to 3 new logos last year. So we are very confident in our ability to execute. The main issue at this point is timing.

Okay. Yes. I appreciate that, Kash. And it sounds like you're pretty well booked into the fiscal second half, lowered outlook and that some of these new bookings are more kind of beyond a 6-month window. Also thinking about growth in the business, obviously, there's that sort of headwind that you helped to clarify in terms of reduction in hardware revenue to the new hyperscaler, the wind down of Penguin Edge.

And so 30 percentage point impact, if we kind of net that against the guidance, maybe 10% growth for this year. net of that in that segment. So moving forward, as you survey the business and you haven't been in the role that long, can you think about what that sort of apples-to-apples growth rate was or is tracking to for this fiscal year? How are you thinking about sort of target growth rates for the advanced computing business moving forward?

So overall, let me give you a data point. So the first half of this fiscal -- our net sales grew about 50% year-over-year for non-hyperscale AI/HPC business, representing 40% of the overall mix of advanced computing, which is almost 2x of what we closed last fiscal. So the growth is substantial in terms of the bookings as well as the revenue that we see.

And we expect that to continue. And as we continue to close the booking, converting the pipeline, we see strong pipeline across all 3 segments that I mentioned between enterprises on-prem AI deployments significant activity with sovereign AI customers as well as new cloud customers.

Your next question comes from the line of Matthew Calitri from Needham & Company.

Matt Calitri here from Needham. Do the new memory launches mark a shift in strategy on that front? Just curious because in the past, the company has talked kind of more about the niche parts of the integrated memory business and noted it's early on things like CXL front. But now it sounds like memory is expected to be a larger driver as part of this AI factory platform. So just wondering if anything has changed there and what gives you confidence there's durable demand here?

Yes. So it is a part of our strategy. the MemoryAI appliances that we launched about a month ago, starting with GPC is a part of us investing more in our AI factory platform strategy. So there are 6 elements to this strategy and MemoryAI is one of the strategic elements where it is very timely if you look at how AI is transitioning from model training to inference. And in the workloads where you are focused on inference, memory becomes an increased requirement because of lower latency as well as larger contact size for inference, powering the agent.

So this is very strategic for our business. In fact, we are leading the market in this area, taking advantage of our unique position at the intersection of memory and AI infrastructure and combining the deep understanding and architecture, we introduced this MemoryAI, KV cache server as one of the products in the line of memory , we are working on other products, and we will continue to invest and in fact, invest more in this area to take advantage of the market opportunity because the timing is perfect and our leadership in the MemoryAI line of products.

To give you a proof point, the -- one of the new logos we acquired Tier 1 financial institution, not only we are deploying the AI infrastructure AI factory deployment for them, they also purchased our CXL-based KV cache server, which is a proof point of as customers are transitioning from training and bringing AI on-premise in their factories, deploying on-premise, focusing on inference and powering agentic AI, it is very strategic for us and the timing is ripe. So we expect to see this demand, and we plan to continue to invest in this area.

Awesome. That's great to hear. And then met with the new CEO in the seat and some moving pieces around sales cycles and supply chains, did you change the guidance philosophy at all or embed any additional conservatism? Any color on the puts and takes there would be helpful.

Yes. Matt, no, no change in the philosophy. We Kash and I very quickly aligned, I think, on how we think about tracking the business and looking at things. And in fact, I think with our new CRO, which -- who came in a couple of quarters ago, he's done a nice job of adding some more rigor to the planning process in our AI business and just improving the visibility there a little bit.

But it's a challenging environment from a supply chain standpoint, and we of course, got a lot of experience managing supply chain in our memory business, and I think that that's an advantage for us in an environment like this.

Your next question comes from the line of Samik Chatterjee from JPMorgan.

This is MP on behalf of Samik Chatterjee. So my first question is I just wanted to double click on your advanced computing guidance. You mentioned that a lag of 3 to 6 months for the revenue which we will book in your second half. But was there a change observed for the bookings which you did in first quarter? Or any change in relative to what were you expecting to do in 2Q? And I have a follow-up as well.

Yes. MP, I think bookings were strong in Q2, really good growth sequentially and year-over-year. I do think that the deployment cycle has lengthened a little bit with some of the supply constraints. In particular, on memory, things have gotten a little bit longer. But we're really pleased with the 5 new logos. And I think demand is good. We're seeing good strength in the pipeline, and it's also diversifying nicely across the non-hyperscale segments such as enterprise and neocloud and sovereign. So I think we feel really good about the demand. I think this is just an issue of a little bit of timing as we can convert bookings into revenue.

Okay. And my second question would be also on advanced computing and your factory related business. Like does NVIDIA coming up with their own reference designs for factory level solutions. Like how does that play relative to you? Like, is that a tailwind for you? Or is that a headwind for you? Like can you please help us understand [indiscernible]

Yes. We believe this is an advantage for us. So we work very, very closely with NVIDIA and some of the wins that I mentioned, for example, the financial institution recently along with our MemoryAI product in this transaction. NVIDIA worked very closely with us, and we are working with NVIDIA leveraging their reference design, combining that with our AI factory platform and complementing NVIDIA's [indiscernible], as an example, to provide full stack to our customers.

So their blueprints are more complementary to our AI factory platform and the components that make up for it. So we are actually quite excited about those blueprints and working very closely with NVIDIA to capture the opportunities, especially as NVIDIA is increasingly focused on enterprise, it aligns with our strategy and go to market.

Your next question comes from the line of Ananda Baruah from Loop Capital.

A couple, if I could. Kash and maybe Nate as well. Earlier remarks were that you're seeing increased momentum across neo cloud, sovereign and enterprise. And you mentioned 1 or 2 of the [indiscernible]. Do you have -- and I think Kash you had mentioned you've made some specific or at least general inferencing remarks, including around agentic. Do you have any specific context you can give us around what your customers are telling you their thrust in inferencing is right now? And maybe the degree to which agentic is showing up there. Like we just want to get a sense of what the customer's activity tone is like behaviorally, say, over the last 90 to 180 days. anything there you can share with us to make it a little bit more experiential for us? And then I have a quick follow-up.

Sure. We believe we are early in the adoption of inference with these customers, but it is increasingly deployed as in customers as they move towards Agentic. inference provides the opportunity for powering the agentic. And when you think about inference, I'll give you an example of why the architecture is changing and why memory is becoming increasingly critical in inference as compared to the model training.

So for example, let's say, if you are writing a book and if you have to write a new sentence without having the memory as a supporting component for you, you will have to reread the entire book before writing the next center. So in the inference, you're doing inference on a lot of data you already have. And if you have a component where the book you have written so far is stored. So before writing the new sentence, you don't have to reread the book that kind of how it is changing for the enterprises and other segments.

And we see customers already deploying it. And the architecture is changing, which is why not only we have the opportunity and advantage to provide them our AI infrastructure as well as the services Increasingly, we are seeing the demand for our MemoryAI portfolio where they are deploying AI infrastructure and increasingly inference, they need products like that to be able to provide that memory component for the inference, so the responses of LLM can be much more faster than they would be otherwise.

I got it. That's helpful. And just 1 last 1 quick follow-up, I'm mindful of the time here in case if anybody behind me. The CXL product, it sounds like -- to the earlier question, it sounds like you guys a little bit more enthusiastic about the CXL sleeve today than you were maybe 90 days ago, you have the new products out of GTC. Is that accurate statement that you're expecting maybe because of these new products, a little bit more and certainly some of the NVIDIA announcements of CES as well. But are you expecting a little bit more revenue a little bit sooner than maybe you were at CXL 90 days ago? And a quick second part to that. Do you need photonics to work before you really get CXL amplification like you need CPO or photonics to work before you can really amplify CXL for scale up. That's it for me.

Yes. So let me address your CXL question first. I think CXL adoption is timely given the transition to inference because, as I mentioned with inference, you need increased memory, faster LLM responses. And what CXL provide compute Express like is you can share the memory between GPUs and CPUs. So what it allows is new memory pooling, which is an advantage in inference workloads. So while CXL was obviously available for the last few quarters, it is driving that -- that inference adoption is driving the adoption for CXL and this transaction that I mentioned where we received an order, it's actually an enterprise generative AI company working on inference workloads. So you can imagine CXL cards make sense for them because those workloads need increased memory and the memory pooling capabilities provided by CXL between GPUs and CPUs are an advantage for those kind of customers.

And then in terms of photonic memory appliance that we are working on in our partnership with Celestial AI, which is now obviously Marvell, that provide increased capability because, obviously, when you have photonic connectivity, then you have increased capacity to share the memory. So it takes it to the next level.

However, CXL in itself is an advantage, we can take it to the next level with the photonic appliance. There is another element which is KV cache that I mentioned, memory AI KV cache server, which is essentially providing much more responsiveness for larger context workloads, again, used in inference. So various requirements, you can think of it as inference as various requirements related to memory and the type of workloads it has and some of it is latency.

So these components between CXL or the CXL-based KV cash, which provides increased responses and larger memory side -- largest context sizes and then taking it to the next level, photonic memory make up various use cases for inference or inference gets mainstream. We will have an advantage of this portfolio helping with various use cases of inference.

Your next question comes from the line of Kevin Cassidy from Rosenblatt.

Yes. Just the gross margin for the memory, your gross margin was up in the quarter and memory revenue was up strong, and I just want to understand that what the dynamics are there. .

Yes, sure, Kevin. We saw a little favorability in memory margins. Some of that is mix, a little bit stronger demand than flash actually, which is a little bit higher margin product for us within the portfolio. And then also some of the pricing increases, we were able to capture a little bit of margin upside on that just based on the timing of our inventory purchases relative to the timing of shipments and sales to customers.

Okay. So you kind of -- as you look out to the second half of the year, you see that catching up price increases compared to... .

Yes. So the price increase is slow, right? If that's an assumption that you use, if price increases are going to slow, then we would see -- we would expect to see less margin favorability from that because there'd be less of a timing difference between or less of a price variation between the timing between purchasing inventory and selling. But we have been using the balance sheet to try to secure inventory where we can. It's a tight market, so it's not unlimited supply. But where we can, we're using the balance sheet to try to gain a little bit of an advantage.

Okay. And maybe just as we're talking about memory as you get to the CXL systems, would you expect that's going to be a higher margin than the module business?

Yes, we do. It's really a solution. It's got software aspects to it, some good differentiation on the hardware as well. So I see that as a nice margin opportunity for us down the road.

At this time, there are no further questions. I will now hand the call over to Kash Shaikh, CEO, for closing remarks.

Thank you, operator. We see AI shifting towards inference with demand expanding beyond hyperscaler to enterprise, neo cloud and sovereign AI customers. We are still in our leadership in this transition, but the combination of our customer demand, product innovation and booking momentum gives us the confidence in the path ahead. We believe we are well positioned at the intersection of AI compute infrastructure and memory, and we are making good progress diversifying our customer base. My focus is on strong execution across product innovation, customer engagement and diversification, disciplined capital allocation and investment in our AI/HPC business to support the long-term growth. We look forward to updating you on our progress.

This concludes today's call. Thank you for attending. You may now disconnect.

Ladies and gentlemen, thank you for joining us, and welcome to the Penguin Solutions First Quarter Fiscal Year 2026 Financial Results Call. [Operator Instructions]

I will now hand the conference over to Suzanne Schmidt with Investor Relations. Please go ahead.

Thank you, operator. Good afternoon, and thank you for joining us on today's earnings conference call and webcast to discuss Penguin Solutions first quarter fiscal 2026 results.

On the call today are Mark Adams, Chief Executive Officer; and Nate Olmstead, Chief Financial Officer. You can find the accompanying slide presentation and press release for this call on the Investor Relations section of our website. We encourage you to go to the site throughout the quarter for the most current information on the company.

I would also like to remind everyone to read the note on the use of forward-looking statements that is included in the press release and the earnings call presentation. Please note that during this conference call, the company will make projections and forward-looking statements, including, but not limited to, statements about the company's growth trajectory and financial outlook, business plans and strategy, market demand and shifts, strategic agreements and existing and potential collaborations. Forward-looking statements are based on current beliefs and assumptions, are not guarantees of future performance and are subject to risks and uncertainties, including, without limitation, the risks and uncertainties reflected in the press release and the earnings call presentation filed today as well as in the company's most recent annual and quarterly reports. The forward-looking statements are representative only as of the date they are made, and except as required by applicable law, we assume no responsibility to publicly update or revise any forward-looking statements.

We will also discuss both GAAP and non-GAAP financial measures. Non-GAAP measures should not be considered in isolation from, as a substitute for or superior to our GAAP results. We encourage you to consider all measures when analyzing our performance. A reconciliation of the GAAP to non-GAAP measures is included in today's press release and accompanying slide presentation.

And with that, let me now turn the call over to Mark Adams, CEO. Mark?

Thank you, Suzanne. We hope you all have a nice holiday season and appreciate your attending our first quarter fiscal 2026 earnings call.

We are happy with our Q1 results. On our last call, we mentioned some headwinds we anticipated in the first half of our fiscal year. Despite these challenges, revenue came in at $343 million in Q1, up 2% sequentially and 1% year-over-year. We view this as significant as we were able to perform well in the first quarter despite not recognizing any hyperscale hardware revenue, which had been a meaningful contributor in the prior year period. Non-GAAP gross margins were 30%, which compares favorably to the midpoint of our full year outlook, reflecting favorable mix and execution in the quarter. As a reminder, our full year outlook incorporates expected variability across the year. Non-GAAP operating income was $42 million, up 1% year-over-year, which led to non-GAAP diluted earnings per share of $0.49.

We continue to see indications of a broader market shift from hyperscaler deployments and early corporate pilot programs toward wider enterprise adoption and more production scale implementations. Within this broader transition, there are early signs that some workloads are evolving from training-centric environments toward inference-oriented use cases as organizations operationalize AI across the enterprise.

As AI systems move into full production, enterprises are increasingly focused on performance, reliability, bandwidth and overall system efficiency, areas where Penguin's ability to design tailored systems can help customers address their specific workload requirements. We believe enterprises are looking for partners who can deliver complete production-ready platforms spanning infrastructure, software orchestration and advanced AI tooling supported with deep technical expertise.

Penguin Solutions brings over 25 years of experience in this arena, starting in high-performance computing, or HPC, and expanding in the last five years to include large-scale AI factory build-outs. This expertise enables us to design, build, deploy and manage complex infrastructure solutions that align with the needs of our enterprise customers.

During the transformation of our company into a leading provider of infrastructure solutions, we have communicated the challenges we faced from having a lumpy revenue model with customer concentration. Our fiscal year 2026 guidance was developed with our ongoing focus on new business development and customer diversification in mind, while recognizing that the timing of new customer deployments can vary. We are encouraged by the customer diversification progress we are making and we'll share additional detail later in the call.

Let me now speak to the performance of each of our lines of business. Our Advanced Computing business achieved revenue of $151 million, up 9% compared to last quarter. We had several customer bookings in Q1, including two new Penguin customers, one in the defense sector and another in the education and research sector. In addition, we continue to see our pipeline expand into new customer opportunities in the financial services, oil and gas, telecommunications, manufacturing and education sectors.

We are also currently engaged in a number of discussions with sovereign cloud customers outside the U.S. regarding potential large-scale AI deployments. This increase in sovereign AI opportunities internationally reinforces the need for Penguin's rapid deployment solutions and services, which enable faster time to production and thus enable a quicker return on AI capital investments.

As customers evaluate the transition from proof-of-concept initiatives to larger production deployments, we believe the complexity of these environments underscores the need for a trusted partner with deep technical expertise across multiple technology domains, combined with proven experiences integrating in an array of technology building blocks to deliver high-performing and highly reliable AI infrastructure. In support of this need, we have recently launched the Penguin Solutions rapid development workshop program, which brings together our architecture leaders, software developers, managed service team members and supply chain experts to help potential customers better understand and plan for the complexity of implementing AI at scale. Penguin's design, build, deploy and manage framework is a proven methodology for high-performing, high-reliability AI factory deployments.

Penguin's decades of experience in complex large data center installations is at the heart of what differentiates us from many of our hardware-centric competitors. The combination of our architectural design know-how, our ICE ClusterWare software platform and our managed services offerings reinforces Penguin Solutions role as a valued partner in helping our customers manage the complexity of AI.

Our Integrated Memory business recorded $137 million of revenue in Q1, up 3% compared to the prior quarter and up 41% compared to Q1 2025. As we head into our second quarter, demand signals for our memory portfolio are strong across our networking, telecommunications and computing customers.

One of the key performance challenges in implementing AI across both training and inference workloads is the limitation of GPU and CPU memory bandwidth. We have been an early developer of Compute Express Link or CXL solutions, which provide an open standard interconnect for high-speed, high-capacity GPU and CPU to device and GPU and CPU to memory connections designed for high-performance data center systems, including those built for AI computing. We are currently shipping early production units through our OEM partners while continuing to expand end-user qualification efforts. Looking ahead, future versions of CXL are expected to emphasize memory pooling, enabling blocks of memory to be dynamically allocated to specific system resources. In parallel, we continue to invest in the design of our optical memory appliance or OMA in collaboration with key technology partners, leveraging a photonic transport layer to further increase the performance of GPU, CPU and memory interconnects.

We are seeing increasing memory demand as customers are looking for unique custom solutions to address their needs in supporting AI workloads. Beyond the core customer migration to DDR5 technology, our customers are evaluating how to best utilize memory to optimize the performance of their AI compute needs. In addition to our legacy OEM customers, we are seeing sales growth in direct-to-enterprise customer engagements and in large hyperscale customers. We believe that our over 30 years of developing specialty memory products originally under the SMART Modular brand for large Fortune 500 customers has positioned us well to capitalize on a new wave of higher performing and higher reliability memory for the AI era. We believe we are well positioned for future growth, leveraging our early investments in new technology to expand our addressable market.

The Optimized LED business operating under the Cree LED brand generated revenue of $55 million in the first quarter, down 18% sequentially. We are seeing weak demand in our China business, along with pockets of softness among certain large U.S. OEM customers. We remain focused on driving profitability in our LED business by leveraging our specialty product portfolio, industry-leading intellectual property and capital-light outsourced front-end operating model. Despite the top line revenue headwind, operating income was $3.5 million, representing an increase of 24% sequentially.

As part of our transition from a holding company to an AI solutions provider, we continue to streamline our corporate structure. To that end, in late December, we signed an agreement to sell our remaining 19% stake in Zilia Technologies, formerly SMART Modular do Brazil for $46 million. We expect this transaction to close in our third quarter of fiscal 2026.

As we look forward, we remain focused on strengthening our partnerships with ecosystem partners such as NVIDIA, AMD and CDW. As the volume of corporate deployments accelerates, Penguin Solutions can provide our white glove design and implementation services to support our customer success. We are also customizing our ICE ClusterWare platform to be compatible with other open source industry software products, enabling a more robust infrastructure solution for managing large AI deployments at scale.

We believe that the combination of our investments in technologies such as inference systems level products, memory advancements and our ICE software, together with our trusted design and managed services capabilities, positions Penguin Solutions as an ideal partner to help customers manage the complexity associated with implementing AI infrastructure.

With a strong balance sheet, a growing customer base, continued investments in differentiated solutions and an experienced team that helps customers design, build, deploy and manage AI environments at scale, we remain confident in our position for long-term future success.

Let me stop now and hand the call to Nate for a more detailed review of our financial performance. Nate?

Thanks, Mark. I will focus my remarks on our non-GAAP results, which are reconciled to GAAP in our earnings release tables and in the Investor Relations materials available on our website.

Now let me turn to our first quarter results. In the quarter, total Penguin Solutions net sales were $343 million, up 1% year-over-year. Non-GAAP gross margin came in at 30%, which was down 0.8 percentage points versus Q1 last year. Non-GAAP operating margin was 12.1%, up 0.1 percentage points versus last year, and non-GAAP diluted earnings per share were $0.49, flat year-over-year.

In the first quarter of fiscal 2026, our overall services net sales totaled $65 million, down 9% versus the prior year. Product net sales were $279 million in the quarter, up 3% versus the prior year.

Net sales by business segment were as follows: in Advanced Computing, Q1 net sales were $151 million, which was 44% of total company net sales and down 15% year-over-year. This sales decline reflects both the wind down of our Penguin Edge business and hyperscale hardware sales in Q1 last year, which did not recur in Q1 this year. Q1 2026 advanced computing net sales, excluding Penguin Edge and hyperscale hardware net sales grew 52% year-over-year. In Integrated Memory, Q1 net sales were $137 million, which was 40% of total company net sales and up 41% year-over-year. And in Optimized LED, Q1 net sales were $55 million, which was 16% of total company net sales and down 18% year-over-year.

Non-GAAP gross margin for Penguin Solutions in the first quarter was 30%, down 0.8 percentage points year-over-year and 0.9 percentage points sequentially, primarily due to the wind down of our high-margin Penguin Edge business, as we described last quarter.

Non-GAAP operating expenses for the first quarter were $61 million, down 4% year-over-year and down 6% sequentially. Operating expenses as a percentage of net sales were down both year-over-year and quarter-over-quarter, driven by lower personnel-related expenses as well as lower subcontract services costs following the completion of our U.S. domestication in the fourth quarter of fiscal 2025. Q1 non-GAAP operating income was $42 million, up 1% year-over-year and up 6% versus last quarter. The combination of net sales growth and operating expense management translated into a 0.1 percentage point increase in non-GAAP operating margin versus Q1 last year. This is our sixth consecutive quarter of non-GAAP operating margin expansion year-over-year. Non-GAAP diluted earnings per share for the first quarter were $0.49, flat versus Q1 last year and up 14% versus the prior quarter. Adjusted EBITDA for the first quarter was $45 million, up 1% year-over-year.

Turning to balance sheet highlights. For working capital, our net accounts receivable totaled $342 million compared to $276 million a year ago, with the increase driven by higher sales volumes and variations in sales linearity across the quarters. Days sales outstanding came in at 51 days, up from 45 days in the prior year quarter and flat with last quarter. Inventory totaled $213 million at the end of the first quarter, down from $247 million a year ago due to order and shipment linearity. Days of inventory was 38 days, down from 49 days a year ago and down from 51 days last quarter, primarily due to the timing of receipts and shipments. Accounts payable were $305 million at the end of the quarter, up from $244 million a year ago due primarily to higher sales volumes and the timing of purchases and payments. Days payable outstanding was 55 days compared to 49 days last year and 54 days last quarter. The year-over-year and quarter-over-quarter movements were due to the timing of purchases and payments.

Our cash conversion cycle was 35 days, a decrease of 11 days compared to Q1 last year and down 14 days versus last quarter due to faster inventory turns resulting from materials shipped during the quarter and the timing of purchases and payments. Consistent with past practice, days sales outstanding, days payables outstanding and inventory days are calculated on a gross sales and gross cost of goods sold basis, which were $605 million and $509 million, respectively, in the first quarter. As a reminder, the difference between gross and net sales is primarily related to our memory businesses, logistics services, which are accounted for on an agent basis, meaning that we only recognize the net profit on logistics services as net sales.

Cash, cash equivalents and short-term investments totaled $461 million at the end of the first quarter, up $68 million from Q1 last year and up $8 million sequentially. The year-over-year fluctuation was primarily due to free cash flow generated by the business over the past year. First quarter cash flows provided by operating activities increased by 125% to $31 million compared to $14 million provided by operating activities in Q1 of last year. The increased cash flow in the quarter versus last year was due primarily to lower investments in net working capital.

We spent $15 million to repurchase approximately 791,000 shares in the first quarter under our stock repurchase program. As of November 28, 2025, an aggregate of $96.5 million remained available for the repurchase of our common stock under the current authorizations. For those of you tracking capital expenditures and depreciation, capital expenditures were $3 million in the quarter and depreciation was $5 million for the quarter.

And now turning to our outlook. Given our solid Q1 performance, we are pleased to confirm our full company net sales and non-GAAP diluted EPS outlook for the year, which at the midpoint calls for 6% net sales growth and $2 of non-GAAP diluted EPS. Consistent with the outlook we provided last quarter, our full year outlook assumes that we will continue to diversify our customer sales mix and does not include any advanced computing hardware sales to hyperscale customers. And also consistent with our assumptions from last quarter, our FY '26 financial outlook reflects the wind down of our high-margin Penguin Edge business. We expect sales from this business to essentially cease by the end of fiscal 2026. The combined effect of these two assumptions in our FY '26 outlook remains approximately a 14 percentage point unfavorable year-over-year impact to our total company net sales growth and approximately a 30 percentage point unfavorable impact to advanced computing.

Regarding sales linearity during the year and consistent with our commentary last quarter, we continue to expect second half sales to be stronger than first half sales. At the midpoint of our full year net sales outlook, we expect approximately 53% to 54% of total company net sales to come in the second half of the year as AI opportunities currently in our pipeline are assumed to book and ship in the second half.

With that said, our full year net sales outlook reflects the following full year growth ranges by segment. For Advanced Computing, we continue to expect full year net sales to change between minus 15% and plus 15% year-over-year. As it has previously, this outlook reflects the Penguin Edge and hyperscale hardware sales impacts mentioned earlier. For memory, we now expect net sales to grow between 20% and 35% year-over-year. And for LED, we now expect net sales to decline between minus 15% and minus 5% year-over-year.

Our non-GAAP gross margin outlook for the full year is now 29%, plus or minus 1 percentage point. We adjusted our gross margin outlook down by 50 basis points to account for a higher mix of memory sales, which have a lower gross margin than our company average. For non-GAAP operating expenses, we now expect a full year total of $250 million, plus or minus $10 million. For non-GAAP full year diluted earnings per share, we still expect approximately $2, plus or minus $0.25. Our FY '26 non-GAAP diluted share count is still expected to be approximately 55 million shares, and our FY '26 non-GAAP tax rate is still forecasted to be 22%. While we expect to use this normalized non-GAAP tax rate throughout FY '26 and beyond, the long-term non-GAAP tax rate may be subject to changes for a variety of reasons, including the rapidly evolving global and U.S. tax environment, significant changes in our geographic earnings mix or changes to our strategy or business operations.

Our outlook for fiscal year 2026 is based on the current environment, which contemplates, among other things, the global macroeconomic environment and ongoing supply chain constraints, especially as they relate to our Advanced Computing and Integrated Memory businesses. This includes extended lead times for certain components that are incorporated into our overall solutions impacting how quickly we can ramp existing and new customer projects and fulfill customer orders.

Overall, we believe our focused execution, disciplined expense management and balance sheet strength provide a strong foundation for sustained profitable growth. We expect these qualities to support our continued progress as we pursue opportunities to enhance long-term shareholder value. Please refer to the non-GAAP financial information section and the reconciliation of GAAP to non-GAAP measures tables in our earnings release and the investor materials on our website for further details.

With that, operator, we are ready for Q&A.

[Operator Instructions] Your first question comes from the line of Brian Chin with Stifel.

A few questions. Maybe just first with the maintaining the outlook, although changing the components a little bit. Firstly, first half -- fiscal first half versus fiscal second half guidance, I guess does that suggest that the February quarter revenue is down maybe low to mid-single digits? And kind of which of the segments is driving that sequential decline? And then also for the full year, raising the memory revenue growth to 20% to 35% year-over-year. But certainly, pricing should be a favorable tailwind. Curious if there are any kind of challenges or constraints shipping product given sort of how constrained some of the memory wafer supply is at the moment?

Brian, thanks for the question. I'll take the first end of it, and then I'll hand it over to Nate.

There's two implications to the kind of forward-looking assumptions. To speak to the memory one, we continue to do a fairly good job navigating the supply constraints you're alluding to, and that allowed us to think through kind of how that business should perform going forward. And so we haven't seen anything material impacting that business and which allowed us to kind of get more granularity on the forward-looking projection, so to speak. On the Advanced Computing piece, and we'll talk about some of this through the call. But when we're winning these new customers, there's a process of kind of getting the award. And then since they're new customers, in many cases, we have to negotiate a master agreement, and then from there that gets converted into a purchase order. There's just more timing involved in new customers.

Now the exciting news is we are winning more new customers, but the predictability gets a little bit tougher for us in a given quarter, so to speak. And that's where, I guess, we're a little bit more cautious. I would say just -- we talked about an award with a major financial institution last quarter. I'm happy to announce that we've signed that agreement, and we expect the PO here shortly this week or worst case next. I can also tell you that we have the same -- we have a major new oil and gas customer that we've also signed a master agreement with and expecting a PO in here shortly as in the next week or so.

And so as we go through that, the timing of then getting the POs in the system and then going out and sourcing components and staging the equipment, all of that becomes just tougher for a first-time customer in the model. And that's where you may sense a little bit of caution as we think about our Q2 and even back half guidance to when things will hit. Nate?

Yes. I mean I would say when you think about the full year outlook, obviously, the most significant changes would be memory looking a little bit stronger, as you alluded to, due to some benefits from pricing primarily, although the overall demand remains very strong and healthy there. But our ability to procure supply is going to be one of the key variables to think about the memory outlook for the full year. And then LED looks like it's going to be weaker. That was fairly broad-based across the regions in Q1. Q2 is always a little bit softer in LED because of Chinese New Year. So I'd expect some sequential pressure probably in the LED business.

I also expect sequentially advanced computing to be down Q1 to Q2, but that's what we expected when we laid things out last quarter. I think Q1 came in right about where we thought. Q2 looks similar to what we thought. And in the back half of the year, we expect some strength due to the -- some of the opportunities we see in the pipeline and bookings that are starting to shape up.

Great. Maybe just for a quick follow-up. In terms of the pipeline in advanced computing, can you maybe elaborate on how some of your expansive channel partnerships with Dell, CDW, you've got sort of the strategic relationship with SKT. I think you alluded to maybe future cloud opportunities on a sovereign basis, and maybe that's contributing to the potentials for fiscal second half. Maybe can you kind of expand on how some of those partnerships are helping with the pipeline?

Yes. I think across the board, most of what you talked about have been fairly exciting for us to engage with and representing a much stronger pipeline for us.

I'll try to break it down. In terms of CDW, you had mentioned, our capabilities in kind of the AI factory environment and large-scale deployments is a great addition to CDW's capabilities and competencies. And we've got some good customer opportunities that just in our managed services and software solution set that we can bring to customers that are of large scale. And they're definitely represented in our pipeline.

If you contemplate our relationship with NVIDIA, it continues to get stronger and stronger as the enterprise deployments scale and grow in number, being a services solutions provider for NVIDIA-based AI factories, Penguin is very well positioned there, and we continue to strengthen that partnership. and evaluating how we can bring the best of what we do with NVIDIA to form a strategic solution offering for our customers long term. And so the engagements with NVIDIA continue to strengthen.

SKT is an exciting partnership. We've talked about on prior calls. There continues to be opportunities with them both in Korea and outside of Korea. And then we've mentioned in my prepared remarks, we've had a couple of -- a few new sovereign cloud opportunities of large scale that are in the pipeline and very exciting just because given the raw investment that is going to go into each of these deployments, we stand to benefit from our involvement. And hopefully, we'll be able to update you on future calls here.

Your next question comes from the line of Samik Chatterjee with JPMorgan.

This is MP on behalf of Samik Chatterjee. So, firstly, I wanted to ask about the enterprise engagements, like you have been talking about the shift from hyperscalers towards then enterprises deploying for pilot programs and then towards broader enterprise-wide applications. Can you please help us understand like what exactly are you seeing, which is helping you see clearly mark that trend out?

And then other than that, I wanted you to double-click a bit on your diversification efforts.

Sure. Let me start with your first question. If I look back over the last three or four years, most of the capital expenditure dollars of massive -- of large-scale deployments was in the area of large language model training at large hyperscalers for the most part, okay? I don't want to be universal in saying 100%, but a majority of the spend that we saw was in a very consolidated set of customers. And if you want to triangulate that data with what's going on in the market, look at where NVIDIA was selling their GPUs as an example, you'll see that their major customers were consolidated to a few large hyperscale type environments. We saw the same thing.

I'd say over the last 6 to 12 months, we've seen the beginning of an evolution where enterprise opportunities are accelerating in terms of just raw volume of enterprise opportunities. And I would say the capital behind that and the planning for future growth and expansion in our customer relationships in the enterprise back that up. And so it's really been an evolution, a shift from early-stage large language model training to corporate enterprise rollout. And just based on our own pipeline activity, but also just raw market data in terms of where the products are going, we're fairly bullish on the enterprise environment as well we are on these larger sovereign AI deals. The combination of that makes us feel pretty strong on our pipeline development and diversification efforts.

Your next question comes from the line of Matthew Calitri with Needham & Company LLC.

This is Matt Calitri over at Needham. Last quarter, you noted an inventory increase to support shipments at the start of 1Q FY '26. And while inventory declined sequentially, it still remains elevated. How should we think about inventory levels as a leading indicator for future shipments? And how is your visibility into the remainder of the year?

Yes, you're right. Last quarter, we exited the quarter with some inventory. That was both in memory, where the price increases, when the prices go up, you're going to see that reflected in inventory. The cost of the goods goes up. And we also had some shipments in advanced computing, which shipped early this past quarter in Q1.

So I think inventory being higher than where it was a year ago is not surprising given that the overall business is larger, especially in memory. It was up 41% year-over-year. But if you look at the inventory days or the inventory turns, they're in a very healthy position. So certainly no concerns there. returning inventory quickly. Our business model is not one where we're buying really ahead of orders. We're buying to orders rather than to forecast generally.

Now in today's constrained memory environment, we'll look for opportunities where we can secure some supply to take some risk off the table, and we have a strong balance sheet that we intend to put to use if the opportunity is out there for us to do that.

Got it. Very helpful.

On the question the back half of the year, I think Mark talked already about some of the opportunities in the pipeline. It remains consistent with what we mentioned last quarter. We expect the second half of the year to be stronger in the first half. That will be especially true in advanced computing, perhaps in memory as well, again, contingent upon us being able to secure supply. But that's consistent with what we said last quarter.

Okay. Great. And then can you expand a little bit upon -- I think in the prepared remarks, you mentioned some work being done to customize ICE to be compatible with other open source platforms. What exactly are you guys working on there?

As we think about the software stack for AI factory rollouts, there's different layers of software. Like, for example, there's cluster management layers, there's security layers, there's orchestration layers. And so what we're trying to do is build a stand-alone stack, which is partly our ICE platform and then partly best-of-breed open software stacking components that allow us to have a unique solution and that we can manage all of it as opposed to our customers having to go out and pick pieces to it. And we're working with customers to define what that might look like in a Penguin stack that's partially our own developed ICE platform and partially best-in-class third-party software that gives the customer the best software features. And by the way, that includes working with companies like NVIDIA to have their software as part of a customized platform for future development deployments.

[Operator Instructions] Your next question comes from the line of Maddie De Paola with Rosenblatt Securities.

This is Maddie calling on behalf of Kevin Cassidy. I was just wondering, given the recent Marvell acquisition of Celestial AI and just a broader shift towards optical fabrics, are you seeing any change in optical memory and related technologies?

I wouldn't say we're seeing any changes. I think it's a strong validation of the market opportunity, broad macro opportunity. People are definitely looking at this dynamic of enhancing the bandwidth performance between memory and GPU/CPUs. So when I think about an established company like Marvell making such an investment that's publicly been announced, I -- it makes me feel good about the direction and the strategy that we're deploying here in developing that type of system-level product in memory.

There are no further questions at this time. I will now hand it back to Mark Adams, CEO, for closing remarks.

Thank you, operator. I would like to thank our worldwide employees for their dedication and commitment. Our Q1 results reinforce that we are on the right path, and we continue to grow our pipeline of new opportunities, helping our valued customers manage the complexity of their AI infrastructure. Thank you all for joining today's call.

This concludes today's call. Thank you for attending. You may now disconnect.

Good afternoon, thank you for attending today's Penguin Solutions Fourth Quarter and Full Year 2025 Financial Results. My name is Victoria, and I'll be your moderator today. [Operator Instructions].

I would now like to pass the conference over to our host, Suzanne Schmidt, thank you. You may proceed, Suzanne.

Thank you, operator. Good afternoon, and thank you for joining us on today's earnings conference call and webcast to discuss Penguin Solutions' Fourth Quarter and Full Year Fiscal 2025 results.

On the call today are Mark Adams, Chief Executive Officer; and Nate Olmstead, Chief Financial Officer. You can find the accompanying slide presentation and press release for this call on the Investor Relations section of our website. We encourage you to go to the site throughout the quarter for the most current information on the company. I would also like to remind everyone to read the note on the use of forward-looking statements that is included in the press release and the earnings call presentation.

Please note that during this conference call, the company will make projections and forward-looking statements, including, but not limited to statements about the company's growth trajectory and financial outlook, business, plans and strategy and existing and potential collaborations.

Forward-looking statements are based on current beliefs and assumptions and are not guarantees of future performance and are subject to risks and uncertainties, including, without limitation, the risks and uncertainties reflected in the press release and the earnings call presentation filed today as well as in the company's most recent annual and quarterly reports. The forward-looking statements are representative only as of the date they are made and except as required by applicable law, we assume no responsibility to publicly update or revise any forward-looking statements.

We will also discuss both GAAP and non-GAAP financial measures. Non-GAAP measures should not be considered in isolation from, as a substitute for or superior to our GAAP results. We encourage you to consider all measures when analyzing our performance. A reconciliation of the GAAP to non-GAAP measures is included in today's press release and accompanying slide presentation.

And with that, let me turn the call over to Mark Adams, CEO. Mark?

Thank you, Suzanne. Welcome to Penguin Solutions Fourth Quarter and Fiscal Year 2025 Earnings Call. Today, I'll walk through both our Q4 and full year results, providing updates on each of our business segments and sharing how we are positioning the company for long-term growth.

Fiscal 2025, the transformational year for Penguin Solutions as we continue to evolve from a holding company structure into a leading provider of AI infrastructure solutions. In addition to delivering 17% top line growth, 190 basis points of non-GAAP operating margin expansion and a 53% increase in non-GAAP diluted EPS, we made meaningful progress positioning the company for long-term success.

Key accomplishments included expanding our advanced computing pipeline and adding several new customers to the Penguin Solutions franchise, supporting our ongoing customer diversification strategy, deploying our first international AI infrastructure implementation, developing and expanding key partnerships with NVIDIA, CDW, Insight, and Dell; rebranding the company as Penguin Solutions.

Moving our corporate domicile to the United States; closing a $200 million investment from SK Telecom; refinancing our debt to strengthen the balance sheet; and strengthening our leadership team with the appointments of Tony Fry formerly of NetApp as our SVP and Chief Revenue Officer; and Ted Gillick, formerly of Dell, as our SVP of Strategy and Corporate Development.

I'll now turn to our financial results for the fourth quarter and full year. We delivered solid fourth quarter financial results. Revenue for Q4 was $338 million, an increase of 9% year-over-year. Non-GAAP gross margin was 30.9%. Non-GAAP operating income reached $39 million, a 16% increase year-over-year with non-GAAP operating margin at 11.6% up 80 basis points. Non-GAAP diluted earnings per share was $0.43, up 18% from the prior year.

Our fourth quarter performance rounded out a strong fiscal year. For the full year, revenue grew 17% year-over-year. Non-GAAP gross margin remained steady at 31% and non-GAAP operating margin improved by 190 basis points to 12.2%. Non-GAAP diluted EPS was $1.90, an increase of 53% compared to fiscal 2024.

We continue to see signs of broad AI adoption particularly within verticals such as financial services, energy, federal and education. As we've noted previously, our expectation has been that the AI pilot systems deployed across industries in 2023 and 2024, could lay the groundwork for broader production scale rollouts in the years ahead. We're now beginning to see indications of this transition with early-stage corporate build-outs taking shape.

Notably, Penguin Solutions was recently selected by a Tier 1 U.S. financial institution to manage an AI infrastructure deployment, the institution's first on-premise GenAI data center implementation. We believe this win reflects the growing confidence enterprises have in our ability to deliver scalable high-performance infrastructure, and we look forward to formalizing this engagement in due course.

At Penguin Solutions, we support customers in navigating the complexity of AI adoption by combining deep technical expertise in advanced cluster implementations with a comprehensive portfolio of hardware, software and managed services. We collaborate with customers to design, deploy and operate these environments with a focus on time to revenue performance, availability and long-term reliability. Our solutions are primarily targeted at deployment serving Fortune 500 enterprises, educational institutions, government entities and systems integrators and neocloud service providers.

While our go-to-market strategy has traditionally focused on direct customer relationships, we are actively investing in forming strategic partnerships that we believe can expand our reach and open new long-term growth opportunities.

Our foundation is built on over 25 years of experience in large-scale deployments beginning with our roots in high-performance computing or HPC. We believe this expertise gives us an advantage in integrating complex AI building blocks such as power and cooling systems, compute, memory, storage and networking. This foundation enables us to deliver the kind of infrastructure that today's enterprise customers need to drive innovation, scale efficiently and stay ahead of the curve in enterprise AI.

I'd like to provide additional detail on our business segments. Advanced computing revenue for the fourth quarter of fiscal 2025 was $138 million, up 4% sequentially from Q3. For the full fiscal year, advanced computing revenue reached $648 million, reflecting 17% year-over-year growth.

Within advanced computing, our HPC AI revenue from non-hyperscalers increased 75% year-over-year, highlighting progress on our customer diversification strategy. Since our last call, we completed the design, build and deployment of Haien, one of South Korea’'s largest and most powerful GPU as a service systems developed by SK Telecom for the Korea Ministry of Science and Technology as a key element in the country's sovereign AI initiative.

We now manage the system on an ongoing basis. We also launched several new AI projects, including with a Fortune Global 500 multinational consumer products company, Fortune 100 federal systems integrator and a Fortune 50 financial institution. This last project is in addition to the win from a different Tier 1 financial institution that I mentioned earlier.

We're also pleased with the growth in our new customer opportunities. In fiscal 2025, pipeline, bookings and revenue from new customers grew nicely, reflecting the ongoing demand for our expertise and solutions. As we've noted on prior calls, revenue in this segment can be lumpy with large deployments in one period, not necessarily reoccurring in the next. That said, we remain encouraged by the level of customer interest in our solutions and the long-term growth potential of these relationships. Our core competency in successfully managing large-scale AI infrastructure build-outs helps customers accelerate their time to a live production environment. We believe our customers value our technology-agnostic approach, which allows us to create a unique overall solution that meets their specific AI infrastructure needs.

Beyond our hardware building blocks, we continue to invest in our Penguin ICE ClusterWare software, a platform that helps customers manage infrastructure assets. Post deployment, our Penguin Solution services organization can provide ongoing operational support to sustain the high performance and high availability of their systems.

Our Integrated Memory segment with sales products under the SMART Modular brand, delivered $132 million in revenue for the fourth quarter and $464 million in revenue for the full fiscal year representing a 30% increase compared to fiscal 2024. This growth was driven by strong demand from customers in computing, networking and telecommunications. We are optimistic about our memory demand in the near term as large enterprises seek out higher performance and reliability memory solutions to support both traditional use cases and increasingly complex AI application.

In line with this demand, we are seeing promising early interest in our Compute Express Link or CXL family of products. As customer qualification efforts continue to expand, we believe we are well positioned for growth as adoption of CXL scales.

In memory, our R&D investments are focused on next-generation technologies. One key area is memory pooling, which has the potential to significantly expand bandwidth and improve memory capacities in GPU environments. We continue to invest in the design of SMART'S optical memory appliance or OMA, with initial product shipments targeted for late calendar 2026 to early 2027. This new offering will be designed to enable memory scaling of the industry's fastest high-bandwidth memory or HBM, which is today a limiting factor in AI cluster performance and efficiency. With a strong backlog, ongoing technology transitions such as DDR4 to DDR5 and a road map that includes innovations like CXL and OMA, we believe memory will be an important growth engine for Penguin Solutions.

Optimize LED under the Cree LED brand. Cree's fourth quarter revenue came in at $67 million, an increase of 9% compared to the prior quarter. Full year revenue was $256 million, roughly flat year-over-year reflecting a combination of secular and macroeconomic headwinds in the LED market. Despite these challenges, we achieved a 250 basis point improvement in non-GAAP operating margin in fiscal year 2025. As we move into fiscal 2026, Cree is focused on capturing market share, operating efficiently, protecting our intellectual property and driving operating profit growth.

Fiscal 2025 was a defining year for Penguin Solutions. We took meaningful steps to transform the company and sharpen our focus on becoming a leading provider of AI infrastructure solutions. We believe that these efforts have strengthened our foundation and positioned us well for long-term success.

Looking ahead to fiscal 2026, we believe we can sustain our growth momentum with the following strategic priorities; growing our enterprise customer base in AI infrastructure deployments; driving innovation across our hardware, software and services portfolio, to create sustainable differentiation; expanding our strategic partnerships to enhance Penguin's go-to-market efforts; operating with discipline and efficiency to position the company for long-term success; and further strengthening our balance sheet to support investments in scale and new capabilities. In setting these priorities, our intention was to align with the long-term interest of our shareholders, customers and employees.

In closing, I want to thank our global team for the dedication and performance in FY 2025. We delivered top line growth, improved profitability, strengthened our balance sheet and expanded the Penguin Solutions customer base. We believe we are well positioned for future success in FY '26 and beyond.

Let me stop and hand it over to Nate for a detailed review of 2025 financials and our outlook for FY 2026.

Thanks, Mark. I will focus my remarks on our non-GAAP results, which are reconciled to GAAP in our earnings release tables and in the investor materials available on our website.

Now let me turn to our fourth quarter and full year results. In the quarter, total Penguin Solutions net sales were $338 million, up 9% year-over-year. Non-GAAP gross margin came in at 30.9%, which was flat year-over-year. Non-GAAP operating margin was 11.6%, up 0.8 percentage points versus last year, and non-GAAP diluted earnings per share were $0.43, up 18% from last year.

For the full fiscal year 2025, total company net sales were $1.37 billion, up 17% year-over-year and aligned with the outlook we initially provided in April and better than the outlook we provided at the start of the fiscal year. Full year non-GAAP EPS was $1.90, up 53% versus the prior year and better than the increased outlook we provided last quarter.

In the fourth quarter, our overall services net sales totaled $63 million, up 5% versus the prior year. Product net sales were $275 million in the quarter, up 9% versus the prior year. Net sales by business segment were as follows; in advanced computing, Q4 net sales were $138 million, which was 41% of our total net sales and down 7% year-over-year. For the full year, advanced computing delivered $648 million of net sales or 47% of total company net sales and up 17% year-over-year. Our strong full year advanced computing growth was driven by our HPC and AI business, which grew 34%. Notably, within our HPC AI business, product and services sales to our non-hyperscale customers were up 75% for the full fiscal year.

For Integrated Memory, in Q4, net sales were $132 million, which was 39% of total company net sales and up 38% year-over-year. For the full year, memory net sales totaled $464 million or 34% of total net sales and up 30% year-over-year. And in Optimized LED, net sales were $67 million or 20% of total company net sales and up 2% year-over-year. For the full year, LED delivered $256 million of net sales or 19% of total company and down 1% versus the prior year.

Non-GAAP gross margin for Penguin Solutions in the fourth quarter was 30.9%, flat year-over-year with margin pressure from a higher mix of integrated memory net sales offset by improved margin rate across all 3 business segments. Non-GAAP gross margin was down 0.8 percentage points sequentially with lower margin rates in advanced computing, partially offset by higher margin rates in both Integrated Memory and Optimized LED. For the full fiscal year, gross margins were 31%, in line with our prior outlook and down 0.9 percentage points year-over-year due to growth in our memory and AI hardware businesses which have lower than company average margins, but are addressing fast-growing market opportunities.

Non-GAAP operating expenses for the fourth quarter were $65 million, up 5% year-over-year and up 1% sequentially. Operating expenses as a percentage of net sales were down both year-over-year and quarter-over-quarter, driven by higher net sales volumes and modest spending increases. For the full fiscal year, non-GAAP operating expenses were $257 million, up 1% year-over-year and down 2.9 percentage points as a percent of net sales due primarily to strong top line growth and disciplined expense management.

Q4 non-GAAP operating income was $39 million, up 16% year-over-year and up 2% versus last quarter. The combination of net sales growth and operating expense management translated into a 0.8 percentage point increase in operating margin versus Q4 last year. This is our fifth consecutive quarter of non-GAAP operating margin expansion year-over-year.

For the full fiscal year, non-GAAP operating income was $168 million, up 39% year-over-year and non-GAAP operating margin improved 1.9 percentage points to 12.2% of net sales. Non-GAAP diluted earnings per share for the fourth quarter were $0.43, up 18% versus the prior year.

For the full year, non-GAAP diluted EPS was $1.90 up 53% versus the prior year and $0.05 better than the high end of our outlook provided in July. Adjusted EBITDA for the fourth quarter was $43 million, up 11% year-over-year, and for the full year was $187 million, up 28% versus the prior year.

Turning to balance sheet highlights. For working capital, our net accounts receivable totaled $308 million compared to $252 million a year ago, with the increase driven by higher sales volumes and variations in sales linearity across the quarters. Days sales outstanding came in at 51 days, up from 49 days in the prior year quarter. Inventory totaled $255 million at the end of the fourth quarter, up from $151 million at the end of last year due to higher sales volumes and order linearity. Days of inventory was 51 days up from 36 days a year ago, primarily due to the positioning of inventory for shipment early in Q1 FY '26.

Accounts payable were $267 million at the end of the quarter, up from $182 million a year ago due primarily to higher sales volumes and the timing of purchases and payments. Days payable outstanding was 54 days compared to 43 days last year due to the timing of purchases and payments. Our cash conversion cycle was 49 days, an increase of 7 days compared to last year due to slower inventory turns resulting from materials positioned for shipment early next quarter. Consistent with past practice, days sales outstanding, days payables outstanding and inventory days are calculated on a gross sales and gross cost of goods sold basis, which were $550 million and $453 million, respectively, in the fourth quarter. As a reminder, the difference between gross and net sales is primarily related to our memory businesses logistics services which are accounted for on an agent basis, meaning that we only recognize the net profit on logistics services as net sales.

Cash, cash equivalents and short-term investments totaled $454 million at the end of the fourth quarter, up $64 million from the prior year and down $282 million sequentially. The year-over-year fluctuation was due primarily to proceeds from the issuance of preferred shares, cash generated by the business and the repayment of our term loan in Q4. The sequential decline was primarily driven by the repayment of our term loan.

Fourth quarter cash flows used by operating activities from continuing operations totaled $70 million compared to $12 million used by operating activities from continuing operations in the prior year quarter. The increased use of cash in the quarter versus last year was due primarily to investments in inventory to support shipments at the start of Q1 FY '26. For the full fiscal year 2025, operating cash flow from continuing operations was $113 million, an increase of 8% versus the prior fiscal year.

We spent approximately $296,000 to repurchase 16,000 shares in the fourth quarter under our stock repurchase program. Since our initial stock repurchase authorization in April 2022, we have used a total of $113 million to repurchase 6.6 million shares through 2025. Earlier today, we announced that our Board has authorized a $75 million increase in our stock repurchase authorization, bringing our total remaining authorization to $112 million.

As mentioned in our Q3 earnings call, in Q4, we completed a refinancing of our existing credit facility. We paid off the $300 million remaining on our term loan using $200 million of cash from our balance sheet and $100 million of borrowings from a new revolving credit facility. This refinancing transaction significantly reduced our leverage, extended our debt maturities and is expected to lower our debt service costs as we reduced our total gross debt by $200 million. Our net debt at the end of the fiscal year was $16 million.

For those of you tracking capital expenditures and depreciation, capital expenditures were $3 million in the fourth quarter and $9 million for the full year, and depreciation was $5 million for the quarter and $21 million for the full year.

And now turning to our outlook. Coming off a strong fiscal year '25, we believe that our strategy and execution capabilities position us well for long-term profitable growth. For fiscal '26, we are initiating an outlook for net sales to grow 6%, plus or minus 10% versus the prior year. There are a few important assumptions to keep in mind with regard to this outlook.

First, as previously disclosed in our annual and quarterly filings, we are in the process of winding down our Penguin Edge business, which is part of our Advanced Computing segment. We expect sales from these Penguin Edge products to essentially cease at the end of this calendar year and have included this assumption in our outlook. While this will result in the phaseout of some profitable business, Penguin Edge has become a smaller portion of our overall portfolio and in the long-term, we do not expect a material impact to our growth trajectory.

Second, we believe that we will continue to diversify our customer sales mix and we have assumed zero hardware sales in FY '26 to hyperscale customers as we don't currently have line of sight to such business in this fiscal year. To be clear and importantly, we do expect our hyperscale services business to continue in FY '26, and those sales are included in our outlook.

The combined effect of these 2 assumptions in our FY '26 outlook is a 14 percentage point unfavorable year-over-year impact to our total company net sales growth.

Last, you may notice that the net sales growth range in our outlook is wider than last year. While we entered this year with a stronger pipeline of AI compute opportunities than last year, we expect our sales volumes to be higher in the second half of the year than in the first half. You will recall that in FY '25, the opposite was true as we had a strong first half of hardware shipments to our large hyperscale customer. That shipment timing led to approximately 52% of our total company sales coming in the first half of fiscal 2025. By comparison for fiscal '26, the midpoint of our outlook assumes approximately 46% of our sales come in the first half of the year. So with our growing base of AI compute opportunities and our expectation of a more back-end loaded year, we felt a wider net sales outlook range was prudent to reflect a broader set of potential outcomes.

With that said, our full year net sales outlook reflects the following by segment; for advanced computing, we expect full year net sales to change between minus 15% and plus 15% year-over-year. This outlook includes the Penguin Edge and hyperscale hardware sales impact mentioned earlier. For Memory, we expect net sales to grow between 10% and 20% year-over-year.

And for LED, we expect net sales change between minus 5% and plus 5% year-over-year. Our non-GAAP gross margin outlook for the full year is 29.5%, plus or minus 1 percentage point. The decline in gross margin outlook versus FY '25 is primarily due to the wind down of the high-margin Penguin Edge business as well as growth in lower-margin businesses, such as Memory and AI hardware. New AI customer wins typically begin with upfront hardware net sales at lower margin during the implementation phase, and we aim to follow those engagements with higher-margin recurring software and services sales. As a result, we anticipate some near-term gross margin pressure as we engage in initial infrastructure deployments, but we view these upfront investments as an important foundation for durable high-margin growth over time.

For non-GAAP operating expenses, we expect a full year total of $255 million, plus or minus $10 million. For non-GAAP full year diluted earnings per share, we expect approximately $2 plus or minus $0.25. Our FY '26 non-GAAP diluted share count is expected to be approximately 55 million shares.

Due primarily to changes in the geographic mix of our earnings and benefits from our recently completed U.S. redomiciliation, we are lowering our FY '26 and long-term non-GAAP tax rate to 22% which reflects currently available information. While we expect to use this normalized non-GAAP tax rate throughout FY '26 and beyond, the long-term non-GAAP tax rate may be subject to changes for a variety of reasons, including the rapidly evolving global and U.S. tax environment, significant changes in our geographic earnings mix or changes to our strategy or business operations.

Our outlook for fiscal year 2026 is based on the current environment, which contemplates, among other things, the global macroeconomic environment and ongoing supply chain constraints especially as they relate to our advanced computing and Optimized LED businesses. This includes extended lead times for certain components that are incorporated into our overall solutions impacting how quickly we can ramp existing and new customer projects.

Overall, we believe our focused execution, disciplined expense management and balance sheet strength provide a strong foundation for sustained profitable growth. We expect these qualities to support our continued progress as we pursue opportunities to enhance long-term shareholder value. Please refer to the non-GAAP financial information section and the reconciliation of GAAP to non-GAAP measures tables in our earnings release and the investor materials on our website for further details.

With that, operator, we are ready for Q&A.

[Operator Instructions] Our first question comes from the line of Kevin Cassidy with Rosenblatt Securities.

Thanks for letting me ask a question and congratulations on the strong fiscal year '25. Thank you for the information about your hyperscale customer. But can you say, is there a project over? And would you say that going forward, we should just take them out of the forecast or should -- are you still active in potential -- more hardware deployments?

Kevin, thanks for the question. We don't look at it like the project's over. We have ongoing services with the customer and still are in discussions for future development opportunities, it's just that in our outlook for the year, fiscal 2026, we don't anticipate any non-service revenue or systems hardware revenue in the year. I'll let Nate comment.

Yes, that's right. I think last year, we entered the year with some visibility to some hardware shipments from the hyperscalers and just wasn't the case this year. So coming into this year, we thought it made sense to just make the assumption that we wouldn't have any hardware revenue from hyperscalers this year. But as Mark said, we continue to have a very good relationship with them and the services revenues continue.

Okay. Great. And also exciting news with the SK Telecom and landing one client with that. And there was a lot of news last week about SK Telecom even being awarded with some business with OpenAI in Korea. Is there any participation availability for Penguin in that relationship?

That's something we can't address or talk to today, Kevin. What I would say is the referenced implementation in our prerecorded material was a great opportunity for the company. It's our first international deployment of significant scale and we went from order to go live in just about 2 months, and it was a significant validation of our capabilities, our rapid deployment framework. And we really value the relationship with SK Telecom, but we're not able to comment on any future opportunities at this time.

Kevin, I'll just add to that. Last quarter, you saw in our filings when we initially booked the deal, it was for hardware. Since then, we've also added services as well. So it's great to see that as part of that transaction and the relationship with SK being strong.

Okay. And maybe just to add on to that a little bit. In the filing, I think it was $34.6 million. Is that -- will that be recognized over the next couple of quarters? Or when does that get recognized?

So we recognize that portion in Q4, but we'll have more in FY '26.

Our next question comes from the line of Michael Ng with Goldman Sachs.

I wanted to just ask about the Penguin Edge and combined impact with hardware and revenue for next year. I think you mentioned a 14 percentage point headwind to revenue growth to the total company. When I do the math, I think that implies a 28 percentage point impact to Advanced Computing, is that the right way to think about it? How would you think about Advanced Computing growth next year, kind of ex these items? And could you just remind us why it kind of strategically makes sense to exit the Penguin Edge business, which I think was about 10% of the segment, but would love to hear your general thoughts there.

Thanks for the question, Michael. I'm going to answer the last part of your question first and then hand it over to Nate to talk about the financial impact and your question on impact of the other elements.

It really wasn't a decision that we could stay or not stay. We had 2 clients in our Penguin Edge business that made up a significant part of that business. And as we announced in our filings, we were going through some last time buys and we're winding down the business. And over time, we were just getting better visibility towards the end of our fiscal '25 to what the impact would be in '26. So it wasn't really a choice of should we stay in it strategically or not, it was 2 large customers that we're winding down on a prior generation of a product and that they were not renewing.

Mike. And so I think on your math, you're roughly right. Advanced Computing is about 47% or 50% of total company sales. So that 14 points all sits within Advanced Computing, which means it's about 28% to 30% of Advanced Computing revenue from FY '25.

Great. And if I could just follow up, please. So if the underlying is growing closer to the 30% to 45%, could you just maybe talk about the key areas of momentum for advanced computing that you're seeing for next year, is it just more of those customer wins that you talked about converting into revenue? And maybe just kind of expand a little bit more in terms of the wideness of the range. I know you had made some comments earlier during the call.

Sure. Let me take a shot at that one. When you think about our model, and we've talked about customer diversification for some time, we've done a pretty good job at the launching of going to market resources into a broader set of customers. And those target enterprise customers plus federal opportunities and those in the education sector. When you combine that into a target set of customers for us, we engage with these customers. We bring them the value proposition we have, we try to identify funded projects that would allow us to utilize our capabilities and helping accelerate our customers' AI implementation. Once we get that opportunity to bid on a proposal, or a request for proposal. We delivered that, and that starts the beginning of a pipeline opportunity, and our pipeline is growing.

Now pipeline is not a booking, as we all know. But the pipeline opportunities are growing as our, just the number of customers that we're engaged with. And so as you talk about the offset to some of these headwinds, we're continuing to add customers to the franchise and some really notable global brands in their respective industries. And we're excited about the direction we're heading, and we're looking to convert those pipeline opportunities into bookings and eventually revenue.

Yes. I mean I think just building on what Mark said, I think it's part of our diversification of our customer base is what you're seeing reflected in the outlook. And as I mentioned in my prepared remarks, too, 75% growth in the HPC, AI business from non-hyperscalers, I think it's just a really positive that at point for us from FY '25, and we think that we can continue to grow a very fast clip in those customers in FY '26.

Our next question comes from the line of Samik Chatterjee with JPMorgan.

I have a couple, and maybe I'll sort of give you both at the same time. You did mention a better second half compared to the first half. And maybe just if you can dive in towards giving you that visibility with some of your customers? And is it really more coming from Advanced Computing visibility? Or is it more memory-driven?

And then just specific to Memory, when you are guiding to 10% to 20% growth. Just curious how much of that is maybe somewhat pricing driven, given sort of what's happened with the underlying commodities here? Or -- and what are the margin implications of what you're seeing on the commodity front -- you feed into your sort of overall systems on the Memory side?

Sure. Let me take the Memory one first. So you're right, obviously, prices are starting to increase. And that affects a portion of our portfolio. It's not all of it. Important to keep in mind that we generally operate in Memory on a value-add basis, right? And so as Memory prices increase, we can generally pass along those price increases, but we don't get additional margin from that. So you would see an increase in revenue, but you would see a decrease in margin rate or really no increase in profit dollars because we operate on this value-add basis.

I would say in terms of the outlook, listen, I think I put a little bit of price increase probably into the high end of the range for Memory, but not a lot. We'll see how things play out there. I think currently, we feel good about the backlog that we've been able to build from Memory going into Q1. And we have pretty good visibility, I'd say, into the first half on Memory. You mentioned the second half being stronger than the first half in our outlook. And mostly, that reflects the AI business. Mark mentioned about the strength of the pipeline. And we haven't converted those into bookings yet. And so the outlook really reflects that, where we have a strong pipeline and some good opportunities, some of which we expect to convert into bookings, but they will be booked later in the first half into the second half and expect revenue in the second half of the year.

Our next question comes from the line of Ananda Baruah with Loop Capital Markets.

Yes. I have a couple, if I could. So just going back on the apples-to-apples revenue growth sort of when you back out -- back out the meta hardware and the Penguin Solutions -- or sorry, the Edge business. So is that to say, if it's a 14% impact year-over-year growth, is that to say that apples-to-apples, like on a pro forma basis, you'd really be guiding 20% growth? Should I think about it that way or am I totally off?

Well, let me explain to you this way and see if it makes sense. The 14% of our revenue in FY '25 came from the Penguin Edge business and the hardware from hyperscalers, right? So I haven't included that revenue in the FY '26 outlook. So yes, if you wanted to remove that from FY '25 and have apples-to-apples with FY '26, you would have a calculated growth rate around 20%.

It would be. I got it. And that's helpful, Nate. And then this -- Nate, the 75% growth from non-hyperscale HPC, AI. Can you give us a sense -- I mean, I guess, can we -- I guess you're giving us the bits we can back into that, I guess, is right, with the overall guide? Is that something we can back into or do we need sort of more parts?

You're talking about how much revenue would that generate?

I guess. I think the comment was -- yes, I think the comment was in 2025...

Yes. So looking at Advanced Computing and non -- there's the AI business, HPC/AI business, there's Stratus and there's Penguin Edge. Focusing on the HPC/AI business, which is obviously our strategic focus. Within that, you have sort of hyperscale business and non-hyperscale business and the non-hyperscale portion of that grew 75% in fiscal year '25. I'm just trying to [zero] really on that strategic focus area for us.

Totally. And I guess what I'm wondering is given the guide [indiscernible] you've given us some different businesses for fiscal '26. Are we able to back into what's implied for that growth in fiscal '26? Or is that something that we need more information to figure out with 75%?

Yes, I think we're basically discussing a trend that happened in '25. And what we're suggesting is that with that focus, we're trying to leverage that and build off that to deliver on the plan for '26. I don't think there's an implied association with what happened in the report for the '25 outcome and -- into the '26 number.

Yes. Ananda, I think you can kind of get to a range with data that we've given you. But the other thing I would just add to that is most of the -- the guidance range mostly reflects opportunity in that HPC/AI space for non-hyperscale customers. So that's where we're seeing that pipeline build, right? And so it's a wide range on Advanced Computing because of that. We have visibility to opportunities, but not visibility yet to the bookings. And so I left the range wider for Advanced Computing than I did on LED or Memory, which you see have tighter ranges on the revenue outlook.

And Mark -- actually Nate and Mark, like Nate sort of to the variance that's left in the bookings that drove -- describing the wider range. What would be, I guess, like what aspects of HPC/AI business, I guess, maybe what end markets or aspects would be the biggest add parts of the market that would sort of maybe drive some of the upside? I guess, is it -- yes.

What I would say -- let me just -- I'll try to answer that. The market opportunities that we're seeing the most near-term customer engagements are in the financial sector. The federal sector, which includes both government and federal integrators, there are some education opportunities that are interesting, but also sovereign cloud opportunities. Those are the 4 that I think can drive us to a successful FY '26.

By the way, this is -- to fully answer, in addition to that, those markets, I think we mentioned in our script, we've had a new CRO start, Tony Fry, who came over from NetApp, and Tony has already brought on team members to target health care and other verticals organizationally to further enhance that. But in terms of '26 outlook, our pipeline reflects the sectors I talked about being, again financial education, federal integrators and government direct as well as sovereign cloud.

Our next question comes from the line of Rustam Kanga with Citizens.

Congrats on the strong close to the year and great to hear about Penguin's pivotal role in South Korea’s Sovereign AI plans. I just had one follow-up. Nate, it was great to hear you kind of call out that you were able to add the services revenue in such a short period of time after the initial hardware deal. And I think you guys have historically talked about hardware as you lead with the hardware and then services follow on. I'm just wondering, is it too large of a leap to make to say that in this instance, you were able to sort of accelerate the time to services from an initial hardware implementation. Is that something that you're seeing? Or is that just a one-off?

Well, I think it's -- we got to be careful and be clear here. The overall solution, and as it was commented in our recorded scripts, the overall solution contemplates hardware systems, software and services. And the timing of when the revenue gets recognized is different. And so the more hardware oriented deals we take revenue credit on, that's normally upfront in any deployment. The software and services, bookings and revenue is typically ratified. So if we get a booking, that doesn't mean that we get the revenue upfront as you know. The revenue happens over the lifetime of an agreement.

And so in this case, like any case, once we install or deploy systems into a data center environment, for example, the hardware gets booked right away, relatively speaking. And we start the clock on software and services that are contracted to and that happens over time. So this wasn't that different than other deployments. It's just that we got both agreements within a certain time frame right after the other.

I think also it's -- listen, each deal can be different than the next. But when we have a large hardware deployments, the value proposition for our services tends to be higher. And so we do tend to see good services attached to those large deployments, and that was the case with SK Telecom.

Our next question comes from the line of Matt Calitri with Needham & Company.

There's obviously no shortage of conversation around AI. And lately, we've heard quite a bit of discourse around how CapEx and revenue seems to be rotating between just a few companies. And then this week, we've had reports out about AMD getting involved in chip shipments with OpenAI and another report today, questioning the profitability of Oracle's GPU strategy. Just curious what your thoughts are on how build-outs are and will progress in this space and what you're seeing in the broader market?

Well, implied in our earlier comments, Matt, is that we still think we're in the relatively early innings of broad enterprise rollouts. If I separate your questions to AMD and OpenAI in that announcement, I think that just goes to show that the capital dollars out building on future large language model training environments as well as inferencing implementations. Again, it's still on the front end, the early end of the market opportunity there. buoyed by enterprise adoption of AI, which is different than the earlier stages that were primarily large hyperscalers making significant investments in their training.

We're seeing and we're starting to see a big pickup in terms of enterprise engagements and the pipeline growing there. Now relative to your reference to the GPU gross margin announcement, what -- I guess I would say when you have a lot of people selling the same thing, it tends to get commoditized pretty quickly. And I'm not commenting on today's announcement only. But if you look at the gross margin of the hardware-only companies that are the large OEMs in the business. Their gross margins have been significantly impacted over time. And so that model is not, in my opinion, is not for everybody, for sure. I think it's -- it will get commoditized if you're selling the same basic underlying solution or chip in this case.

So I think there are 2 different issues you raised. I definitely think the market is on the early stage of deployments, especially around the enterprise opportunity. And I think the announcement with AMD and OpenAI that was in the press this week, certainly, another good example of the CapEx spending. Today's announcement that you're referring to on the gross margin piece is something that we see when there's large hardware-only type environments and competitors.

That makes sense. Very helpful there. And then as the memory market seems to be heating up here and good commentary from you guys there and guidance there. How are you differentiating your offering there? Or to a certain extent, is it just a matter of who has availability to ship this stuff?

Well -- and Matt, I know you're relatively new to our story from Needham and thanks again for jumping on the call today. Our business is largely is differentiation because we buy our supply of Memory silicon from the likes of SK Hynix and others. And we deliver a value add in terms of a system or subsystem level solution and we get margin above the industry gross margin for the commodity itself being the memory chip. And so we differentiate ourselves both through design and firmware and software and performance reliability. And so those categories are elements of our differentiation allow us to charge more than the industry charges for the Memory itself. And so it's largely a differentiation model if we're not differentiating on the design wins, we're not going to get a lot of them.

There are no additional questions waiting at this time. I would now like to pass the conference back to Mark Adams, CEO, for closing remarks.

Thank you, operator. Our Q4 and full year results validate that we are on the right path, helping our value customers solve the complexity of AI infrastructure. Thank you all for joining today's call.

That concludes today's call. Thank you for your participation, and enjoy the rest of your day.