Navitas Semiconductor Stock

Good evening, depending upon where you are in the world, and welcome to today's webinar Breaking the AI infrastructure Power Wall. Navitas 2.0 completes the Grid-to-xPU power portfolio with Claros acquisition, brought to you by EE Times, Navitas and Broadcast by AspenCore. I'm Colleen Heckman, and I'll be your moderator today. Today's webinar is a panel style discussion.

[Operator Instructions] And now on to the webinar, breaking the AI infrastructure Power Wall. On today's panel, we have Chris Allexandre, President and CEO of Navitas Semiconductor. Chris joined Navitas as President and CEO and Board member in 2025.

Chris is a seasoned semiconductor executive with over 2 decades of global leadership experience across sales, marketing, operations and business unit management. He's held senior executive roles at leading semiconductor companies, including Texas Instruments, Fairchild, ON Semiconductor, NXP, IDT and Renesas Electronics. Joining Chris today is Llew Vaughan-Edmunds, Chief Marketing Officer at Navitas Semiconductor. Llew is leading the company's global marketing organization while driving strategic initiatives across corporate strategy, business development and M&A. Llew brings more than 25 years of leadership experience in global semiconductor industry, having held senior roles at Infineon, International Rectifier, onsemi and Applied Materials.

And finally, our third panelist is Dan Kultran, VP and GM, IVR Business Unit at Navitas Semiconductor. Dan is the VP and General Manager of Navitas' IVR business unit and is former CEO and Founder of Claros.

Previously serving as the Chief Technology Officer at Epirus, he contributed to the development of systems addressing modern challenges such as asymmetrical technologies and drones. It's with great pleasure, I now am going to toss things over to the Navitas headquarters where the gentlemen are standing by. Gentlemen, take it away.

Welcome, and thank you for joining us today to this breaking the AI Infrastructure Power Wall webcast. As you surely know, we have closed the Claros acquisition, and this marks a defining step in the Navitas 2.0 transformation. Why does it matter? Power has become a bottleneck for AI compute.

Delivering more usable power more efficiently is an enabler of more compute output. There is a direct link now between power efficiency and compute capacity. In today's webcast, I'm with Llew, our Chief Marketing Officer; as well as Dan, ex-CEO Founder of Claros and now VP, GM of the newly created IVR business unit of Navitas.

We will cover and explain the Power Wall, the architectural changes needed to break it and the Claros -- how Claros completes our technology portfolio. Let me do a brief disclaimer. Our discussion includes forward-looking statements about our market, technology road map and the expected benefits of the Claros acquisition.

This reflects our current expectation and are subject to risk and uncertainty. Please review the disclaimer and the risks described in our SEC filing. Before we get going and go through the presentation, I want to share my excitement in joining forces with Claros and welcome again the entire Claros team to Navitas.

As mentioned earlier, this is a very important step in our Navitas 2.0 transformation journey. This is more than just a portfolio addition. Claros connects our existing GaN and SiC capabilities with the critical final stage of power delivery to the processor. Those are complementary technologies. GaN and SiC address conversion and distribution across the higher voltage power chain.

Claros brings integrated high-density regulation close to the xPU. We have one single system-level objective. As the hyperscalers take more and more ownership of the full power chain from Grid-to-xPU, we're now able to serve the entire need with one system-level objective, reduce losses and power density constraint across the chain so that more power can support more AI compute.

We think Grid-to-xPU really described well the breadth of our combined portfolio in power and the customer problem we are addressing. Let us start with Navitas 2.0, which we started about a year ago. This is a strategic shift away and deliberate pivot from consumer market towards high-power application with AI infrastructure at the center.

As a matter of fact, as you might have seen from our past earnings release, we expect mobile to be insignificant by the end of '26 and almost all revenue to come from high power already. Essentially, this means the pivot has happened. We focus on 4 markets: AI data center, grid and energy infrastructure, high-performance computing and industrial electrification. AI infrastructure includes both data center and the energy system supporting, and this is really where the company is heading. We have 3 technology pillars: GaN for fast, efficient conversion, high-voltage and ultra-high-voltage SiC for demanding high-voltage and high-power application and now Claros' IVR for highly integrated power delivery to the processor.

Our roots go back to strong technology innovation as outlined by the number of patents now surpassing 450, and this will continue as well as our strong legacy with over 350 million units shipped. We remain focused on execution and scaling. Our operational priorities continue to be focused on strong and reliable U.S.-based manufacturing and supply chain, supported by our Foundry Plus model and strategic partnerships that we have announced.

We remain focused on strengthening the financial model through revenue scale, margin expansion and progress towards profitability. And as commented at the announcement a few weeks ago, Claros acquisition does not materially change our plan for breakeven point. Now that we have integrated Claros, let's review our SAM and long-term opportunity, which is much larger. Our identified 2030 SAM expand from the original $3.5 billion GaN and SiC base to more than $8 billion, including new [ Gen 5 ] technology and Claros IVR. We have announced a new [ Gen 5 ] product back last summer. What is expanding the market? Approximately $1 billion come from SiC [ GeneSiC ] technology and at least $3.5 billion from the VPD and the IVR opportunity.

As you can see, AI infrastructure dominates with more than $6.9 billion and nearly 85% of our SAM, explaining why we outlined this as the core focus of the company. High-performance computing and industrial electrification remains important, but most of this opportunity aligned directly with our AI infrastructure focus.

This is the addressable market, not a revenue guidance, but it outlines we are in the right market with the right technology. Our task is now to translate this product portfolio into customer adoption and sustainable growth. Before giving Llew and Dan to go through the details, I want to show you what Grid-to-xPU really mean and the overall power AI infrastructure rearchitecture coming up.

Read the diagram left to right. There are 3 essential jobs: bring power from the grid to an 800-volt DC infrastructure, convert it efficiently inside the data center and regulate precisely at the processor at super high current. Grid to 800-volt DC, solid-state transformers, battery storage and power conversion system support efficiency, power density, resilience and bidirectional energy flow.

Inside the data center, high-density DC/DC conversion steps power down towards the lower voltage rail needed by compute tray while the 800-volt bus supports distribution. The final step, package and embedded IVR brings power down towards the 0.7 processor core, where high current and rapid load change makes essentially demanding.

All this together brings one system benefit, higher efficiency, fewer conversion stages and greater power density. The value comes from improving the whole chain, not just one conversion point. And as we discussed earlier, this directly impacts the amount of compute. And now to Llew.

Thanks, Chris. This slide really sets up the problem we're trying to solve. We call it the AI Infrastructure Power Wall. It's a series of compounding challenges from conversion losses and thermal constraints to power density limitations, all the way from the grid to the processor. If we look at today's AI data center power architecture from left to right, there are challenges developing at every stage.

First, we start with traditional line frequency transformers. These have been in existence for over 140 years. They are bulky, heavy and increasingly incompatible with the power density demanded by modern AI infrastructure. As data center power requirements move into hundreds of megawatts and ultimately gigawatts, the industry needs a much more efficient and scalable way of bringing power from the grid into the data center.

Then we move into the AC to DC power supplies. Now these convert the AC power to 48-volt DC into the rack. But as the AI rack power continues to increase, these power supplies are reaching practical limits in terms of power density, efficiency and thermal performance.

The next challenge is power distribution at relatively low voltages such as 48 volts, delivering more power requires dramatically higher current. As rack power increases, the current flowing through the system quickly becomes enormous, driving significant [indiscernible] losses, heat generation and the need for practically large amounts of copper.

We see the same issue inside the compute tray. Power still has to be converted from the 48 volts down to 12 volts through an intermediate bus converter or an IBC. This additional DC to DC conversion stage introduces power losses, heat and additional components that consume valuable board space.

And finally, we reach what is becoming perhaps the most significant power delivery challenge is the processor itself. Whether that's a GPU, a CPU, a TPU or even a custom accelerator, these processors are consuming dramatically more power and demanding thousands of amps at sub-1 volt.

So when we talk about the AI infrastructure power wall, it isn't a single bottleneck. It's a cumulative effect of limitations across the entire power delivery chain. From the grid to the processor, legacy architectures are becoming pushed beyond what they are designed to support, creating compounding losses, thermal constraints and power density limitations. And as AI power continues to scale, simply improving each stage incrementally is no longer enough.

We need to fundamentally rethink the power architecture from the grid all the way to the processor to break through these power walls and enable the next generation of AI compute. So, the question becomes, how do we redesign the architecture to break through each of these power walls. So, this slide shows how we see the architecture evolving and how Navitas is breaking through the AI Infrastructure Power Wall. The fundamental change is the transition to 800-volt distribution, fewer conversion stages and dramatically higher power density across the entire architecture.

Starting at the grid, solid-state transformers or SSTs can replace traditional line frequency transformers, providing a much compact, efficient and intelligent approach to power conversion. These SSTs convert grid power directly to 800-volt DC while enabling bidirectional power flow, active power management and seamless integration with battery energy storage systems improving both power density and the resilience of the AI data center.

Now Navitas' ultra-high-voltage silicon carbide technology enables the high voltage, the high-frequency power conversion required by these next-generation SST architectures. At the next stage, increasing rack power is driving a transition towards much higher power density requiring 3-phase AC to DC power supplies. This allows significantly more power to be delivered to each rack while improving efficiency, power density and thermal performance. Our high-voltage silicon carbide MOSFETs support these higher power architectures with efficiency, switching performance and reliability required for next-generation power supplies.

One of the biggest architectural changes is the move away from distributing 48-volt DC for the backplane to individual IT trays. Instead, AC to DC and DC to DC power conversion moves into a power side card dedicated to supporting high-power IT racks exceeding 1 megawatt with 800-volt DC distributed directly to the rack and compute trays directly.

This higher distribution voltage allows significantly more power to be delivered while keeping current at practical levels, reducing I-squared-R losses, heat and the amount of copper required in the backplane.

At this stage, high-voltage silicon carbide and medium voltage gallium nitride become key technologies, delivering the efficiency, switching performance and power density required for these next-generation architectures. The next step is the 800-volt native compute tray. Now we can bring that higher voltage much closer to the processor and use very high density, high-frequency DC-to-DC conversion close to where the power is actually consumed. Navitas recently collaborated with NVIDIA to demonstrate 800-volt to 6 and 800-volt to 12-volt DC-to-DC power delivery boards that eliminate the traditional 48-volt IBC stage.

Now this improves system efficiency, reliability, cost and compute density. Our high-voltage and medium-voltage gallium nitride FETs and GaN power ICs are ideally suited to this architecture with a high-frequency switching capability, enabling smaller, more efficient and significantly higher power density conversion directly within the compute tray. And finally, we get to the last few millimeters and perhaps the most critical stage of a power delivery. This is where vertical power delivery and embedded power delivery enabled by Claros' IVR technology fundamentally changes how we power the processor.

In summary, the result is a fundamentally different Grid-to-xPU architecture, higher voltage, fewer conversion stages and higher power density. Together, Navitas' SiC, GaN and Claros' IVR technologies address the power bottlenecks at every stage, breaking through the Power Wall and enabling AI compute to continue scaling. Now let's take a closer look at the processor power road map and why power demand is rapidly outpacing the ability of today's power delivery architectures to support it.

This chart really illustrates the magnitude of the challenge. We're seeing an extraordinary increase in the power required by next-generation AI processes. The trend here in processor power grows from around 1.4 kilowatts in 2025 and over tripling to 4.4 kilowatts in 2028 and heading to well above 10 kilowatts in the next years. We're talking about an order of magnitude change in processor power over a relatively short period of time.

And because these processors operate at less than 1 volt, supplying that power means delivering thousands of amps. It also means responding to very rapid changes in current as the processor moves between workloads. The higher voltage distribution we just discussed helps us bring more power into the rack and compute tray efficiently, but we still have to get that power through the final stage and into the processor itself. There are 3 broader constraints behind the Power Wall shown on the right.

First, we simply can't rely on Moore's Law and transistor miniaturization to solve the problem. Improvements in the processor itself has to be matched by improvements in how we power it. Second, transmission and conversion efficiency become much more important at these current levels. Even small inefficiencies matter when we're delivering thousands of amps, making the final few inches of power delivery exceptionally critical.

And third, packaging and thermal management have to keep pace. We need to deliver more power within a limited physical footprint while still being able to cool the system effectively. Together, these constraints mean that having enough power available at the rack doesn't guarantee we can't deliver it effectively to the processor.

Power delivery itself is becoming a fundamental constraint on the next generation of AI compute. So, the key message is straightforward. If we want AI compute to continue scaling, our ability to power our processors has to scale with it. Dan will now take you through a closer look at that final delivery challenge and how Claros technology addresses it. Over to you, Dan.

Thanks, Llew. Let me double-click on that final delivery challenge and explain why it becomes so difficult at these current levels.

Today, voltage regulators typically sit beside the processor. So, the current has to travel across the board to reach the chip. As you just showed, we're asking that path to carry thousands of amps. That creates 3 challenges you see on this slide. First, electrical losses.

At these currents, even a very small amount of resistance matters. And because resistance losses increase with the square of the current, doubling the current means 4x the loss in the same path.

As this slide highlights, lateral power delivery can lose up to 30% of the power, power that never reaches the processor. Second, these losses become heat. We're already trying to cool a very demanding hot processor. And now its power delivery system adds another thermal load. That makes cooling more complex and uses part of the system's thermal budget that could otherwise support more compute.

Third, the board space. You can see in the picture how much space the voltage regulators occupy around the processor. As current demand grows, adding more power circuitry puts even more pressure on the limited space. So, building on your point, this is why the final few inches matter so much. We need to shorten the current path and make power conversion more compact. That's the challenge Claros IVR technology stack is built to address.

Let me show you how we bring that conversion closer to the processor. Let's talk a little bit about that next-generation power delivery architecture. The first step on the left-hand side is vertical power delivery or VPD. This is already starting to happen as compute density increases. Instead of placing voltage regulators beside the processor, we move them directly underneath it on the backside of the board.

Think of it as shortening the delivery path from centimeters to millimeters. A shorter path means less resistance, less energy loss as heat and less parasitic inductance. That last point I made about inductance matters because AI workloads can change power demand very quickly, lowering the inductance help the regulator respond faster to those changes and keep the processor running in a stable manner.

The next step is embedded power delivery or EPD. Here, we bring power conversion into the processor package or the substrate itself, further shortening the delivery path. We see EPD as the long-term destination for XPU power delivery. VPD gets us closer today and EPD takes integration further as processor architectures evolve. Our IVR technology stack is designed to support both. Moving the regulator underneath the processor through VPD and EPD is an important step, but how we build that regulator matters just as much.

Many early VPD implementations take the familiar components from lateral power delivery, the power transistor, discrete capacitors and inductors and assemble them into a module underneath the processor. That improves the delivery path, but the module can still be very bulky and the power still have to travel laterally.

As power demand keeps increasing, simply rearranging those components only takes us so far. That's why we have spent years developing the Claros IVR technology stack. We designed the digital control, power transistors, high-density capacitors, high-power embeddable inductors and packaging all together as one integrated solution. That lets us optimize how they work together and how much space they need.

The controller coordinates power delivery and current sharing, the power devices perform the conversion, the integrated passive and advanced packaging help make the solution extremely compact and thin. This is the technology stack now we're unveiling as part of Navitas. Let me show you how it translates into products for both vertical and embedded power delivery architectures.

On the left is Claros IVR technology stack, as we just discussed. In the middle are the 2 power delivery architectures and on the right are the products that support them. The first is packaged IVR or Navitas pIVR. Think of it as an integrated IVR chip that can sit underneath the processor, replacing the larger modules I mentioned earlier. By integrating all the passive elements needed into our package, we can deliver more amps per given area that true higher current density is especially valuable when space underneath the processor is limited.

The second product is embedded IVR or Navitas eIVR. Here, we integrate silicon chiplets and advanced passive directly into the processor substrate that brings conversion even closer and creates a path to support much higher current level. We develop both products because customers need improvements before embedded architectures are ready. Package IVR gives our customers a way to adopt our technology sooner, while embedded IVR supports deeper integration over time. Both of these are built on the same underlying Claros' IVR technology stack.

Having a great technology stack is not enough. Let's look at how we build a platform to scale. We start with a repeatable 10-amp building block. We can combine 4 of those building blocks into a 40-amp chip or 16 into a 160-amp chip. To illustrate compactness, our 40-amp chip, which fully integrates all passive elements measures just 3.3x 3.6 millimeters, delivering over 3.4 amps per millimeter square. Then we scale again by combining chips to meet the processor's total current requirement.

The examples here show 64 chips delivering approximately 2,500 amps using the 40-amp configuration or roughly 10,000 amps with the 160-amp configuration. So, there are 2 levels of scalability, how many building blocks we put into a single chip and how many chips we combine at the system level. This system-level technology is called PowerArray. Our PowerArray technology uses an integrated digital controller, which coordinates the chips and manages how they share the load.

That modular approach supports both package and embedded power delivery architecture. We can build on the same foundation and configure it for different processors as their power requirement grows. Let us bring it all together. The Claros IVR technology gives us an integrated platform for delivering power closer to the processor with products for both vertical and embedded power delivery.

For customers, the benefits are straightforward. More current in less space, faster response as workloads change and an architecture that can scale with increasing power demands. Together with Navitas gallium nitride, high and ultra-high voltage silicon carbide technologies, this extends our power platform from the grid all the way to the xPU. Llew, you showed us how we can improve power delivery into the rack. What I have just covered is how we carry that improvement through the last few inches to the processor itself. As AI grows, we see significant opportunity to help customers turn more of their valuable power into useful compute. With that, I'll hand it over to Chris.

Thank you, Llew and Dan, for detailing the problem we are trying to solve and how we plan to solve it with our breaking the AI infrastructure Power Wall strategy and Grid-to-xPU portfolio.

Let me recap and bring it back together. The AI infrastructure power conversion runs from the medium voltage grid through solid-state transformation, high-voltage DC protection and rack conversion to the low-voltage processor core. Navitas' existing strengths, GaN and high-voltage and ultrahigh-voltage SiC address the conversion and distribution stage leading towards the compute system.

Claros adds the last step. IVR address the critical final conversion and regulation stage, enabling packaged and embedded approach close to the xPU, complementary, but not redundant. Higher voltage distribution and ultra-high current processor delivery are different problems requiring complementary technologies, but overall in the same system.

This combination strengthened our ability to engage with hyperscalers and AI power platform across the power architecture. At closing, I want to bring it back to one slide we already shown in the last earnings, i.e., how those power architecture leads to dollar content increase and revenue scaling.

Not one single transition. Our opportunities span several architectural change. We are not waiting for every data center to adopt the final grid to core architecture. Number one, SiC in high-power supplies. Higher rack power drives silicon to SiC replacement in high-density AC to DC supplies, including system with approximately 50-volt output. Two, 800-volt power side cars, moving power shelves and battery backup into sidecars expands opportunities for SiC and GaN in conversion and power backup.

Number three, the new AI factory architecture. Native 800-volt distribution into compute trays drive high GaN density conversion closer to the xPU. Rising processors power also drive the opportunity for IVR that Dan explained. Solid-state transformer, the high conversion from medium-voltage AC to 800-volt DC creates an opportunity for ultra-high-voltage SiC in Grid-to-Core infrastructure.

We expect opportunities for greater semiconductor content per system as these evolve alongside the growth in AI infrastructure itself. With that, we want to thank you for joining us today. In simple closing words. The problem, power delivery and efficiency are becoming critical enablers to the next wave of AI compute.

The Navitas 2.0 answer, a high-power portfolio combining GaN, high-voltage and ultra high-voltage SiC and now Claros IVR Grid-to-xPU. With Claros, we now address the critical final stage with an integrated, scalable platform for vertical and embedded power delivery. Breaking the AI Infrastructure Power Wall means turning better power delivery into more AI compute. Thank you.

Well, thank you, gentlemen, for an excellent and insightful panel discussion. To those audience members that ask a question, please note that we will address those offline after the webinar.

For more information related to today's webinar, please visit any of the resource links located in your resource folder on your audience console. Within the next 24 hours, you will receive a personalized follow-up e-mail with details and a link to today's presentation on demand.

So, thank you again for attending today's webinar, breaking the AI infrastructure power wall. Navitas 2.0 completes the Grid-to-xPU power portfolio with Claros acquisition, again, brought to you by EE Times and Navitas.

This webinar is copyright 2026 by AspenCore. The presentation materials are owned by or copyright by Navitas and the individual speakers are solely responsible for their content and opinion.

So, on behalf of my guests and Navitas, I'm Colleen Heckman. Thank you for your time, and we hope you have a great day.

Welcome to Citi Day 1 TMT Conference. My name is Kelsey, one of the analysts here at Citi covering U.S. semis. We are very pleased to have Chris Allexandre from Navitas Semiconductor here with us today.

So Chris, it's been roughly a year since you stepped in as CEO and launched the Navitas 2.0, shifting Navitas away from lower-margin consumer markets to higher power, higher-value applications. So perhaps we can start there. Tell us where are we in our transformation strategy.

Okay. So thank you for having me, and welcome, everybody. As you said, when I came on board a year ago, the prime goal was to transform and pivot the company from mobile, low-end consumer to high power, right? And what I would say today is that pivot is pretty much done. So now the job is to basically, from there, scale the company to, of course, bigger scaled, more valuable and profitable company ultimately. The change was, first, go-to-market. As we move from mobile, essentially focused on China, OEM, ODM, moving to the high-power market, which are AI data center, grid infrastructure, high-performance compute and industrial electrification, we have to completely pivot the go-to-market, right? So focusing on the hyperscalers, the xPU companies, the ODM, OEM, merchant power companies as well as anybody in the ecosystem, right? So that was a big change.

The other change was capitalizing on the underlying technology of the company, i.e., GaN, because we were one of the leader or pioneer in GaN as well as high-voltage and ultra-voltage SiC to basically use those technologies to drive the disruption that we're going to talk about in the AI infrastructure. That was fundamentally more accelerating some of the investment that we already made, partly in getting GaN pivoted from mobile, as I said, to AI infrastructure as well as the ultra-voltage SiC. The last thing was to kind of make sure that the team is in place. So a lot of work was done into upgrading the application team and the leadership team to build the culture.

I said earlier that essentially today, the pivot is done, and you can see that in the revenue. I'm sure you're going to ask a question about where are we in that transition, right? At the end of the year, we started that journey where mobile was nearly almost the entire company, up to 85% of the company. At the end of the year, mobile will be insignificant. So basically essentially gone despite having grown double-digit per quarter for the last few quarters, right? So that's kind of where we are. Year #1 was transition the company, and now we can start to build from there.

Okay. So before we head into the details, perhaps can you give us a simple version of why is GaN and silicon carbide gaining traction today? Why now? And also, if there's a rule of thumb to think about which part of the power stack are we using GaN, which part of the power stack are we using silicon carbide? And how do you see both materials accelerating from here?

I assume you want an answer without getting into an EE type of framework.

Yes, yes. A simple framework.

Simple framework. So as I talked about, we moved to high power. The 2 main markets we're focusing on within the high-power markets are the AI data center and the grid infrastructure. We call that AI Infrastructure. The reason we call that AI Infrastructure is you can't transform data center without changing the grid, and data center is the reason why you transform the grid, okay? So that's number one.

Two technologies that we have. One is SiC and the other one is GaN. And I'll come to the -- why they are using some form of conversion or power conversion. But think about power equals current multiplied by voltage. If you increase the power and you don't increase the power -- if you increase the power and you don't increase the voltage, you're going to increase the current. The loss, which is what basically makes the power inefficient, is proportional to the square root -- square of current.

So think about, let's take an 100-kilowatt DC load, if you run it at 50 volt, which is today, most data centers are using 50 volt, that gives you 2,000 amps. If you move it to 800 volt, you go down to 125. So by construction, the system is a lot more effective. That's why there is a move to 800 volt. Like we've seen 12-volt to 50 volt back in the days of server, now we're moving from 50 to 800, right?

So that's the first thing to keep in mind. SiC is a very good technology if you're running at very high voltage in a consistent way, so where power gets sustained and where robustness makes a difference. So think about SiC as a very good technology for AC-DC conversion, where you're converting high-voltage AC into DC, very good for high-power DC-DC conversion, as well as anything that goes into grid tied application.

We'll come to other applications that are kind of growing in the future, which is as you raise higher voltage, circuit protection and power protection becomes a big thing, okay? We'll talk about JFET, which is another technology. But that's kind of the SiC. The GaN is very good in high power, high-density type of application where you're switching the frequency at much higher. So if you think about how you're moving this DC-DC conversion closer to the GPU or xPU or compute load, that's where GaN kind of gets its way.

So the best way to think about it is both of them operate in different part of the conversion. There is a bit of a gray zone. If you think about some BBUs, okay, some DC-DCs would kind of use GaN and SiC as well. But directionally, both technology are very complementary to enable the grid rack conversion about the 800 volt, which I mentioned earlier.

So as you think about the rollout of 800-volt architecture in phases, like is there a material that will see a greater adoption first because of like the sidecar that we'll be implementing before we move on to the SST final state?

So go back to what I said earlier, right? So if you're trying to make more efficient systems, and to deliver more power, you have to increase the voltage. So we see multiple drives, multiple inflection points, as we call them, up to the 800 volt, okay? The first one is actually very interesting, is pre-800 volt. Most systems today run at 50 volt, which means the busbar of your rack is actually 50 volts, which has been around for quite some time. But as we deploy more data centers, as we deploy more powerful and high-compute data centers, you need more power.

To deliver more power, you need to have higher density of power because the space available for the conversion is the same. So the first inflection is actually 50 volt, I would say, classic systems moving to higher power. And that drives an acceleration of the SiC content. Essentially, when you go from a couple of kilowatts of AC-DCs to much higher power AC-DCs, you're replacing silicon by silicon carbide. So that drives the first inflection. But again, this is kind of pre-800 volt.

Now you mentioned sidecar. This is not a very creative name, but it means what it is, okay? You're moving the AC-DC power shelves and the DC-DC power shelves from the top of the rack to a sidecar. Why do that? First of all, you make space in the rack. You keep the 50-volt busbar, but you then introduce the 800 volt. So think about this as introduction of the 800 volt. And while the first one, pre-800 volt, is happening now, this one is happening in the first half next year. So what is the benefit of that?

As I said, if you move the 800 volt as the output of your sidecar, you can raise the power one step more, right? So we moved from a few kilowatts to tens of kilowatts. Now we move it to 20 to 30-plus type of kilowatt PSUs, right? Now you still need to convert this 800 volt back to 50, and you're doing that in the DC-DC space outside of the compute rack. So that's kind of the sidecar.

So the sidecar is a very quick, not dirty, but a very quick way to implement 800 volt, get some of the benefit without moving to the third inflection, which is where you move the DC-DC inside the compute rack. When you do that, you're actually essentially making the full rack 800-volt native. The busbar of the rack is 800 volt. When you do this, you have to move the DC-DC into the compute rack. When you do that, this is where technology becomes essentially GaN.

So inflection 1, replace silicon with silicon carbide, get high content, thanks to that, raise the power level. Inflection 2 is you're getting one step further, 800-volt output, still kind of classic DC-DC conversion that kind of step the GaN to next level and then the SiC to next level, a bit of GaN. Third one is really acceleration of the GaN. The fourth one, which is kind of far out, is if you think about it, the sidecar is kind of add-on to basically put something in between the grid and the data center.

But ultimately, the plan is to improve the quality of the -- the grid or kind of modernize the grid, as we talked about, which is the replacement of the transformers by something I'm sure you heard, which is a solid-state transformer. At that point, you're delivering 800 volt straight to the data center. That's '28 and beyond because you need to upgrade the grid, which is a bit more challenging, okay, from a reliability standpoint and of course, from a security standpoint.

But think about those transition in 4 steps, which I think from an investor point of view, derisk the transition from an AI data center side, derisk the transition, but also outline the importance of having both GaN and SiC because both technologies, as we talked about, live in their own space with some overlap in the middle. But if you don't have both, you don't participate to the big conversation. And if you're a hyperscaler, the only thing you care is how do I move 35,000 volt all the way down to sub-1 volt in the most effective way, which is less conversion, higher power conversion, higher efficient conversion. This is where GaN and SiC have something in common.

How do you think about the growth rate of this material market, GaN, silicon carbide? Or broadly, like what is Navitas' content per kilowatt or per megawatt? How do you think about that?

So everybody is asking what's the content of SiC and GaN and depending on the stage and kind of very difficult to model, right? So we've published models to give investors kind of a reference, right? And the way I would talk about it is if you think -- so the models are based on assumptions. So the -- you have to take an assumption on how many data centers we're going to deploy between now and 2030. It could be 220 gigawatts, it could be 300 gigawatts, okay? It's getting closer to 300 by the day. But it's kind of a window, which is a range which is reasonable.

And that is the assumption you're using, 200 to 300.

Then you have to factor how many of those will be 800 volt. How many of those will be 800-volt sidecar? How many of them will be 800-volt native? And of course, this depends on time line. You can't -- when you start a data center, you're going to stick to an architecture. You don't change it halfway through. So data centers that are deployed in a year from now might actually well use sidecar, okay? And actually, some of them will move to native. So you have to make some assumption there.

But high level, what we've modeled is if you think about inside the data center, the GaN gets you to $15,000 per megawatt. That's replacing silicon. So that's the content which is today in silicon, okay, moving to GaN. When you look at SiC, there are 2 portions. One is the -- inside the data center, the AC-DC conversion I talked about, which is about $10,000 per megawatt, which includes both the SiC for AC-DC conversion as well as the JFET, which I mentioned earlier for the power protection.

And then you have another 10, maybe up to 15 in the long run for the grid modernization. And that SiC is at much higher power, okay? So that's kind of 2,000 volt, 3,000 volt [indiscernible] volt. So that gives you a sense, total is 25-ish, which is still a small portion of the total power that is being consumed per megawatt in data center. But as you guys understood from the discussion we had earlier about the 800-volt transition, this is basically gradually a replacement of silicon.

Silicon will continue to exist for a long time, okay, for multiple reasons, okay? Because not everything is switched on day 1. Nothing is actually a digital switch and control and many technologies need silicon. But when it comes to this power conversion, what you're going to see is gradually a replacement of silicon on the first stage and the second stage with GaN and SiC. And we'll talk about the third stage, which is as you get closer to the GPU or the xPU.

Okay. So you talked about it's going to be a gradual replacement. So what is the key metric that a customer looks at to drive that decision to do that switch?

First of all, it's the need that drives the transition. So customers don't decide if they're going to use 800 volt or not. It depends on the level of power density you're trying to achieve and the level of the rack level. So we always refer to what is driving all this is the rise of the power density of the rack. If you want to do high-power rack, you have no choice to move there. I think what people -- customers are looking for is, number one, do you know what you're doing, okay? That's a good place to start from, right?

So we didn't start GaN and SiC yesterday, okay? Actually, Navitas started as a GaN company, pioneered GaN with another company that got acquired by a very big German company, kind of pioneered GaN into the first market that got GaN to mainstream. So we shipped 350 million units of GaN. Different market, but the underlying need is the same. It's how do we -- it was when the chargers were moving from 20, 30, 40 watts to 100 watts, okay? And you had to move from silicon to GaN, right? So that's number one. Do you know what you're doing, okay? Are you starting -- are you planning to start? Or do you have years behind you?

Number two is, are you ready? And I think particularly in the AI data center, but the grid as well because the grid used to be slow, it's not that slow anymore. Customers expect technology to be ready, ready to use and so you can ramp. The third one, which I mentioned earlier, is the fact that because you can't ignore -- you can't look at each conversion independently, having both the ultra-voltage SiC, the high-voltage SiC and the GaN makes a difference. And we compete with different people on the SiC side and on the GaN side. The SiC competition is traditionally the SiC vendors that were focusing on EV that are pivoting to data center.

In reality, this is not the same because the voltage are different. So EV is lower voltage, it's high voltage, but lower voltage, where data center is 1.2 kV and above and the grid is more than that. The aspect of how reliable the technology is also different. So we compete with different set of people. On the GaN side, we also compete with a different set of people. But people having both is actually a very small pool, which I think is also very important and probably underappreciated for those that don't meet customers like I do every day. I'm talking about the merchant powers, okay? The big -- the Deltas, the Flex, the LITEON, the Vertiv, the Eaton, those guys do care. And of course, the hyperscalers, right? But those guys do care if you have both technologies.

Right. So maybe you can elaborate on that. How do you work directly with the customers, the hyperscalers? Or do you work more with the Flex, Deltas, LITEONs of the world?

All of the above. So if you think about hyperscalers, and they are different type of hyperscalers, okay? So let's make a difference between the Microsoft of this world, the Metas of this world, the Amazons of this world and the Google of this world, which are more vertically integrated. And here is, kind of, in between. NVIDIA is a GPU company, but it's actually more than a GPU company, okay? They are doing racks, okay? And when you do racks, you kind of almost do data centers. So I put those last 2 in kind of into the same camp. When you talk about silicon to GaN, silicon to silicon carbide, moving from 5-kilowatt PSUs to 30-kilowatt PSU, 50-kilowatt PSUs with liquid cooling, high integration of the DC-DC close to the compute load, they are involved.

The Google and NVIDIA.

It doesn't mean that they will do the design by themselves. The answer might be different on the NVIDIA side and Google. But they might spec the design and get it done by the Deltas, the Flex and so on, right? So you have to basically work with those guys and their architects and the people are thinking about power to enable the full power of the GPU or the xPU. And I say xPU because we talk about GPU all the time, but it's actually more an xPU play.

Then you have to work with the Deltas, the Flex, which are basically the buying customers. And those guys are the buying customers. So they are basically accountable to the hyperscaler to deliver quality on time, passing qualification and so forth. So we work with those guys. And it's not like we convince one and we implement the others, you have to convince both, which is very similar to the high-end computing market, right? And then you have to work with architects, okay? There are system integrators and people that kind of look at this at the high level.

What is interesting is in all my career in semiconductor, the hyperscalers are taking a bigger seat in driving this transformation. The question we ask ourselves is, would 800 volt have happened if NVIDIA didn't drive the transformation? The answer is yes, but way later than it's happening now. Would the grid transformation or modernization would have happened if the AI guys, i.e., NVIDIA, Google, would have said -- or Microsoft, they said, listen, we have to change the grid. It would have happened, but a lot later.

So I find it interesting that the hyperscalers xPU vendors are taking a bigger influence on the overall Stage 1, which is super high voltage down to 800 volt, 800-volt to [ 5 or 6 ] and [ 6 ] to GPU because if that -- if this doesn't get fixed, they can't get the full benefit of the AI workload they are trying to run. Let's remember that when we started this, the #1 bottleneck was how fast and how powerful can you make a GPU and xPU. Then it moved to memory. How can you basically make the data flow between the 2 as seamless as possible. The next bottlenecks are power and copper to fiber. So the -- what I call the power wall is today what's limiting the full capability of the xPUs.

Many xPU companies are underclocking their processor because they can't get enough power delivered to the processor. So the 800 volt is solving part of the problem, but we'll talk about the last centimeters. That's why we acquired a company -- we announced the acquisition of a company just a few weeks ago because if you fix all this and you do a great 800 volt, you still need to get more amps delivered to those thousands of amps processors and xPUs.

Exactly. So...

Good segue.

Good segue. Exactly. So talk about Claros. Claros brings about vertical power delivery. So we're essentially going to be competing with some of the incumbents right now and also offering it kind of end-to-end across the entire power tree. Talk about that rationale for that deal. You highlighted some of that. It's going to expand the SAM for Navitas. So perhaps you can elaborate on that.

So first of all, I would remind everybody that we announced the acquisition, and we have not closed yet. So everything I'm going to say is based on what we've said so far and what Claros themselves has talked about. The target for the close is around beginning of Q4, okay, October time frame.

So what is the rationale for GaN and high-voltage, ultra-voltage SiC company to acquire a VPD-IVR company? Go back to what I said, you can't just look at the first stage and the second stage independently. First stage is 35,000 volt to 800 volt. Second stage is 800 volt to 6 volt or 50 volt, whatever. You can't look at that independently. And what is the next -- so let's assume we are 800-volt implementation. What is the bottleneck today is the first stage. You don't deliver enough power, which is really current to those processors, right? So this has to be fixed.

So you mentioned the incumbent. There is no better way to get into a market than when there is a disruption. GaN and SiC company are competing with silicon company because there is disruption, which is 800 volt. Otherwise, it stays in the same group of people. It's the same on the first stage. So again, just to zoom out a little bit, we use 2 words here, VPD and IVR. So let me specify them because there's a lot of confusion. The first one is structurally how you deliver power. So today, you deliver power to the processor on the laterally. So on the PCB, you have a lot of MOSFETs, silicon with caps, inductors and a lot of passives to deliver thousands of amps to those processors.

Vertical delivery is basically doing that vertically on top, on the back. You can do that with MOSFETs. The first implementation of VPDs are actually taking those discrete MOSFETs, capacitors, inductors, putting in a big module. You put that on the back of the PCB, and that's a vertical power delivery. That doesn't fundamentally solve the problem. It helps, but doesn't solve the problem. The IVR is a new concept where you integrate everything on a piece of silicon. So think about control, think about FETs, think about drive, capacitors, inductors.

So the challenge here doesn't -- is not about how you take all those discrete things into a package. This becomes how you integrate into a monolithic piece of silicon. That's where you basically get to the next level when it comes to how you deliver power. And you can do an IVR in a chip, which will become a vertical power delivery. So you put that on the back of the PCB and you deliver power. And you can do that in a kind of IVR way where you put that in the PCB or even in the substrate of the xPU. So the reason why I say that is there's a confusion that people think IVR is a play where you integrate the power inside the PCB and inside the substrate of the xPU.

You can use IVR in a vertical power delivery way. And actually, all the IVR companies are doing chips with your package. And if you look at Claros, what they talked about, which is they have a 3:1, which is 3.3-volt entry, sub-1 volt output, 40 amp, no caps, no inductors, nothing outside, everything integrated. And the reason why they have something great is their concept is it's a mesh or they call that an IVR array, where you can put many of them together and they can control through current sense, the delivery of the current from each of those chips.

So what this means is if you need 2,000 amps, you put 50 of them. If you need 4,000 amps, you need you put 100 of them. Of course, there is a limit on how many of those chips you can put together because you inject latency and other things, right? But their vision is let's be as scalable as possible to deliver to various xPUs, architecture that can basically fit them. So going back to your point, how does it compete with the incumbents? Well, the incumbents are like the silicon before 800 volt. And that's going to continue for a while.

So don't -- I want to be clear. I'm not saying that silicon will be replaced. We talked about $20,000, $25,000, maybe $30,000 per megawatt. The total power content is probably $100,000, $150,000. It's going to take time. But this disruption will force the replacement of MOSFET or MOSFET and the IVR. And that's why you saw ADI acquiring Empower. And what is interesting is that disruption came from mostly start-ups that saw that were not in the tunnel effect that most companies have, which is I have MOSFET, so let me optimize my MOSFET to -- you have to change the game. And those companies, all of them started basically with the idea of how do we basically not get from a few hundred amps to more, how do we get to thousands. And that's kind of where the IVR technology came into play.

Is there a way to think about the content as we shift from lateral to the VPD to IVR? It seems that there's a lot more engineering work that's involved and eventually even the use of GaN as well.

So it's difficult to talk about content, but what I would tell you and you referred to the SAM increase, right? So based on 220 gigawatt deployment, we said the SAM for GaN SiC, removing EV, removing China. So talking here about the U.S., not competing with China for China. And the grid is about $3.5 billion in 2030. Our large competitors like Infineon and others can actually refer to a bigger SAM. But for me, $3.5 billion feels fairly big, correct? Then you add $1 billion for the JFET, which I referred to, which is a SiC technology, targeted at more power and circuit protection. So eFuse, ORing, circuit breakers, onsemi, Infineon kind of in the JFET, right? So that's added a $1 billion. So that's $3.5 billion to $4.5 billion.

The IVR-VPD, which is the last step, the last centimeters to the xPU is basically adding as much as that. So it's doubling the size because as you get closer to the GPU or the xPU, you have more content. So for us, the key point is we do believe -- I do believe that, as I said, if you do one stage only, you're not going to sit at a big table because the architects are looking at how do we basically enable 35,000 volt to 1 volt. And if you only participate to the first 2 ones, I mean, it's pretty decent, $4.5 billion is -- but lot of the SAM is in the last one. And it's also touching the xPU. And since I said the xPU guys or the hypers are driving, they definitely want to be there.

So ADI is not a MOSFET company. They acquired Empower. We're not a MOSFET company. I think strategically, I think you're going to see all the power companies have to somehow get into high power. That's why you hear everybody talking about I'm going to have GaN and working on GaN. And I think it's going to be very difficult to do it without IVR in the midterm.

Okay. So I know I have a couple of minutes left. I'm going to open it up to the floor. Does there anyone else have any questions? And I can continue. Yes. So on the IVR, perhaps could you talk about when do you see that broad deployment for that inflection?

As I said, the first inflection is in a VPD way. So like a chip or an array of chip on the backside of the PCB.

This is happening now?

That is actually starting to happen, okay? The big modules are happening now, but I would say the tiny IVR chips, Empower and others, Claros, this is happening in '28. So again, I refer to what we said at the announcement of the acquisition. It doesn't change our organic plan, '26, '27, which is really driven by SiC and GaN inflection that we talked about start to kick in, in '28. I think the Claros CEO talked about his leading customer, which is an xPU customer, ramp in '28. So our job post-integration will be, of course, to execute on that, add more. But I think you have to think about this being a '28, '29 acceleration more than changing any of the short-term, midterm '26, '27 and beginning of '28 plan.

Okay. Could you briefly quickly talk about your foundry strategy? Because I know you guys are fabless. How do you compete with some of the IDMs, the incumbents in the space?

So I think there is a bit of a belief out there that the only way to compete is to be an IDM. And I won't give a name, but I would take the fastest-growing semiconductor power company today is a fabless company, okay? And I never heard Michael talking about owning his own fab. So that's already a proof point that you can be, right? So -- and by the way, I know other companies without naming them that have their own fab and that they are in the world of hurt right now, okay, with underutilization and so forth, right, being a SiC company.

So you have to be careful with this vision that dogma, which is you have to be an IDM to be successful. What you want is, first of all, SiC and GaN and IVR have different needs, okay? So we work with Global. We used to work with TSMC. We're now moving to Global on the GaN side. We work with X-FAB on the SiC side, and Claros is working on the 14-nano with Samsung. Those are different needs, okay? You can't put all that into a single wafer fab.

Number two, there is a way where if you own your moat and your process. So what I mean by that is you use the baseline of your foundry partner, but you have some moat which are additional differentiation at the device level, at the process level, at the application level. And you have -- you invest capital into securing capacity, expanding capacity, then you can actually have what the customer needs. The customer needs guarantee of supply, you can expand, you control your yields, your quality, you can do that without owning the fab. I'm not saying that owning a fab is not good.

I'm just saying that when you are moving fast, emerging with new technologies, financially, it's actually not a very good model. And what matters the most is basically have this fine line about investment, owning moats controlling your destiny and investing in partnership with your foundry. And by the way, I know IDMs that are moving fab-lite, okay, without naming one large European U.S. company in the control automotive, they are moving fab-lite. And they talked about how efficient this model is. But of course, we compete with heavy IDM. And customers are asking us, hey, how do you ensure it is reliable? How do you ensure you have enough supply? Do you control your destiny? And this is where this foundry plus model that I talked about is key, right?

Thank you. So I know we are running out of time. Perhaps one last message for investors. What about Navitas' story is currently underappreciated by investors?

I think I will repeat myself a little bit, but I'll leave you with 2 things. Number one is, we have transformed -- we started to transform this company about a year ago. The revenue is more or less where it was 1.5 years ago, but the makeup is completely different. So investors look at the quality of revenue, but have not yet fully understood this transformation that we've done, moving from mobile being 85% of the company to mobile being insignificant. And we see at the end of the year, mobile will be insignificant, let's say, low single-digit and AI infrastructure will be more than 1/3. That's a big pivot when it comes to what makes the revenue.

We talked about the fact that we have backlog well into '27. This was not the case 18 months ago when you're focusing on mobile, right? So that is underappreciated, but I would say, give us 2, 3 more quarters of double-digit consistent growth and I think this goes away. So I would say this is not -- this by construction. The #1 thing that gets completely underappreciated, and I get -- Tonya and I get questions every time we meet investors, less and less, but still is people think it's a digital switch, okay? When is 800-volt NVIDIA, GaN happening? What day? What time? What month? And we spend a lot of time talking about the inflection. We talk about Infineon does too. We talk about a lot of time explaining the different transition.

This is actually not a digital switch, okay? Stage 1 or inflection 1 is happening will continue. We're going to ship silicon carbide into AC-DCs for a long time. Sidecar is happening next year, and this will continue for a long time. And then native 800 volt is going to happen second half next year, early '28, and this will continue. So that fragmentation and succeeding opportunities is what's totally underappreciated, which I think will drive consistent growth. And we have not even talked about the grid, which I think is another kick in '28 and beyond.

That's the one thing which I would leave all of you with is don't think about this, this is not winning the iPhone, okay? This is server and this is AI. And I think there are multiple technology disruptions and change that will drive content expansion. Of course, we have to win every step of the way.

Of course. Thank you. Thank you very much.

Thank you. Thank you very much. Thank you.

Hello, and thank you for standing by. My name is Lisa, and I will be your conference operator today. At this time, I would like to welcome everyone to the Navitas Semiconductor Second Quarter '26 Earnings. [Operator Instructions]

I would now like to turn the call over to Brett Perry of Shelton Group Investor Relations. Please go ahead.

Good afternoon, and welcome to Navitas Semiconductor Second Quarter 2026 Financial Results Conference Call. Joining us today are Navitas's President and CEO, Chris Allexandre; and CFO, Tonya Stevens. PAUSE I'd like to remind listeners that management's prepared remarks contain forward-looking statements, which are subject to risks and uncertainties, and management may make additional forward-looking statements in response to your questions. Therefore, the company claims the protection of the safe harbor for forward-looking statements that's contained in the Private Securities Litigation Reform Act of 1995. Actual results may differ from those discussed today, and therefore, we refer you to a more detailed discussion of risks and uncertainties in the company's filings with the Securities and Exchange Commission, including Forms 10-K and 10-Q.

In addition, any projections as to the company's future performance represent management's estimates as of today, July 27, 2026. Navitas assumes no obligation to update these projections in the future as market conditions may or may not change except to the extent required by applicable law. Additionally, the company's press release and management's statements during this conference call will include discussions of certain measures and financial information in GAAP and non-GAAP terms. Included in the company's press release are definitions and reconciliations of GAAP to non-GAAP items, which provide additional details. For those of you unable to listen to the entire call at this time, a recording will be available via webcast for 90 days in the Investor Relations section of Navitas' website at www.navitas.com.

And now it's my pleasure to turn the call over to Navitas' President and CEO. Chris, please go ahead.

Good afternoon, and thank you for joining us on today's second quarter 2026 earnings call. We appreciate your continued interest and support as we execute our strategic transformation to Navitas. In the second quarter, delivered increasing revenue of 22% sequentially, coupled with a stronger third quarter guidance. High power markets grew more than 50% year-over-year, serving as further evidence of the building momentum in our GaN and high-voltage IC product, especially in our focused area of AI infrastructure. We're also delivering on our Navitas 2.0 transformation. We are well ahead by over 1/4 of expected action for nearly all sales to be coming from high-power market by year-end, with revenue contribution for mobile and low-end consumer being insignificant.

We continue to deliver step by step on what we said we would do, and this quarter serves as another proof point of our consistent progress. Over the past several quarters, we have aggressively pivoted the entire organization to focus on high power market, where Navitas GaN and high-voltage SI technology can deliver meaningful differentiation and increase long-term value. The resource reallocation and organization realignment is now substantially complete with new leadership in place and a refreshed product and technology road map we are sharpening our focus on AI infrastructure, which comprise both AI data center and the grid energy infrastructure required to power them. Combined, this AI infrastructure market represent the vast majority of our long-term serviceable addressable market for GaN and high-voltage sic and underpin our future growth trajectory as a high power company.

Turning into a closer look at our reported results and progress for the second quarter. As previously mentioned, total revenue increased 22% sequentially to $10.5 million, driven by growth across our high power markets. High Power represent the majority of our overall revenue mix with revenue contribution from mobile in Q2, declining both sequentially and year-over-year as in the prior quarter. I also want to highlight that both GaNn and SIC contributed to our sequential growth with a particular acceleration in our 6 business during the quarter. As expected, we also delivered expanded gross margin as a result of more favorable revenue mix towards higher value, higher power product and improving scale.

Notably, our strong momentum continues to build and accelerate into the second half of the year. Our expanding backlog extend beyond '26 coupled with record book-to-bill supporting our expectation for continued double-digit quarterly growth through the second half of the year. The third quarter will also represent a return to year-over-year growth, driven entirely by high power markets. This also translates to achieving mid-single-digit revenue growth for the full year, while similarly having substantially exited the mobile and the low-end consumer market. This is a significant change in the revenue composition for the company and clear evidence that we are delivering on Navitas 2.0 transformation, with growth increasingly driven by a combination of AI data center, and grid and energy infrastructure.

We expect AI infrastructure market will represent more than 1/3 of our total sales by year-end, setting the stage for continued momentum in 2027. While we are nearing completion of our transformation to a high Oporcompany, our focus continues to be grounded in 4 key pillars: market focus, technology leadership, operational efficiency and financial discipline. Starting with our focus on high power market. The rapid adoption of AI is driving immense market demand for overcome critical power bottlenecks across AI infrastructure including both AI data center and Green Energy. As a result, Navitas unique ability to deliver high-power products, leveraging both can and high-voltage technologies, we are benefiting from accelerating momentum to enable customers' high-power application within data center as well as the grid and energy infrastructure needed to supply them with power. Together, those 2 areas of AI represent the large majority of our long-term sand growth trajectory and where the company is headed.

In AI data center, we are currently generating growth ahead of the market transition to DC. For example, increasing power level in CDC power supply units are driving the need for higher density which, in turn, is accelerating the replacement of silicon with our high votes. We are also actively engaged with hyperscales merchant power customers, telecenters OEM, ODM and multiple programs ramping in the second half of '26 that will accelerate throughout 2020. We're also seeing strong traction in DCDC PSUs and battery backup units where both our SiC and GaN solutions are being designed in. Again, this activity is happening today in advance of the 800-volt transition.

In fact, we continue to believe that the transition to architecture for mac-generaton AI data center will happen in 2027. as various XPUs, GPU with hyperscalers will introduce it at different times, and it will unfold in a series of steps. Each step will represent an inflection point that drives increasing momentum and explosive growth for Navitas high-power GaN and high-voltage on content. I will briefly walk through each of those inflection points, which also outlined in the earnings related slide deck that we've posted to the Investors section of our website. What's clear is the evolution to Elunate is inevitable as it remains the industry's only path forward to achieve much higher power and higher density AI. The first infection point second half 26 ramp and accelerating in first half '27. SiC adoption in CD CPS is being driven by power scaling and density requirement independent of the 800 volt DC initiatives.

As the AI day center racks require more power, it is driving CDCPSUs, which ultimately drives high density and therefore, accelerating the replacement of silicon by sick even with 5-volt DC output. This is already underway, and the growth is happening now and will continue throughout 2027 and beyond.

Following, there will be a second infection ramping in mid-2027. First, the introduction of the 800-volt bus bar in the sidecar rack with power system elements such as CDC porches and BBU moving from the IT rack to the power side car with output of 800-volt DC to the IT rack. This change is bringing additional high-voltage content in higher-power SCDCPSUs now with 800-volt output plus new SiC and game content in top of right CPSUs and BBU. We are in advanced system design and reliability testing with several key customers and are preparing the ramp.

Then the third inflection point, ramping mid to late 2027 really accelerating in late '27 and early 2028. The integration of the high-density DC/DC conversion tightly into the GPU NXP trays using GaN for its superior switching frequency and power density in megawatt scale rack across various GPU, XPU and hyperscalers at various time. At that point, fundamental change happened in data center IC rapper architecture. 800-volt comes in straight to the servers. This is what most are referring as native on 800-volt. We're highly confident in our position for 2027 ramp with our GaN. Seamlessly -- the ACD CPS will continue to be in higher demand for high-voltage IC with increased power level and density on top of BBUs and other power system.

Lastly, there would be a fourth inflection point 2028 and beyond. This is where solid state transformers come into play an on-site data center, taking mid-voltage AC electricity for utility grid and directly converting to 800-volt which get distributed across the data center. This is the full DC evolution with ultra-high voltage sick and GaN across grid monetization, solid state transformers and end-to-end power delivery from grid to core with full wide-band gap solution. Complementing this significant opportunity within AI data center is the equally large and even longer duration market opportunity in grid and energy infrastructure. Today, we are actively advancing design activity and sampling across BSS, solar farm converters, PSUs and solid state transformers application.

Our recently introduced 2.3 kV and 3.3 Gen6 modules obviously excellent feedback, and customers have begun requesting volume samples for system-level testing in the second half of the year. We're also seeing early interest in our new ISOLEL-T0247 family, which offers unique advantage in liquid cooling application. Importantly, I want to reemphasize that Navitas remain technology agnostic, and we are prepared to offer customers the optimal solution, whether that began or high voltaic across the full power chain from grid to rack. This unique flexibility is allow us to capture water content per system as well as support multiple architectures. As previously mentioned, both GaN and SIC are contributing to the current growth, and we expect AI infrastructure to drive the substantial majority of our revenue and growth going forward.

Turning to our second key pillar. Technology leadership is essential to our success, and we continue to diligently invest in innovation, and expanded product road map for both GaN and high-voltage. On GaN, we are advancing our preface platform solution, including the 80 to 6-volt DCDC power delivery Board demonstrated at recent industry events with a 800 to 12-volt version in development. We have kicked off a new program utilizing Navitas unique solution to maximize system efficiencies in the secondary side or 800-volt data center optoologies.

Our industry-leading DSN 8x8 to site cool package continues to gain broad adoption with superior power density, thermal performance and board space savings. And our 650-volt $11 million remains the lowest RBSM high-voltage GaN device in the industry, and we have a significant number of customers preparing for mass production. Additionally, our medium voltage honorable game is seeing increasing adoption for secondary tied and other applications.

On the high-voltage genetic technology based on our property trench assisted planner architecture continues to differentiate with its best-in-class railability efficiency and manufacturability. -- attributes that are increasingly critical as voltage scale from grid and energy infrastructure application. We recently introduced our isolated TO247 product family spanning 1.2 kV to 3.3 kV, delivering module-like performance, the standard discrete footprint with integrated isolation for direct cooling and simplify customers' manufacturing. As mentioned earlier, we're also seeing customer traction in both IDC and grid and energy infrastructure application.

We also recently expanded our SCS portfolio with newly introduced 12 kb JF product line to be released early next year initially targeting AI data center, Solstad transformers and energy grid infrastructure application. Our new JFE product line opens door to address an additional $1 billion of incremental TAM by 2030. Also, we continue accelerating towards our ambition to deliver best-in-class ultra high-voltage SiC technology and product and are already in discussions with select customers regarding the planned third quarter release of our new 6.5 KBC technology, which we expect to unveil very soon.

Additionally, we are currently engaged with on the development of next-generation 10 KV SiC devices with a prominent lead customer and expected announcement in coming weeks. In addition to expanding our existing SIC portfolio and technology, last week, we announced a strategic partnership for MagnaChip to license our genetic Gen 4 and Gen 5 trials assisted planar technology, spanning 1.2, 2.3 kV, 3.3 kv and high voltage supported by our supply chain and material ecosystem, the technology will report it, qualify and internalize in their fab in South Korea. This partnership delivers 2 primary strategic benefits. First, it enabled expanding adoption of our SIC technology across more target markets, expanding Navitas technology beyond the technology current focus; second, and longer-term, this collaboration facilitate establishing of another foundry source of Navitas SIC wafers, ultimately strengthening our supply chain resilience and supporting our ability to efficiently scale Genex solution.

Our deliberate strategic decision to prioritize AI infrastructure over automotive, unlike some of our competitors has allowed us to bring focused, high-performance product to fact to market faster. I think both GaN and SiC is also seen by customers as a key differentiator and allows us to focus on customer needs, independent of any technology buyers. Additionally, this has allowed us to secure initial design wins with key customers that will continue to support our long-term growth trajectory for years to come.

Operational efficiency. With respect to operations, we are making excellent progress on our strategic partnership with GLOBAL FOUNDRIES lead part from our pivot to 8-inch gain are on track for customer sampling and qualification before year-end, and we expect to have initial qualified product in early 2027. This transition will enable U.S.-based GaN manufacturing, supporting national security application and long-term supply chain resilience. I also want to note that we have secured appropriate buffer capacity at TSMC would ensure a smooth transition for existing customers throughout '21 and beyond. In addition, we continue to further strengthen and streamline our supply chain, consolidating to fewer, more strategic OSAT partners that are better equipped to support high power at scale.

Internally, we are also increasingly leveraging AI tools across designs, operations and other functions to accelerate execution and improve efficiency as we scale. In terms of the fourth pillar, maintaining financial discipline continues to be a fundamental operating principle. Over the past 9 months, have retransformed the organization with realized significant efficiency and have held operating expense essentially flat. With our transformation now substantially complete. And with a clear facility into accelerating report, we are prudently increasing investment in specific areas, including expanded product development like our Jet Horizon to strengthening customer support for key committed program and enhancing operational readiness for upcoming ramp of volume shipments.

Each of these objectives are directly aligned with our goal of capturing a substantial multiyear growth opportunity for GaN and high-voltage IV solution across AI infrastructure markets. Also, we recently raised additional capital to further strengthen our balance sheet and support ongoing strategic execution. More specifically, with $567 million of cash at quarter end, we now have increased flexibility to fund strategic investments in our business, including our Foundry Plus program, capacity expansion and supply renovation agreement with our foundry partners as well as potential strategic inorganic opportunities. That being said, I want to be clear that our immediate and overarching focus remains on driving strong top line growth together with gradual gross margin expansion through improving mix and scale while maintaining an unwavering path towards becoming a profitable high power compete.

In closing, I'm very pleased with our continued progress and growing momentum. Q2 represents another proof point that we are executing on our strategic Navitas transformation. We are delivering on our commitment to achieve quarterly growth by year-end will have substantially completed our transition to a high power company and expect to be back to year-over-year growth. This majority of the growth is being driven by AI infrastructure market. This is also supporting our expectation for continued double-digit growth for the second half, setting the stage for continued growth momentum into '27 and beyond. With our substantial cash balance and market leadership, we are well positioned to deliver sustained growth as we capitalize on the opportunity to enable the AI revolution with our differentiated Hibogame and high-voltage sake.

With that, I'll pass the call to Tonya to review our second quarter financials and the third quarter outlook.

Thank you, Chris. Before I begin, please note, unless otherwise indicated, I will focus my comments on non-GAAP results. A detailed reconciliation of all non-GAAP to GAAP financial measures can be found in our press release published earlier today.

Revenue in the second quarter of 2026 was at the high end of guidance, increasing 22% sequentially to $10.5 million. This represents an increase of approximately $1.9 million from the $8.6 million in the first quarter. As Chris highlighted, the double-digit growth was driven by increased traction in high-power markets, which grew more than 50% year-over-year and reflects a notable improvement in our revenue composition as our mobile and low-end consumer business continues to be a smaller portion of overall revenue. We continue to expect this historical business to become insignificant by year-end. As a result of improved product mix and higher quarterly revenue, gross margin expanded by 50 basis points sequentially and 100 basis points year-over-year to 39.5%.

Our accelerating shift in overall revenue mix towards higher value, high power markets and away from mobile and low-end consumer remains fundamental to our ongoing gross margin expansion strategy. We continue to expect gradual improvement in gross margin throughout the year as we drive top line growth in high-power markets, coupled with expected return to year-over-year revenue growth. Operating expenses for the second quarter were $15.5 million compared to $15.0 million in the prior quarter and $16.1 million in the same quarter a year ago. Operating expenses for the quarter continued to reflect our commitment to focused and disciplined spending -- was at the high end of our guidance range as we began making incremental investments in the business, particularly in new R&D programs to accelerate growth.

Having diligently maintained effectively flat OpEx in recent quarters during our strategic transformation, we are increasingly focused on the resources and investments required to support the longer-term success and sustained growth of the transformed company. As such, we are targeting a prudent increase of approximately $1.0 million to $1.5 million in quarterly OpEx beginning in the third quarter. This equates to a roughly 10% increase yet remains meaningfully lower than our expected top line growth rate. The incremental OpEx will be allocated to scaling the business, including investments to accelerate new product development strengthen our engineering and application support for key committed programs and reinforce operational readiness in advance of expected growth in ramping shipments.

Loss from operations in the second quarter was $11.4 million compared to a loss of $11.7 million in the prior quarter and $10.6 million in the second quarter of 2025. In Q2, weighted average basic and diluted shares outstanding were approximately $240.7 million, resulting in a Q2 loss per share of $0.04 flat to the $0.04 per share loss in the prior quarter and compared to a loss of $0.05 per share in the year ago second quarter.

Before moving to the balance sheet, I want to briefly provide additional context related to our reported GAAP net loss for the second quarter. Results on a GAAP basis included a noncash charge of $203 million related to the October 2021 business combination earn-out share provisions that were contingent upon stock price appreciation targets. These earnout shares were deferred merger consideration paid out to stockholders in connection with the company's SPAC transaction. This earn-out was fully recognized and settled by the end of Q2 and no further charges related to it are expected. As such, going forward, there will no longer be an associated line item for the change in fair value of this earn-out liability reported under other income or expense on the company's statement of operations.

Turning to the balance sheet. Cash and cash equivalents at the end of the second quarter 2026 were $557 million compared to $221 million at the end of the first quarter. The increase in cash and cash equivalents primarily reflects the additional capital raised during the quarter of approximately $373 million at an average stock price of $21.89, which meaningfully strengthened the company's balance sheet and overall financial position. As a reminder, the company continues to have no debt.

In addition to bolstering liquidity and working capital flexibility, the significant added capital ensures ample resources for accelerating our continued transformation into a scaled high power company. This includes strategic investments in support of advancing our foundry plus initiative potential capacity expansion and supply reservation agreements with our U.S.-based foundry partners as well as potential pursuit of selective strategic opportunities.

With respect to inventory, we ended the second quarter with $19.5 million of inventory compared to $14.9 million in the prior quarter reflecting the start of our build of appropriate buffers of TSMC wafers to ensure a smooth transition for our customers. This buffer inventory is also reflected in an approximately $15 million increase in Q2, prepaid expenses and other current assets on the balance sheet until the wafers are received as inventory in future quarters. The sequential $4.6 million increase in Q2 inventory and $15 million prepaid for future anticipated wafer receipts primarily reflects our measured investment to support customers' future anticipated AI data center growth.

More broadly, channel and distributor inventory remains at healthy levels. Moving to guidance for the third quarter of 2026. We expect accelerated sequential growth with revenue increasing 28% to $13.5 million, plus or minus $0.5 million. At the midpoint, this also represents a return to year-over-year growth, while reflecting a completely different revenue composition as we rapidly shift away from mobile and low-end consumer with growth driven by high power markets and specifically AI infrastructure. Non-GAAP gross margin is expected to be 39.7% plus or minus 100 basis points which at the midpoint represents a 20 basis point increase, reflecting a continued favorable shift in revenue mix towards high power markets and some additional improved scale.

As previously discussed, we are moderately increasing our investment in OpEx going forward to further accelerate our expected future growth. Non-GAAP operating expenses are anticipated to range between $15.5 million to $17.5 million. That concludes our formal remarks. Operator, please open the call for questions.

[Operator Instructions] Your first question comes from Quinn Bolton from Needham & Company.

Congratulations on the nice third quarter outlook. Chris, I can say I wanted to start, there's been a lot of noise in sort of chatter in the market that 800-volt architectures may be delayed, including confirmation, I think, that NVIDIA's cyber may have been canceled to be replaced by something as of now that's unannounced. But given some of this noise around 800-volt architecture, can you just sort of comment on what you're seeing in terms of adoption of 800-volt and whether there's any impact on your 2027 revenue outlook as a result of perhaps architecture shifting around?

Thank you, Quinn. This is Chris. And I would bet you would ask that question. So we provided in the early comments, the steps, okay? And I call that the inflection point. So first of all, I want to reiterate that thanks to the fact that we have both Gale and , we are able to grow ahead of the donor. Number two, you probably saw that what I call the infection 2 is the introduction of the 800-volt ethanol through the power sale car, right? And by the way, that studies as well for the plus or minus 400-volt which is used for the XPUs and the ASIC. And you can see that there's already a step, okay, in the usage and the step-up in the content in both GaN and SIC.

Moving to the third inflection point, which is I think what you're referring to so-called native where the DC/DC conversion moves down to the GPU tray. Of course, I'm not going to comment on NVIDIA and Cyboplan. I would -- I believe they've made a communication about their plan, and I would refer you to that. What I would tell you, though, is I think there is a misconception in the Edward being a digital switch. If you look at the inflection #2 is actually the start of the interval for the psychiatrack. That's number one, which will drive more sick and more game content.

Number two, even if you look at infection which is where the gun content really step up as you move the DCDC into the compute try. And you have no choice that you use gain because of the switching frequency. The 1 thing I would tell you is you have multiple GPUs, you have multiple XPUs, you have multiple platforms. What we see is that this ramp will happen in steps throughout '27, of course, accelerate in '28. But it's not a 1 thing, okay, and 1 customer. So the short answer to your earlier question, do we see that as a change in outlook as can the answer is no. And I think that goes back to multiple times, you heard me saying that adding both GaN is a strategic advantage for us to capitalize -- to capture content and is even more so today.

Got it. And then I was wondering, Chris, if you could talk about sort of applications for your new silicon paid JET product line that you discussed on the call. Is that mostly AI infrastructure, energy grid infrastructure, kind of what are some of the initial applications you'll target with the silicon shaped?

So thank you for that question. It's actually a very strategic decision that we made to expand our SAM. As I mentioned, this will add nearly $1 billion of SAM by 2030. I think on sums even referred to $1.3 billion up by 2030. This is essentially a product that is very well suited for safety critical applications. So the focus here is going to be both AI data center and the energy grid infrastructure. So you find it in applications like if use owing, of course, solid-state circuit breakers, anything that helps to protect as you move to higher power, the protection circuit protection and power protection has become a bigger thing. And I think it's a time that will actually accelerate in the future.

I'll just give you an example. I just met an SST customer, right? And we've been talking about Ultra voted for a while with them. Just the fact that we can offer 1.2 kV up to 3.3 kV, the sand that we could capture in that SST went up by 4%. okay? So this is a significant -- and thank you for the question. It's actually -- I'm glad you this, this is actually a significant decision that we've made to expand the portfolio with JF.

Your next question comes from Jon Tanwanteng.

Okay. First of all, congrats. And then second, I was wondering if you could talk a little bit more about the MagnaChip deal. Is that a volume or fix type of deal -- and when do you expect it to contribute would be this year next? And then after that, do you expect any more licensing to follow on the back of that as well?

So first of all, thank you, John, for the question. This is Chris. I appreciate the question. We just announced that partnership with MagnaChip which by the way, goes beyond 6, but we just announced it the portion. First of all, it's a validation of the technology merits and benefit of franchises MOSFET from Genetec technology that we've been in business for quite some time. The way you have to use it, this is not so much about the licensing. I mean, of course, it will, over time, play in our revenue stream, but this is not the prime objective.

Number 1 is expand our fab because per the press release we've made, MagnaChip is actually going to focus on markets that we don't serve. So it's actually going to augment our ability to reach more customers, more market and more SAM with our genetic technology. Number two, is it creates an opportunity for us to partner with Madathip in the foundry concept. -- as we talked about, as we see the huge demand ahead of us and the seek growing at a 60 to 70 CAGR in the context of center grid, I think having more opportunity for us to secure capacity is essential, right? So we're not creating competitor. We are creating an extension of Navitas, and we are very much looking for to the partnership in the years to come with Magnership.

Got it. And I appreciate that color. Second, could you possibly comment on the Wolfspeed litigation? What's going on there? What do you think your chances might be? And kind of what's the risk?

Okay. So I'm sure you understand that I cannot comment on the specifics of pending litigation. But what I want is to give every some context around the litigation. And I'll refer to the whole speed because you had a question about Walspeed, but I refer as well. to the Renesas litigation that just came last week, right? And the other thing I would say is everything I'm going to say is actually on public record. So number one, Wolfspeed already suit us because we stopped buying wafers from them a while back. Then they sued us or they sued 2 of our employees that worked at Wolfspeed in the past, including 1 that they had left, okay, in their cost reduction effort to iBank. Then they even tried to file and they felt a restraining order when third parties recruiters who are calling their people for job position we had online on the web, they felt.

And now they to us for patent infringement, okay, in both GaN and SiC. So in my opinion, this is the last step in campaign of arhassment and intermediation litigation and looks like a disparate move. Then 2 weeks later, just last week, okay, Wednesday, Renesas suites. I'm not sure it's clear for everybody, but I want to make sure to understand that based on the public record, Renesas would own up to 49% of Wolfspeed. So is all this a coincidence, the week before the earnings and all this confidence, I'll let you decide.

Then as I said, the timing is desire, okay? And curious, okay? We've been in GaN and SiC for more than a decade, and yet we just got sued by Wolfspeed. I let Renesas more than a year ago. in June 25. I'm coming up to 1 year anniversary in Navitas and yet we just got sued by Renaissance last week. All this the week before earnings. I don't think this is a coincidence. So let's face it, and I'll give you my view there. You don't start litigation like this if you are winning market share, your technology is superior, -- you heard today our financial results. You heard the momentum we are building. I give you the detail of the full inflection point we see for both GaN and SiC and the momentum we have with customers. We're making a lot of progress.

So sorry for the long-winded question answered, but I'll leave you with 2 things. Number 1 is what we filed in the 8-K when the Wolfspeed litigation came. We respect IP and technology. Actually, the company is a result of a decade of innovation coming from start-ups, okay, in both can GaN and SiC and we'll defend ourselves. Number 2 is we let everybody draw their own conclusion on why -- now Wolfspeed and their major shareholders are running to the cardhouse instead of competing in a fair way in the marketplace. And that's going to be my only comment on this case during that call.

Your next question comes from Madison Depalo from Rosenblatt Securities.

This is Marty Calling on behalf of Kevin Casey. Just in regards to the Magnatip partnership, what other technology licensing opportunities are you considering? And then I have a follow-up after that.

So we licensed to MagnaChip, as I mentioned, Mary, the Genetic technology. We always consider -- we're not in the business of licensing our technology. We had the business of serving customers and growing the top line of Navitas. And starting this multiyear growth journey, I talked about with the infrastructure. But we're always open to license our technology to partners and people we can partner with.

Okay. Great. And then you mentioned the record book-to-bill and backlog extending beyond 2026. How much of that is of the expected 2027 growth is supported by the committed programs versus programs that are already -- or are still in qualification.

Yes. So I'll start. This is Tonya. -- at. Thank you for your question. We are breaking out what percent is committed in 2027 or what percent relates to our backlog. What we can say is what gives us confidence is the various inflection points that Chris described in his prepared remarks and them coming on top of each other, so it's a compound growth effect. The fact that we have both GaN and SiC which are both critical to gaining content a few competitors have both and having both allows us to participate, like Chris said, in all of those inflection points and then also what gives us confidence is the number of programs that are moving through qualification and into production, including design wins and DVTs, EVTs and PBT.

Many it's a very good question. I have 2 things. Number 1 is, I mean, you probably saw that we directionally gave you a sense of beyond Q3, how the business is going to continue, right? And the reason we did that is despite mobile going down even faster than we talked about we're going to grow more than we expected. And we are surprised and I'm sure you are surprised by the momentum that we have in the business and the outlook that we have today. And that's prevote, as I mentioned. And it's a very important I think, to understand. Now this is not 1 program, as Tonya said -- this is multiple hyperscalers, multiple OEM ODM, multiple power level of the CPUs -- and that continues to be inflection #2, okay, which will be some high in '27.

So for me, what gives me confidence is this is not like there is a bit of a shift of the view. I think up to now, the view was the growth of Navitas will come from 1 large GaN big sockets that will come with the 800-volt transition native, high infection #3 and the tick will come from the grid. But this is very different. Today, what we see is across CDCs, CDCs, BBUs the 1 bolt in the compute trade, it's tens and 20s of programs. different programs, different board, different customers. Some of them are sick. Some of them are high voltage 6, some of them all Tivoli, -- some of them are GaN. Some of them are both sick and gas. We've seen in a couple of cases, that we have, especially for CDC PSUs and BBUs, both sick and game. So that's what gives us confidence, Marie. Of course, we're not going to guide on 2017. We already gave you kind of directionally how Q4 is going to look like, just to make the point that the transition to Navitas 2.0 and to be a port company is essentially 1 quarter ahead, okay, what I talked about 6 months ago. And that's all driven by this 800-volt and the sidecar rack acceleration that we see.

Your next question comes from Joe Moore from Morgan Stanley.

Great. In terms of the 800-volt side car, when you talked about mid-2027 timing, I feel like there's some side cars in the market maybe sooner. So can you talk about what's the progression for Navitas to penetrate that business?

I think you absolutely -- thank you, Joe. This is Chris. You're actually absolutely right. I think when I referred to the to the mid-27 -- it's actually really kind of when things accelerate. I think what you're referring to the sidecar rack earlier ramp is the plus or minus 400 volt which I think is also more attached to some ASIC and XPU. But you're absolutely right that we see, in particular, with CDC PSUs and DCD CPSUs and to some extent, BBU as well PAUSE that the Sadara 80-volt or plus eliponalvolt is going to ramp earlier next year, okay? But from a meaningful -- what I tried to give in the slide and the remark is trying to give a sense of the step functions of the inflection I think clearly, there's going to be an acceleration in mid of the year. So Q2, but we see program ramping associated to the sidecar in the first half of next year.

Very helpful. And then in terms of the other markets, you talked about infrastructure as a third exit in the year. Can you talk about what's happening on the performance compute and in the noninfrastructure electrification side?

So on the high-performance compute as the high-end computers are moving and accelerating the use of much higher power type of architecture, including even embedded GPUs. We see a raise of the power level of the PSUs, okay? I mentioned that in our earnings -- we moved from 65-watt, 200-watt type of charges. Now we have customers doing 200 plus 20 watts. So we see an acceleration in the GaN usage, and that's benefiting us. I would refer to some announcements that were made by large U.S. OEM in computing, for instance, that came up with a super high-end GPU-enabled notebook that basically includes a 280-watt charger, which is full of GaN with a significant content. At that point, you have about $5 to $6 of content of GaN.

And then when it comes to the even high level, we just released with a customer a 1,600 watt, okay, platform that basically helps to power super high the super-high gaming platform. So those are, of course, not as high volume but the content is so much higher, but I think it has contributed to us. And I think this was -- this business, as we mentioned in the last earnings has actually helped us to compensate and really kind of neutralize the move for mobile ahead of the AI data center growth, right, which I mentioned with Air instructure 1/3 of our revenue by Q4.

[Operator Instructions] Your next question comes from Tristan Gerra from Baird.

This is Tyler and on for Tristan. Drilling on the last question, what are your expectations for revenue mix between the high-end compute and data center exiting this year?

Yes. So I'll start. So we don't break down our revenue by our high-power markets, the 4 high power markets being data center infrastructure, the 2 of those combined being AI infrastructure, then performance computing as well as industrial electrification. But Chris did give more context relative by the end of the year, we expect the AI infrastructure to be 1/3 or greater of our total revenue by year-end.

I think the way you should think about this is basically, over the last 12 months, we pivoted from being essentially mobile exposed to essentially being nonmobile exports. And in the last earnings, I referred to mobile being insignificant by the end of the year. The reason why we kind of gave a sense about the year-over-year growth by the end of Q4 is to kind of really outline that it's actually even less than insignificant. So we're not going to get specific about the numbers here, but I think I said in my earning -- in my early script that basically, we are 1 quarter ahead of my expectation in terms of mobile being one.

So that gives you a sense, right? And then the other thing that we gave color is the fact that 1/3 of Q4 revenue is coming from AI infrastructure. And you can see really the erinfrastructure being the acceleration of our growth Q2 to Q3 and Q3 to Q4, which I think is why we came higher than the Street expectation.

Yes. And we've also said on prior earnings call and reiterate this time that AI infrastructure is growing at over 50% quarter-on-quarter, both in Q1 and in Q2, and we expect it to accelerate. It's accelerating every quarter.

Very, very helpful. We've heard price increases across the industry. Are you seeing this trend for your products as well? And does that vary across silicon carbide and can.

So we've seen price increase in silicon. I think you've seen that across the board and in other technologies like memory and so forth. I'm not going to get specific about price increase with customers. However, as I said before is, as tension come, you expect the pricing to go up. But right now, we're focusing on getting our customers to adopt this new technology and transition to the new architecture. So price increase in the core market -- and I'm not referring to the market we move away from, has not been so far focused on our side.

Your next question comes from Richard Shannon from Craig-Hallum.

Let me ask a couple of questions. First 1 for you, Chris here. When you talk about the 4 stages of inflection with an AI data center, is there any particular stages of inflection that you feel relatively more or less confident about the share you're going to get? And if so, do you have any -- any way to characterize where those differences come from like GaN versus silicon carbide or where you have both or anything else? And I recognize the difficulty in answering a question about stages in terms of time when obviously, the 4 stages a couple of years out here, but I'd love to get your sense on that, please.

So first of all, I think on Stage 1, it's happening now, okay? And we are very excited about the amount of program and really something I mentioned in the earlier remarks is the acceleration of the replacement of silicon by silicon carbide, okay? And as you move to a high density and a higher efficiency high power level, there's an acceleration there.

And then when it comes to stage 2 -- and we are in a very advanced engagement and situation with the customer. I mean, at this stage, this is not any more prototype, right? This is basically a large quantity system-level testing, system level liability, should it be an CDC at 18, 23,27,30 kilowatt or CDC15 to 40 kilowatts or even a BBU, right? So -- what I like about Stage 2 is that it's multiple platforms, multiple hyperscalers and multiple merchant power per hyperscaler. So it's a lot of program, which I think give us kind of fairly good confidence that we're going to be able to capture a share.

When you go to Stage 3, what I like about this is we move from -- this is 1 customer, 1 large GPU vendor flipping to 800-volt native as people call it, being now looked at not just be, not just in the GPU rack, but computer rack, but across multiple racks across multiple XPUs across multi politics. And I think the fact that we've been in GaN for so long, I think, give us a leading advantage -- and I think I'll refer to the announcement that we made, the partnership that we announced in the past with some GPU vendors or other hyperscalers.

And then Slide #4, I think the 1 thing I would change compared to what I said earlier Richard, is steps in before the big jump is SST, okay, when really the grid delivers you right on evolve. But what we see is a lot more application than SST. I referred to BSS last time. PCS and solar. And what I like is that, of course, the big jump is in '28 with the SST -- but really kind of we start to see some nice ramp in '27 as well with the other application, right? So I think it's hard to give you a where I think we're going to win more than the others. What I like is that we don't chase 1 thing here, okay? It's multiple presales, multiple oat, multiple merchant power, it's sick, high voltage, how riveted and GaN. So give me confidence that we've been able to capture share.

Okay. Great. Chris, for all that detail. Second question is for Tonya, on the OpEx here. A couple of questions. You got a little bit wider range than you've had in the past quarters, your $2 million worth -- and I may have also missed any dynamics of how to think about OpEx going forward here. But what's the variability or the size of the range? And how do we think about this over the next few quarters? Any seasonality, any other investment cycles? Or should we expect it kind of largely flat for a period of time?

Sure. Great question. And the way you should think about OpEx and OpEx expanding, and we talked about this in the last earnings call is relative to it be meaningfully less our top line growth and our revenue growth. So at the midpoint of our Q3 guide, that's a 28% revenue increase. And even at the high end of our OpEx guide that would be approximately a 10% increase. So meaningfully less as in the 1/4 to 1/3 range is how we think about it. But you're right, we see a bigger step up Q2 to Q3 than we're expecting going forward because as Chris and I both talked about in our prepared remarks, we've held OpEx relatively flat for several quarters and then are also meaningfully and purposefully investing opportunities to accelerate revenue, and you're seeing that culminate in our revenue and our programs.

We talked about investing in new R&D projects like the JFET, like ultra high-voltage, the 6.5 kV, 100k and beyond more customer support programs as we ramp in the data center, including application engineering. And then that robust supply chain to make sure we're already ahead of demand. So that's kind of how you should think about that. And we've been doing all of that while keeping OpEx flat in the past and having less of a focus on China market. So the first thing we did is make sure all of our resources shored up and focused and shifting toward R&D versus other OpEx. And then even within R&D, sure it was all focused on high power markets before we started investing again. So that's how I would think of it, but still meaningfully less than the revenue growth, and you see the revenue growth accelerating. So you see a little bit of an uptick in OpEx.

I'll add something, Richard, I think our focus and eyes on getting this company to get profitable, has not changed. So the focus is accelerated top line growth and enabling the business with OpEx increase as a fraction of the revenue growth to stay on path for being profitable. With the larger number of programs I mentioned with the multiple inflection points with the fact that we feel there's a big opportunity for us to expand our portfolio, which mean expand our SAM. We decided with the growth coming sooner in second half compared to what we had estimated 6, 9 months ago when I started, we decided to pull the trigger a lot faster, and that's a conscious decision.

That concludes our question-and-answer session. I will now turn the call back over to Chris Allexandre for the closing remarks.

Thank you, operator, and thank you, very many, for your interest and your question. I'll leave you with a couple of things, right? 5, 6 points, which I want you to take from this call. Number 1 is the transformation to Navitas 2.0 is exertional complete. By the end of the year, as we told you, we are back to year-over-year growth despite mobile massive headwind, 4 quarters of sequential growth, double digit and a complete change of the mix of the revenue with essentially all revenue by the end of the year being high power and mobile being gone.

So when I took that role a year ago, we talked about transforming Navitas. I think today, it is transformed and now the focus is that we execute the strategy, right? The transformation is working. I talked about having both gallons being super critical. And we talked about the benefit in the inflection points offering both. We talked about some platform using both GaN. We talked about the fact that this 1/3 of our revenue by the end, right? So this is all kind of and showing that the transformation is working. The 1 thing I want to also highlight is this not 1 customer. This is multiple hyperscalers, multiple merchant power multiple platforms, okay? We refer to CDC PSUs, which is the first infection. But I think we got a question earlier, CDCPSUs, DCDCPSUs, BBUs, SSDs multiple things, right?

So the way I do this is the AI is the catalyst of the large sand that we go after. We added $1 billion with Jeff. Now it's and the revenue transition to Navitas 2.0. So 20 is actually who we are, not who we're going to become. And that came 1 quarter earlier than expected to be nest with you and create to the team and the Navitas employee that did this amazing job to transition this company. And now it's about execution and operational discipline to basically be on the path of a multi-year growth journey and back to provide, which I mentioned.

And that's what I want to leave you with. So it is a very important quarter for us because it's not talking about performing. It's talking about transformed, okay, which is very important.

Ladies and gentlemen, that concludes this call. Thank you all for joining. You may now disconnect.

Thank you for standing by. My name is Tina, and I will be your conference operator today. At this time, I would like to welcome everyone to the Navitas Semiconductor Q1 2026 Earnings Call. [Operator Instructions].

It is now my pleasure to turn the call over to Leanne Sievers. You may begin.

Good afternoon, and welcome to Navitas Semiconductor's First Quarter 2026 Financial Results Conference Call. Joining us today are Navitas President and CEO, Chris Allexandre; and CFO, Tonya Stevens.

I'd like to remind our listeners that the results announced today are preliminary as they are subject to the company finalizing its closing procedures and customary quarterly review by the company's independent registered public accounting firm. As such, these results are unaudited and subject to revision until the company files its Form 10-Q for its quarter ended March 31, 2026. In addition, management's prepared remarks contain forward-looking statements, which are subject to risks and uncertainties, and management may make additional forward-looking statements in response to your questions. Therefore, the company claims the protection of the safe harbor for forward-looking statements that is contained in the Private Securities Litigation Reform Act of 1995.

Actual results may differ from those discussed today, and therefore, we refer you to a more detailed discussion of the risks and uncertainties in the company's filings with the Securities and Exchange Commission, including Forms 10-K and 10-Q. In addition, any projections as to the company's future performance represent management's estimates as of today, May 5, 2026. Navitas assumes no obligation to update these projections in the future as market conditions may or may not change, except to the extent required by applicable law.

Additionally, the company's press release and management statements during this conference call will include discussions of certain measures and financial information in GAAP and non-GAAP terms. Included in the company's press release are definitions and reconciliations of GAAP to non-GAAP items, which provide additional details. For those of you unable to listen to the entire call at this time, a recording will be available via webcast for 90 days in the Investor Relations section of Navitas website at www.navitassemi.com.

Now it's my pleasure to turn over the call to Navitas President and CEO. Chris, please go ahead.

Good afternoon, and welcome to everyone on the call and webcast. We appreciate you joining us on today's call. I'm pleased to report that Q1 is reflecting another quarter of solid progress and growing momentum on our transformation to Navitas 2.0, highlighted by the company's return to top line sequential growth.

For those of you that may be new or still coming up to speed on our story, I want to begin with a brief high-level summary of our ongoing strategic transformation and Navitas 2.0 vision. Over the past 2 quarters, we have meaningfully reaccelerated our pivot away from the company's historical mobile and low-end consumer business to focus the entire organization on high-power markets, where Navitas GaN and high-voltage SiC products can deliver long-term differentiation and value.

Today, we are singularly focused on 4 high-growth, high-value market segments, AI data center, energy and grid infrastructure, performance computing and industrial electrification. Our go-forward objectives are to rapidly achieve scale in these higher-value markets in support of driving sustainable and profitable growth.

Turning to an overview of the quarter. Our Q1 financial results demonstrated solid quarter-over-quarter improvement, and we observed growing momentum across our high-power markets and expanded customer engagement. Highlighting the quarter, we achieved the expected return to growth in Q1 with revenue increasing 18% sequentially. The renewed growth was driven by our high-power markets, which also represented a growing and larger majority of total revenue as we continue to reduce reliance on the company's historical mobile and consumer business.

Although far too early to declare victory, we effectively completed our realignment of the entire organization, and Navitas is back to growth, driven by our high-power markets. In fact, revenue from our high-power business grew 25% year-over-year with all 4 of our high-power end markets increasing sequentially in Q1.

The increased contribution from high-power market also drove a favorable mix in our overall revenue mix, resulting in improved Q1 gross margin. Consistent with our previously communicated expectation, we anticipate continued sequential top line growth and gradual gross margin expansion throughout '26. The ultimate success of our strategic transformation continues to be grounded in 4 pillars: market focus, technology leadership, operational efficiency and financial discipline.

With respect to market focus, we continue to see new technology adoption accelerating across multiple end markets and customers, both of which are increasingly driving towards GaN and high-voltage SiC solutions. Without question, AI is the primary catalyst driving this momentum and leading to the broadening adoption of high-power solutions across all 4 of our target end markets.

Collectively, this market represents a serviceable addressable market of $3.5 billion by 2030, split roughly 50-50 between GaN and high-voltage SiC with combined CAGR exceeding 60%. We are definitely focused on the largest portion of the TAM, which I'd like to refer as the AI infrastructure comprised of unique but related growth opportunity across the AI data center and the grid and energy infrastructure, each of which are fundamental to enabling the AI evolution.

Today, the aggressive increase in compute power density is accelerating GaN and SiC adoption in data centers, while the required modernization of the energy green infrastructure to support these data centers is driving increased need for high-voltage SiC. It is uniquely positioned as one of the very few companies that can claim deep long-term experience in both GaN and high-voltage SiC technologies.

We're also agnostic and readily offer customers the ability to choose the optimal solution for their application architecture. As a result of our proven capability in both SiC and GaN, we believe it allows us to address more of the power chain and ultimately capture with content per system.

Briefly providing the trends and opportunities specific to each of our 4 targeted end markets, starting with AI data centers. As a technology leader in both GaN and SiC power delivery, we support all major AI data center architectures with industry-leading power density and efficiency. Again, having both technology is a strategic differentiator and our ability to fully support a given customers' chosen approach translate into more opportunities across more applications and greater potential lower content for Navitas.

As conveyed at the recent NVIDIA GTC event in March, AI data center is rapidly evolving towards a HVDC architectures, leading to expanding content opportunity driven by the need for exponential power levels, increased density and top-tier efficiency. Our immediate focus remains on expanding sampling of our newest GaN and SiC product, enabling qualifications, preparing for scale ramp and supporting hyperscalers and OEM customers in their ongoing design and development efforts, spanning from AC-DC PSUs and DC-DC PSUs and affordable HVDC brick designs at higher power level capacity.

In grid infrastructure, we continue to advance active engagement across a series of new and existing customers with notable acceleration in design activity in the United States. AI remains a prominent underlying catalyst as all industry participants increasingly acknowledge the existing energy grid is not capable of supporting the projected future rollout of AI deployment. This market where technology and scale are equally important, represent a large and long-term secular growth opportunity for our current and future high-voltage SiC products.

Navitas GeneSiC technology position us as a leading enabler of the grid and energy infrastructure modernization efforts, providing customers with more reliable and higher density power through our recently introduced 2.3 kV and 3.3 kV modules and a road map to even higher voltage.

In performance computing, we are seeing sustained healthy adoption of GaN in higher power chargers solution for high-end laptops and mobile workstations used for gaming and AI development. Our opportunity in this market continues to be driven by the dramatic increase in power requirements with CPU moving from 15 to 30 watt to 45 to 80 watts in ultra-end AI notebooks with the integration of GPU requiring up to 120, 175 watts. As a result, we expect to benefit from growing demand and momentum in performance computing market application throughout '26 and beyond.

Finally, in industrial electrification, we are continuing to see customer traction in both GaN and ultra-high voltage SiC in high-performance applications such as DC-DC converter and megawatt chargers, industrial pump, motor control and heavy equipment electrification.

With respect to our second pillar, technology leadership, we remain fully committed to ongoing innovation in GaN and high-voltage SiC driven by focused R&D investments and demonstrated by expanding customer engagement and co-development projects. On GaN, we have continued to accelerate sampling of our 100-volt and 650-volt devices to more OEMs and ODMs. Customers pursuing the 80-volt HVDC architect today are testing GaN, and we believe most are doing testing with magnetized devices.

We are focused on enabling and supporting customers in this transition from silicon to GaN like we have always successfully done in our past. More recently, we have seen some customer internal reality, system-level testing on our newest GaN devices. During the first quarter, we continue to deepen our collaboration with OEM, ODM and hyperscalers, including demonstration of enabling new GaN architecture that feature high power efficiency and reability, which is leveraging Navitas's more than 10 years of GaN experience and expertise.

One of those highlights was our recent release of a 20-kilowatt 800-volt to 6-volt DC-DC platform using our latest 8x8 60-volt GaNFast test aiming at 97.5% efficiency. This platform solution was formally unveiled in March at GDC and showcased at NVIDIA MGX. As a reminder, we also previously released an industry-leading 800-volt to 50-volt AI DC-DC power fully GaN, 60 Volt and 100 Volt, delivering best-in-class efficiency and density. This respective platform are generating strong interest and prospective customer engagement due to their demonstrated ability to deliver the highest power density, efficiency and performance for next-generation AI data center architecture.

Today, our team remain focused on execution, including product delivery, qualification and preparation of targeting growth for GaN-based 800-volt HVDC architecture in 2027. On high-voltage stitch, we continue to strengthen our technology with a focus on high power density and ability, which represent both the primary market drivers and our key differentiators in terms of silicon and packaging.

Following the introduction earlier this year of our new industry-leading Gen 5 GeneSiC technology based on our patented French-assisted planer architecture. In March, we released our 1.2 kV Gen 5 SiC product tailored in packages to address the higher power density DCDC and ACDC unit in PSU application. We have since delivered samples to OEM and ODF, and they are currently being evaluated by most PSU vendors. Initial customer feedback has been excellent with report up to 50% increase in power density and greater than 98% system efficiency and improved cool.

Turning to operational efficiency. The prior restructuring action initiated late last year, which I discussed in detail last quarter, have been substantially complete. As previously mentioned, today, the entire organization and its resource are fully aligned to focus on the high-power market. This represents a substantial strategic repositioning from where the company was just 9 months ago.

Our team is moving fast and working very hard and their collective mitigation is impressive. Recognizing the tremendous opportunities ahead, we plan to continue adding selective engineering skills and competencies to accelerate customer support over the coming quarters. Also during the quarter, we completed our leadership transformation with the appointment of our new CFO, Tonya Stevens, who formally joined the team in late March.

We now have the full leadership team in place, including new leaders in operations, engineering execution, sales and marketing, business units and finance, all of whom joined the company in recent weeks and months from larger companies with strong track record in execution and scale. Importantly, this new appointed team and our employees have demonstrated strong buying and excitement for Navitas 2.0, and it's a privilege to lead this transformation alongside each other.

We also continue to make progress on our strategic technology and foundry partnership with GlobalFoundries non-GaN. We are confident this will enable our planned 8-inch pivot in 2027 for GaN manufacturing in the United States. At the same time, we are starting to build appropriate buffers with TSMC to ensure a smooth transition for all existing customers. Additionally, we have begun actively scaling our supply chain to support upcoming growth and demand, and we are leveraging AI internally across design and most of the functions to allow us to scale even faster.

Our fourth pillar is financial discipline, which we are committed to as we execute our scale-up plan and transformation to Navitas 2.0, a consistently growing and profitable high-power company. This includes remaining diligent with respect to prioritizing our investment in high-power program, maintaining leverage OpEx and focusing on high-margin, long-term engagement that build multinational customer relationships. We made significant progress in Q1 with our previous restructuring effort and full mine towards high power market now substantially complete. Going forward, we'll continue to drive efficiency across the organization and are committed to disciplined investments in the business, even as we target a much larger market opportunity.

Our focus remains on top line growth and margin expansion, driven by improving scale and mix of our high-power business in support of achieving long-term profitability. In summary, I am very pleased with the continuous progress and great momentum we have achieved in such a short period of time. We're taking further steps towards positioning Navitas as a high-power company. We anticipate continued sequential revenue growth in the second quarter and throughout the rest of '26.

Q1 was the first clear proof point and the growth in high-power market demonstrate the momentum of our Navitas 2.0 strategy. We also anticipate gross margin to steadily improve as volume growth drive better fixed cost absorption and our revenue mix increasingly favors the high-power business. Mobile contribution will continue to diminish this quarter and become insignificant by year-end. At that time, we expect our business and revenue will be defined almost entirely by high-power market, a transformation that positions us well for sustainable long-term growth and profitability.

Before I turn the call over to review our financials, I'd like to take this moment to welcome Tonya Stevens, our newly appointed CFO. I'm thrilled to have her join our executive team. She brings over 30 years of exceptional track record of financial leadership in the semiconductor industry, most recently at Lattice Semiconductor. I look forward to her valuable contribution as we grow the business and scale our operations to a larger, financially disciplined and profitable company.

With that, I'll pass the call to Tonya to introduce herself and review our first quarter financials and second quarter outlook.

Thank you, Chris. Before reviewing the financials, I would like to take a moment to introduce myself and share my motivations for joining Navitas. My corporate finance career spans more than 30 years and began with 7 years in public accounting. I've since spent the majority of my career in the semiconductor industry, including 17 years at Intel in Corporate Finance and the last 7 years at Lattice Semiconductor as Chief Accounting Officer and previously Interim CFO.

I'm incredibly excited to join Navitas for several reasons. The team is comprised of extremely talented and capable leaders and individuals who are laser-focused on executing the company's strategic objectives in a rapidly advancing and high-velocity environment. Together with its compelling technology portfolio, the company represents a pure-play GaN and SiC opportunity to scale up and capitalize on the substantial AI-driven secular growth in high-power markets. It's a privilege to be part of the Navitas leadership team, and I look forward to meeting many of you that I haven't met already over the coming weeks and months.

With that said, I will now review the financial results for the first quarter of 2026 and then discuss our outlook for the second quarter. Please note, unless otherwise indicated, I will focus my comments on non-GAAP results. A detailed reconciliation of all non-GAAP to GAAP financial measures can be found in our press release published earlier today.

Revenue in the first quarter of 2026 exceeded the high end of guidance, increasing 18% sequentially to $8.6 million on a GAAP basis. This compares to revenue of $7.3 million in the fourth quarter and $14.0 million in the first quarter of 2025. As Chris highlighted, the return to sequential growth was driven by high-power markets, which grew approximately 35% from the first quarter 2025 and now represents a large majority of total revenue as the company continues to reduce its reliance on historical revenue contribution from mobile and low-end consumer business. Notably, we expect high-power markets to continue driving sequential growth throughout 2026.

The higher quarterly revenue and improved revenue mix drove a 30 basis point expansion in gross margin, which improved to 39.0% from 38.7% in the prior quarter and 38.1% in the first quarter of 2025. The shifting mix of total revenue toward higher value, high-power markets and away from mobile and low-end consumer is key to our gross margin expansion strategy. We expect sustained gradual improvements in gross margin throughout the coming year.

Operating expenses for the first quarter were $15.0 million compared to $14.9 million in the prior quarter and $17.2 million in the same quarter a year ago. Operating expenses for the quarter reflect our commitment to focused and disciplined spending, particularly in SG&A, which created the opportunity to invest more in R&D projects quarter-over-quarter in support of our strategic pivot to Navitas 2.0 while keeping total operating expenses flat.

Loss from operations for the first quarter was $11.7 million compared to a loss of $12.1 million in the prior quarter and $11.8 million in the first quarter of 2025. In Q1, diluted shares outstanding was approximately $230 million, resulting in Q1 loss per share of $0.04 compared to $0.05 loss in the prior quarter.

Turning to the balance sheet. Cash and cash equivalents at the end of the first quarter 2026 were $221 million compared to $237 million at the end of the fourth quarter, and the company continues to have no outstanding debt. With respect to inventory, we ended the first quarter with $14.9 million compared to $13.3 million at year-end. The sequential increase in inventory primarily reflects our measured investment to support future anticipated revenue growth.

With respect to channel and distributor inventory, as a result of previous streamlining actions taken during the latter part of last year, we now have a significantly healthier channel inventory profile. Going forward, we are committed to disciplined monitoring and management of these inventories to ensure we are well positioned to respond quickly to end market demand. Overall, the balance sheet remains very strong and provides the company with an extensive amount of liquidity as well as ample flexibility in terms of working capital to execute our strategic objectives and anticipated growth.

Moving to guidance for the second quarter of 2026. Consistent with the company's previous communications, we expect continued sequential growth with revenue increasing to $10.0 million, plus or minus $0.5 million. At the midpoint, this represents over 16% sequential growth compared to the first quarter of 2026. Non-GAAP gross margin is expected to be 39.25%, plus or minus 75 basis points, which at the midpoint represents a 25 basis point increase, primarily reflecting the ongoing shift in revenue mix toward higher power markets.

Non-GAAP operating expenses are expected to remain approximately flat sequentially between $14.5 million to $15.5 million as we continue to emphasize disciplined cost management. Moving forward, we may choose to selectively invest in OpEx to accelerate growth at a fraction of the rate of revenue growth.

That concludes our formal remarks. Operator, please open the call for questions.

[Operator Instructions]. Our first question comes from the line of Tristan Gerra with Baird.

I know it's still probably a bit early, but would you be able to talk about the dollar content that we could expect per rack for silicon carbide on the first-generation 800-volt architecture? Then what type of ramp in content should we expect with Kyber for both silicon carbide and GaN?

Tristan, this is Chris. Thanks for the question. If you refer to our prior communication, right, we gave guidance in terms of content per megawatt because that's how the best way to kind of define the content we talked about for GaN in the range of $10,000 to $15,000 per megawatt, really driven by the massive 800-volt HVDC when the DCDC gets inside the rack, as we discussed primarily.

In the ACDC PSU, there is about $5,000 to $8,000 per megawatt, which is coming from both the higher power of those PSUs. If you refer to GTC, right, NVIDIA announced that at the end of the year, the PSUs, the ACDC are going to get to 18.5 kilowatts, which is much higher factor is if we look at the power level from today's PSUs, the ACDCs, which are in the range of 5 to 10 kilowatts to 18.5 kilowatt for NVIDIA, but even 25 to 30 for other hyperscalers, there's a ratio of -- when power goes up by 2, the SiC content goes up by 5. There is a non-linear increase, right? I'm not going to get specific in terms of content because it really depends on the architecture, 1 phase, 3 phase to 3 phase, but refer to the $5,000 to $8,000 of content for the SiC inside the center, which is mostly AC/DC PSUs and the mental model, which I just mentioned, which is when the ADCDC from, let's say, 5 to 10 kilowatts to 18 to 25 to 30 kilowatts, there's about 2.5x content acceleration compared to per rack.

Then for my follow-up, specific to silicon carbide, clearly, pricing has been coming down drastically in '24, '25. Given the ramp that you see, do you expect pricing to stabilize? I know you're going to be in the very high voltage. How different is that pricing dynamic there than in the lower voltage, but also do you expect at some point supply-demand balance in silicon carbide?

We don't participate, as you know, to the low-voltage SiC business in mostly industrial and EV, right? What we see is for inside data center, the ACDC mostly use 1.2 kV and above 65 sometimes and 1.2 kV and above, right? Where the driver today is more speed, reliability and density. Of course, this is a competitive market, and as the hyperscalers are driving more power and more PSUs and more PSUs per rack, there is quite competitive.

Today, what we see is what the customers are pushing us on is how we execute and how we help them to get to the best scalability and the best density of power, which I think save a lot more money at the system level than a cheaper device.

Your next question comes from the line of Madison DePaola with Rosenblatt Securities.

This is Maddie calling on behalf of Kevin Cassidy. You highlighted that GaN and SiC are both playing vital roles in AI power and that you guys are uniquely positioned to win both technologies. I know you mentioned this, but can you provide any more color on how having both capabilities is helping in customer discussions or design win activity in data center over your larger competitors?

Maddie, this is Chris. First of all, I think we focus on the high-power markets, right? We have 4 markets. Each of them have a different flavor of architecture and technology. If I refer to AI data center, it's mostly a GaN and SiC play. If I look at grid infrastructure, it's mostly a SiC play. Of course, high-performance computing is more GaN play and industrial is actually both a SiC and GaN play, right? If you look at the first 2, which is what your question is, right, if you look at the evolution of the architecture, so let's zoom out a little bit, right?

Today, in the current architecture, the traditional architecture is 50-volt bus bar where the voltage from the grid, which is 480, 400-volt ACs convert to 50-volt DC, right? That's mostly use SiC, okay? That's been going forward, right? The first step, and I think I referred to what has been announced at GTC, right? The first step is to the 800 volt is the introduction of 3-phase much higher power, which I referred to in my answer to Tristan. The first phase is most higher power 3-phase AC/DC, right, where you convert the 400-volt, 480-volt AC into 800-volt DC, okay? That's the first phase that's going to start at the end of the year, early next year, right? That on the AC/DC we use mostly SiC.

Now there is a DC-DC conversion to that. If you refer to what NVIDIA announced at GTC, there is a DC-DC top of rack converter, right, at 15 kilowatt for instance, both use either GaN or SiC. I think both is already right there, enabling customers to have a choice depending on the preference. What is very interesting is when you move to the next step, which is the second phase of the 800-volt DC architecture, where you get to, let's call it, high-density rack, megawatt rack will be Kyber for NVIDIA or more other high-density racks for the Googles of this world and the others, right? That's where you move the DC-DC conversion inside the trade inside the rack.

When you do that, you have no choice than to use GaN. Because the level of density, the level of power requirements make it impossible to use silicon, but also silicon carbide doesn't have the switching frequency. That's where you're moving to GaN. The fourth step is when you replace -- that's more on the grid side, when you replace the AC-DC PSU on the side rack basically by SST. If you think about this is a continuum of architecture change and evolution and having more offset to see the [inaudible] should it to be current generation, next generation, next, next generation and how the guys are evolving from current architecture to next phase of into high-voltage, high density, even down the road with the reorganization and the restructuring of the grid.

Our next question comes from the line of Quinn Bolton with Needham & Company.

This is Shadi Mitwalli on for Quinn. My first question is for Chris, but do you have any big picture takeaways from GTC in APAC in March, especially in regard to the direction of GaN versus SiC in 800-volt data centers?

My takeaway was kind of what I just mentioned to Maggie. First of all, we've talked about 800-volt architecture now for more than a year. It's happening. I think NVIDIA was very clear that they see at the end of the year, early next year, this what I call the first phase of the 800-volt HVDC architecture where you basically do the ACDC at a much higher level of power with SiC and then you do a DCDC where you can use Gan and SiC, but also outlining that as you move to next step, the move to much higher density rack is kind of enabling GaN content to move next. That's my takeaway from the GTC is 800 volt is happening.

Now keep in mind that there is -- we talk about NVIDIA here, but there are other hyperscalers. They might have a different path, they might actually go even faster to the next phase where you get the power, the DCDC part enabling directly on the train and in the rack, which will accelerate the GaN adoption. I would say I come out of GTC with a stronger conviction having both makes a huge difference. I think we talked about this before where I said Navitas is uniquely positioned because we have both SiC and GaN. I think it's actually very hard for a supplier to sit at the big table if you either have GaN or if you have a SiC. There's only a handful to not say a very few number of suppliers who have both. That's the key differentiation. That's my takeaway on top of the fact that [inaudible].

Then my follow-up is just on the product landscape for GaN. As you're sampling with hyperscalers, what are some of the key specs that matter most of them when evaluating GaN products? How does your portfolio measure up against those requirements?

What we said before is we've sampled both high voltage, so 650-volt GaN as well as mid-voltage GaN 100 volt. We've done that in different flavors of package, depending on the level of integration and density that the customers are looking. In the last quarter, we mentioned we've done the initial samples, since then, we've now delivered the final samples, which is basically the samples that will go to production. We are working with customers on -- they move from, I'd say, device level testing to board system level testing. The feedback we get is our technology as well as packaging offering is actually adequate to what they're trying to do.

Your next question comes from the line of Richard Shannon with Craig-Hallum Capital Group.

This is Tyler Anderson on for Richard. I was just wondering, could you talk about why customers would want to upgrade transformers that aren't connecting to data centers? Have you heard of any talks within the government to force the upgrade of transformers?

I'll start by the last part of your question. We have no knowledge of any forcing function or requirement for the government to move from traditional transformers to SST. What I would tell you is if you look at -- and I think we've made some slide in the past in our investor package, if you look at the transformers today are very kind of old school, so to speak. They are operating at a low frequency, which is in the 60 Hz. They have limited efficiency, which is less than 95%. They are heavy metal. They are very large and very weak. As you move to an explosion because that's what we're talking about explosion of rollout of AI data center, which basically pull on the grid a lot more energy, you have to install a lot more transformers. That's going to be, at some point, impossible if we keep the convention transformer. The move to SST is a bit of a necessity as we scale up and deploy the hundreds of gigawatts, in the next few years.

The other thing I would refer to is we keep referring to SST, but when we talk about grid and energy, this is going beyond the SST. SST is going to be the last step of evolution. Today, you have much higher level of power of transformers, megawatt converters. You have grid-type solar farms that are being deployed. There's a lot of grid type applications that are being deployed, which we see as a growing driver even in '26 and '27 ahead of the big acceleration of the SST, which will come really in late '27, early '28.

I'm also wondering if there's anything around the switching. I'm seeing something about -- and please, I understand I may be wrong on this or going down the wrong path, correct me if I am. Aluminum conductor steel transformers, I'm seeing things about them wanting to focus on the switching. Would you be able to benefit from that upgrade in the switching?

I mean, yes, you will. I think the grid companies have realized that the only way to make the grid, as I say, compatible with the acceleration of power is really to get to this new form of conversion, less conversion, less steps moving from super high voltage great DC to a form of electronification of the grid for lack of better. I think this will require and isolation transformers basically.

Then have you heard of any conversations around the lack of supply of transformers accelerating anything with your customers?

I have not, but I will not be surprised that the requirement for volume in terms of classic transformers and the dependency on metal and a few other things might actually play also in the expiration of the modernization of the grid.

[Operator Instructions]. Your next question comes from the line of John Tanwanteng with CJS Securities.

This is actually Jeremy on for John. Can you just talk a little bit more about the sequential improvement you're seeing heading into Q2, if that's mostly data center driven and if you're meaningfully ahead of where you thought you were going to be a quarter or 2 ago?

Yes. This is Tonya. Jeremy, so I'll start and let Chris add. Relative to your point in high markets, if you remember in Q4, we talked about high power being the majority for the first time in the company's history, and we talked about it being greater than 50%, mobile being less than 25% and the vast majority of the company last year.

Now in Q1, high power continued to grow. It was a large majority of the company, like you heard us say. Throughout the year, to your point, we expect it to continue to grow as a percent of the company. We exit the year almost an entirely high-power company and that being driven by what you said, the data center and the grid and infra, the AI infrastructure component of that. You saw in our press release and our discussions, high-power grew 35% year-over-year from Q1 of '25 to Q1 of '26, and we expect that growth to accelerate in the second half of '26.

Again, driven by both components, but the key catalyst is that AI component. The momentum is driven by all of the high-power markets, but particularly the AI infrastructure, and that's data center and energy grid.

I add something, Jeremy, thank you for the question. First of all, if you look at Q4 to Q1, when we grew 18%, we said, as Tonya said, that the high-power markets, grew as a percentage of the company, mobile went down. That means that the growth of high power was actually much higher, than 18%, the top line of the company and grew 35% year-over-year. Now we don't break down by markets. We don't -- we referred to kind of high power, but we also said in our script that all markets grew sequentially. As we see, this will continue throughout the year.

Now I'll give you one data point. Tonyia referred to and I referred that also in my script about AI infrastructure. What this means is we are combining within the high power, we're combining data center and grid. The reason why we do that is what I've noticed is the driver of the grid is data center. At the end of the day, you cannot look at AI data center and grid energy as 2 independent markets like computing would be. This is really kind of intertwined. That business grew 50% quarter-over-quarter from Q4 to Q1. That's the only color I'm going to give you.

As the company grew 16% -- sorry, 18% quarter-over-quarter, the combination of data center and grid infrastructure grew 50%. That's stronger than expected. You asked me where I think we were -- we are versus where I thought we're going to be. That's stronger than expected. The reason why it's stronger is that we all see it, it's an acceleration of rollout. We have not seen yet the content going up. I talked about the fact that content is going to go up. The content is going to go up because when you move from a 10-kilowatt PSUs to 18.5 kilowatt PSUs or even a 25 to 30 kilowatt PSUs, the ratio is 2x power leads to 5x content. The stick content and growth is going to accelerate.

Today, what we are seeing is just the growth of AI. Then next year, we're going to see even an acceleration of GaN as power gets -- the DCDC gets inside the rack. I think what we are seeing here with the 50% is that the AI data center is accelerating. I will also tell you, even though we don't guide by market that what we see today for Q2 and as a reminder, we are confident in our guide for Q2. We're seeing that AI infrastructure that grew 50% quarter-over-quarter, Q4 to Q1 is actually going to grow faster. That growth is going to accelerate throughout the year. That's before even the step-up in content.

Yes, I would say we are a bit ahead of where I think we're going to be. I look at Q2 guide with confidence, the benefit of being high power is we have longer visibility. We used to be in mobile where you get -- you're still chasing orders within the quarter. I think the high power market, in particularly data center and grid infrastructure are giving us a much longer visibility. I look at Q2 with confidence. We think, as we said before, that this growth will continue throughout '26.

One last follow-up. Any update on the use of cash this year and next in support of the growth ramp? What are your thoughts on when cash flow breakeven is likely to occur?

Yes. I'll take that one. Coming into Navitas and being new, when you look at the strength of our balance sheet, and I even referenced that in my script, a very strong balance sheet. We have over $221 million in cash and no debt at the company. That gives us a pretty long runway to support our working capital needs and CapEx flexibility. I'm confident we can execute the objectives and the organic plan consistent with what I said in the script.

Again, we remain focused on profitability. Like Chris said, we remain on track and maybe a little ahead of where we thought we would be to profitability. We're very focused on that. Nothing's changed in our thoughts around profitability and in fact, potentially accelerated a bit.

Jeremy, you can make the math. I mean, at today's gross margin, and today's OpEx, it will take us to be in the high 30s from a revenue standpoint to be profitable. Now we're guiding 10% for Q2. We said that we expect that growth to continue throughout the year. There is no reason to believe based on what we just discussed that the momentum that we are seeing in data center, grid infrastructure as well as the other hypermarket will slow down. You can extrapolate that to when we're going to be profitable. I'm not going to guide specific. What I will tell you is when we look at our business, both Tonya, myself and the leadership team is getting to breakeven is a key objective.

We're going to spend what we have to spend to optimize and to drive our growth, but being financially efficient, and make sure that we get to breakeven at some point is a key priority for us.

Your next question comes from the line of Quinn Bolton with Needham & Company.

Welcome Tonya. Great to have you on board. I wanted to follow up, Chris, you mentioned that at least on the 800-volt GaN opportunity, you've kind of moved from device level testing to board level testing. Kind of can you walk us through what the following steps would be to get to final production and sort of the time line if these higher power racks go to production, say, second half of calendar 2027, when do you think those designs would be sort of fully locked down? Did that happen at the end of this year? Or could that continue into 2027 in terms of the testing process?

Thank you, Quinn. You're very consistent asking the same question in the quarter, so I appreciate that. Nothing has changed really. I would change the answer, as you said, depending if you're looking at the first phase of the DC to the second phase. Again, for everybody to understand, the second phase is when the DC-DC conversion gets inside the rack. The big difference is in the first phase, you're designing AC-DC PSUs, DCDC PSUs. You're working with the hyperscalers, but really the implementation of that is at the merchant power, ODM, OEMs, the Delta, the Flex powers, the Vertiv, [inaudible] and so on.

We know where we are with those guys. We first delivered the samples, both the 1.2 kV SiC that we mentioned, the Gen 5 in the new package as well as the gas devices. We now have delivered the final samples, which I think is the sample that will get to production, which I think is important. For those boards, we are, as I said, moving from component level testing to standard testing. What does it mean? Well, the customers have done a couple of prototypes, they're optimizing the systems, the layout, the ELI performance, the efficiency, we are highly active and supportive of this with our application engineers and our field application engineers. That's kind of where we are.

The next step is once they've done some level of system testing, then they're going to do system reality and system validation at the next level. I would say for the first phase of the 800 volt DC since this is meant to ramp at the end of the year to earlier next year, I mean, we're going to get clarity very quickly. As I told you before, for me, I'm not going to comment on design and engagement with customers unless the customer wants to, but you're going to see the proof point in the backlog and as we go.

Now when it comes to the second phase, which is really driven by the hyperscalers, when the DC-DC conversion gets in the train inside the rack, mostly with GaN because there is no other technology that helps you to do this 800-volt 50 or 800-volt 12, 800 volt 6, inside the rack. I think today, we are still working with the hyperscalers and getting the ODM to be comfortable. One of the reasons why we're spending so much time developing those references that we've announced earlier this year, the 850 or 860 is that it gives comfort to the hyperscalers and the customers on how to deploy. It's only 6 to 9 months behind. If you ask me when we're going to get proof points of the in-tray GaN-based DCDC current, probably Q1 to Q2 next year.

Again, this is a duration. Customers, what I see is I measure my team and the engagement with customers in terms of the number of samples we ship 10 samples or 50 samples. When you get to 5,000 samples, it's not samples. It's quo build. My team on the amount of energy that the customer is spending on testing the technology and putting us in from an apps point of view in terms of helping us. I see that energy, that momentum, that number of samples are going up. That's why I'm comfortable in the momentum we are building.

However, as I said in the past, I think the proof is in the pulling, and we are not going to talk about pipeline. We're not going to talk about customer engagement unless the customer decide to, but we're going to refer to growth and outlook and guidance and backlog, which I think is what you should expect in terms of success.

I guess a follow-up just longer term, do you guys have a view? Or are you seeing customers push the intermediate bus voltage to 48, 12 or 6 in that 800 to step down? Do you think that 800 to 6 ultimately wins? Or do you think there's going to be a mix of different intermediate bus voltages across different hyperscaler platforms?

In the first phase, as we talked about at the end of the year, the bus bar at 50. I think you're referring to the true in-tray 800-volt HVDC. At this point, I would say it depends on the hyperscaler. I think you're going to see different flavors. You have seen that we announced GTC, NGX with NVIDIA and they are going to fix. I think that's kind of one of the trends we see. With that scale back to 12. It's a possibility. Some other hyperscalers might decide to scale 50. You might see some hyperscalers ramping next year with the in-tray massive volt HVDC keeping 50 volts as a bus bar, but moving the DC-DC conversion from top of rack to inside the rack.

The short answer to your question is I think we're going to see multiple flavors. Directionally, I would say the trend is the same, reduced number of conversion as you move to higher density rack, which means that the secondary voltage is going to go down over time.

With that, I will now turn the call back over to Chris Allexandre for closing remarks.

Yes. Thank you for joining us today. As I said earlier on, too early to declare victory, but what I see is the company is on track and accelerating the pivot and the transformation to Navitas 2.0. We have a lot of work to do still ahead of us. If you look at our momentum in high power, the growth in high power, the growth in AI infrastructure, which I mentioned quarter-over-quarter and the trend that we have ahead of us, I'm confident this will continue.

I want to close by thanking our team,s Navitas team a lot of work. This was a big pivot that we asked the team to go through moving from historical consumer low-end mobile type of business to high power. It's a big shift in terms of geographical coverage and in terms of product mix. I want to thank them for the effort, the reliance and the effort that we are putting into making that happen. Of course, our customers, okay, that are supporting us as well. Thank you.

Thank you again for joining us today. This does conclude today's conference call. You may now disconnect.

Great. Welcome back, everybody. I'm Joe Moore, Morgan Stanley Semiconductor Research. Very happy to have with us today, Chris Allexandre, CEO of Navitas. Thank you for joining us.

Thank you, Joe.

So first time we've done this as a fireside, you've been CEO for about 6 months. What have you learned? What are your priorities? How do you think about sort of Navitas 2.0?

So yes, as Joe said, I've been now CEO for the company for the last 6 months. The first thing I've learned is, and I think we all see it, is the size of the opportunity, okay? So the high-power market, AI data center, grid infrastructure and other markets are moving to a much higher level of power and density and efficiency, which we have never seen before. So it was very clear for me that we had to pivot the company to mobile. And that came and drove some of the pivot we've done, which mobile was historically where the focus and the revenue of the company was, and I wanted to prove it that even quicker. The other thing I've learned, which did -- was not clear to me before I joined is the importance of the grid, okay? You heard me talking about the grid.

Everybody talks about AI data center, which is very important, right? But without a change of the grid infrastructure, you cannot enable the size and the magnitude of the AI data center rollout that we're going to see in the future. And the last thing I've learned is the importance of where we're coming from. We talk about Navitas 2.0. I'm going to give you a bit of a sense of the priorities we have. But 2.0 doesn't mean 1.0 didn't matter, okay? In 1.0, we pioneered GaN in the mobile space, which was really kind of taking the mobile chargers to much higher power, okay? And we've learned a lot about this, okay? And we apply this learning into the data centers and other applications. And we had a very kind of very specific high reliability SiC technology. And now as we take it to much higher voltage for the grid, we kind of [ roll this out. ] In terms of priorities, we use -- when I rolled out the new strategy of the company, I used 4 pillars, which I'm going to keep using, okay, in the future to give kind of investors a bit of consistency on how we make progress.

Number one is the market focus, and I'm going to come to that in a second. Number two is the technology innovation, okay, and technology enablement. Three is operational efficiency and, of course, financial discipline. Market focus, I talked about the 4 markets we go after, okay, AI data center, grid infrastructure, performance computing, which is not to be ignored, and industrial electrification, right? The -- when we guide our Q1 and closed our Q4, I said that the revenue of mobile was now less than 25% of the company and high power is now the majority of the company. And as we're going to grow throughout '26, the growth will come from high power and mobile will go down to an insignificant level. That's something we are focused on in '26 to make sure that we pivot every resource we have in the company. Every dollar we spend is on the high power, but of course, we have to pivot the revenue.

Number two is technology innovation and enablement to the customer, right? And this is technology, GaN technology and SiC technology. This is product. We've sampled our 650-volt GaN for the 800-volt HVDC. We've sampled the 100 volt as well. We've sampled the new super high-voltage SiC modules that will enable the grid. It's also a system, okay? We've introduced 800-volt, 50-volt HVDC board, we announced a few weeks ago that was in partnership with one hyperscaler. This is all about enabling the transition, right? Operational efficiency is important. We are in the transition of moving our GaN foundry partnership from TSMC to GF. I think we'll talk about that later. I'm sure you have a question about this. This is very important. And last but not least, it's my eyes on getting this company profitable, okay? And to be profitable, you have to grow the top line and of course, grow the gross margin, which we will do with the mix, but keep the OpEx in a controlled way, right?

So those are kind of the priorities. That's what defines 2.0. 2.0 is clarity of what we go after, those 4 markets, clarity of who we are. We are a GaN and high-voltage SiC company. It's how we do it. We do it with speed because the customers we talk to, they want everything by us today and capitalizing on the 1.0 and all the experience we have to deliver value to the customers.

And the first indication of how decisive you are is this decision to exit mobile so quickly. Like it was tempting at all to sort of -- it's been clear for a while that there are headwinds in that market, but it's tempting to hang on to that revenue and use it for gross profit generation? Or what led you to the decision to do it that way?

So I would lie if I would tell you that on day 1, we made that decision to move that fast. But very quickly -- so what I did when I joined the company, the first thing I did is to take my backpack and go around the world, meet our employees and our customers. And everybody told me the same. They said, we need GaN because we can't do the AI data center without GaN, okay, partly the 800-volt DC. We need a grid and we need high-voltage SiC. We talked about that. And we need yesterday. And we're a small company. So every cycle we spend on mobile, every cycle we replace ourselves with the new mobile without innovation, which is just basically keeping what we have was a waste of engineering cycle.

So the decision was we're going to pivot faster. That's why we kind of slash the outlook. And today, I look back, this is what, 6 months ago, I look back, and this is probably my best decision because what's happening is, first of all, Mobile is now less than 25% of the company, will be insignificant by end of '26. People ask me what is the French English meaning of insignificant? I say less than 10%, okay? That's in my language -- that's translation.

What it means is it doesn't make or break the company. But the most important is the engineering and the time we spend. We flipped the company. So by moving away from mobile, the entire team is now obsessed and focused on getting the road map, the technology, the support, pivot from China to the U.S. We hired more people. We reduced also where we didn't have the right skills. So I would tell you, Joe, it was like doing it in a smooth way would have been missing the opportunity of growing faster down out of the gate. And it's proven by the guide to Q1. Q1 is up, okay? And it's all driven by [indiscernible].

And you're now in a position where you can grow potentially each quarter this year. Your confidence around that. And I think the company had a history of sort of having a tough time hitting expectations, your thoughts around that going forward?

So I mean -- so first of all, we're not in the business of guiding multiple quarters. We guided Q1. We're in a quiet period, but I would say -- what I would say today is we are confident in our guide and very confident in our guide. We also said that we're going to grow quarter-over-quarter multiple quarters in a row to give investors a perspective on how we are transitioning away from mobile and with mobile going down. So we're confident in that. And why we're confident? Number one is Mobile gives you 4 weeks visibility if you're lucky, okay? High-power gives you multiple weeks and longer visibility. So the customers we are dealing with now are giving you kind of midterm to longer-term visibility. So we have backlog in place for Q2, and we know what Q2 is going to look like. That's number one. And we have ramps coming up in Q3 and Q4.

Number two, I think the challenge that the company had is, to your point about the pivot is to spend too much time trying to rescue, okay? The anchor, okay? And an anchor is going to go down. You can scream as hard as you want, it's going to go down, right? So there was so many cycles trying to save the mobile [ soldier ] and that was very hard. So I think that led to the miss, always trying to think we're going to do better. I think by pivoting, walking away from backlog in Q4, as I said, was the wisest decision because now we are basically focused on this, on the right thing to do. So I'm confident we are not going to give a model to '27 before later in this year, but we reiterate our messaging in the last 2 earnings, okay? We're going to grow quarter-over-quarter throughout '26. The mix of mobile going down, high power going up and the scale will improve gross margin gradually, which I think will be the best proof points for investors looking at how we change, okay, and becoming a different company.

So you've given us a SAM, a 2030 SAM. I think $3.5 billion unless that's changed across the 4 markets. Can you give us an overview of your comfort level around that sizing? Pretty big number for a company.

Yes. Well, what's interesting is when I talk to many investors, they tell me I under call the SAM. So because rightfully so, some of our competitors, they call for $5 billion, okay, wide band gap SAM in the 4 markets we're talking about. For me, I don't want to spend too much time on it $3.5 billion or $5 billion doesn't matter, right? It's growing fast. It's a 60% CAGR for the next few years. The most important here takeaway is, number one, the make or break of that market is AI data center and grid infrastructure. If you look at the $3.5 billion, 3 of that is AI DC and infrastructure or grid infrastructure. Computing is going to be $400 million. It's a good business. We'll talk about that later, and industrial will be another $600 million. But 2/3 of that SAM is those two markets. And by the way, you can argue it's one market because what is driving grid? I mean grid, we've talked about since I was in high school.

We know grid...

We talked about upgrading grid, right? Edison would come back today would see the same transformers than we've seen 50 years ago. AI is a complete different use case. It's a catalyst that drives -- you cannot deploy the 250-plus gigawatt of AI data center in the next 5 years with the grid we have today. So for me, AI data center and grid is -- it's the same thing. It's inside data center or outside of data center. The other thing that is an important takeaway is -- and we have both GaN and SIC, okay? That might not be applicable to other companies that only do SiC or only do GaN. We have both, and we think it's important to have both is that the SAM is split in half between GaN and SiC, which is very counterintuitive. You asked me what I've learned. This is something I've learned, like I didn't expect that when I came on board. And the reason why it's the case is, number one, GaN is mostly outside of computing and industrial.

GaN is mostly 800-volt HVDC play. As you move to high integration, high-density GPU racks, you need to move to GaN, that's the 800-volt HVDC. That's a GaN play. SiC is playing also before that. So today, you have in the traditional AI architecture, you have to convert high-voltage AC to DC. Today, 50 volt, tomorrow 800 volt, right? That needs AC-DC, either inside data center or as what they call the side car. that needs SiC, okay? So SiC is in business today of AI, if you prefer. But what is interesting is it's insignificant today. We are sampling today. We're shipping today actually in that market, but it's not that significant. But there is a growth with the AI going. But what's happening is today, everybody shipped 5-kilowatt, 8 kilowatt, maybe 12-kilowatt type of PSUs.

In 6, 9 months, whatever we have 30-kilowatt type of PSUs because as you get more power inside the same rack, you cannot just increase the number of PSUs. And the content per kilowatt is not linear. So if you think about a couple of dollars of SiC content in a 5-kilowatt PSUs, you're talking about $200 per kilowatt of SiC content in a [ 30-kilowatt ] PSUs. So that drives SiC growth. And of course, the grid. What is also interesting thing to think about is -- and of course, nobody looks beyond 2030, but how much of the grid will be upgraded by 2030, not the vast majority. So SiC is half of the SAM in 2030, but the grid re-architecture is just the beginning. This is going to be -- and think about AI being a 5 years to 10 years build-out, grid is a 50 years build-out. So for me, and that came to me as a surprise, SiC has a long leg because of the grid infrastructure as well and playing on both sides, inside and outside data center.

And how -- there are a lot of competitors in each silicon carbide and gallium nitride. There's a smaller subset that have both. And there's obviously relevant ones, but how important is an advantage to have both of those technologies?

First of all, my -- I've been in 25 years plus in semiconductor. I always believe it's better to give the customer a choice than projecting your own religion. So I hear people -- 2 years ago, people said GaN is not going to make it in data center, oops, okay? That was the wrong call. Then people said it's going to be a SiC play, oops, okay? It's a GaN play. Then people said, well, SiC has no market outside of EV, oops, okay? So my point is we don't know what the customer wants. So I don't want to push SiC because I have SiC. I don't want to push GaN because I have GaN. We go to customer and say, what do you want? And we enable -- so that drives a completely different conversation. If I come to you, Joe, and say, what do you want? I have both. That drive a different conversation if I come and say, well, I have water, so you better drink water, right? So for me, this is very good to have both. That's number one.

So we are not challenged to be religious, okay? Number two is it's a continuum. You can look at grid, which is how you convert hundreds of watts -- hundreds of volts of AC into 50 volt or even 800-volt DC. And then you can look at how you convert this to 50-volt DC all the way down to GPU. It's a continuum, okay? You -- it's different customers, but they all talk to each other. And I think NVIDIA, in particular, has driven a complete ecosystem to make sure that you have to think about this as a continuum, right, because it's the same power chain. So I think both is actually very helpful. And as you said, there are not many vendors that have both, in particular, the ultra-voltage SiC. SiC, I know SiC got bad press, okay, because of the EV and it's hard to make money in the low-voltage SiC. That's why we don't participate. But it's a different game to do 1.2 kV, 1.7 kV or even 2,000, 3,000 volt, SiC for the grid. It's a different spot. So I think both is actually helpful.

Okay. Great. So maybe we could look a little bit more of each of the end markets, starting with the biggest AI data center. Can you talk about the shorter-term opportunities that you see there? And then longer term, the focus running 800 volt?

So everybody wants to talk about AI DC, right? So -- but I'm not surprised you start with that. So as I said, there is -- everybody talks about this change, which is the 800-volt HVDC. And this is not a GPU play. This is a rack play, okay? Most people ask me, are you attached to this GPU or this GPU or this [ TPU ] or this [ X DPU? ] I said it's not about the [ TPU or X DPU. ] It's about the rack. So if you look at NVIDIA as an example, right, the Oberon to Kyber. What is the difference? It's not just the GPU. It's number of GPU per rack. When you get to a point where you're compounding so many GPUs in the same rack, you cannot use traditional way of converting power. Too many conversion means more loss. More loss means less efficiency, okay? And the whole system doesn't work and you have less space. So the #1 big transition is the transition to Kyber for NVIDIA or any form of high-density rack, XPU.

And I say XPU because we talk about GPU, but you can talk about TPU and any form of XPU, right? So that's one. And that will drive the 800 volt all the way down to GPU. So here, you basically convert the grid to 800-volt DC, that's the entrance into data center. And then you go all the way. And you can go to 50 volt first, maybe you go down to 6 volts later, okay? Or maybe you go all the way down to 6 volt because it's -- it's obvious less conversion means less loss. So that's number one. And that's really a GaN play. The other thing is the grid is still AC. You still need to convert AC to DC. Two ways to do that. One is through the grid upgrade, which is the SSTs and all the things we're going to talk about. The other one is what some of the hyperscalers have called to replace the PSUs, the side car, okay, which is those rack of PSUs that convert, let's say, 480-volt AC to 800-volt DC, right? That's full load of SiC, right? So I don't want people to think that data center is just a GaN play, okay, both. But the 800-volt HVDC is a GaN play, in my opinion, and the PSU is a SiC play. That's for data center, right?

And if you're me, how do you pick winners? So there's 14 suppliers to NVIDIA power supply business on 800-volt. How do we go about? What are the milestones along the way that show us that Navitas is going to play there?

So first of all, revenue is the source of truth, okay? That's how I grew up in my 25 years. You're never going to hear me talking about pipeline and all this what I call monkey money, right? So because on Saturday, when I go and do grocery shopping, they only take dollars, okay? They don't take pipeline. So outlook, revenue guide, this is what matters. But you're right, how do we know where we're going to win business? So it's a multi-doors type of approach. Number one, you have to validate that you know what you're talking about. I hear some of my competitors, and I respect everybody, they say, we're going to have GaN. Well, okay, we're going to have GaN, but the customers need GaN today. And I'm not leaving a company that's been in GaN for 10 years.

I can tell you the last 10 years of experience makes a difference, okay? We're not going to develop a new super high-voltage SiC. We have it today. So validating that it works helps when you have it today instead of having in the future. Number two, once you prove it works, and it's a system play and it's a device play, you have to basically get chewed by the customer, okay? They're going to take high temp, low temp, they're going to go all the corners. That's validating that the device works. And then you have to go to rack level validation, okay, which is not like device level, but rack level. And then you get to supply chain and where you produce and do you have like a reliable multi-source, good Tier 1, Tier 2 OSATs and foundry. So it's a multistep approach. People ask me, when do we have the ChatGPT moment of Navitas that you have won 800-volt DC.

I said we're going to see it in the order book. It's hard to tell you when -- and by the way, it's not me who's going to comment, it's the hyperscalers and the OEM, right? I don't like long-term planning like this about business. But I would tell you, people underestimate how much having done that before makes a difference. If you think about GaN has always been this esoteric technology that nobody is comfortable with, right? It's hard. It's not as simple to drive than silicon. And we're helping customers. When we released the 800-volt, 50-volt board a few weeks ago, which was the highest efficiency at 10 kilowatt. This is to help customers. This was done with a hyperscaler to help customers to see how to do it, right? If you've never done that before, good luck. That's my point.

Yes. Okay. And then you mentioned the similarities between grid infrastructure and AI, but can you talk about that opportunity as well?

Yes. So the similarity is what's driving the use case, okay? So I joke -- I mentioned Edison, but if you would come back and look at the grid today, it's the same grid we had 50, 100 years ago. Huge transformers made of coil and metal and operating at 60 hertz. I mean it feels like my first year of college, okay, in electrical engineering. This doesn't work, okay, with AI workload. So -- why they are driven is because the AI is the use case. EV was not even the use case, okay, to change it. AI is the use case. You cannot deploy remotely half of what we're trying to deploy with the -- so first thing I would tell everybody is don't think grid is slow. That was slow, okay? That's not slow, number one. Number two is that it's really a U.S. play. And we will come to U.S.-centric and focus on national security. Europe, energy is very regulated, okay? U.S. is deregulated or more -- less regulated.

And I think the hyperscalers are involved in how the grid gets redefined. I talked about necessity being the force of nature, okay? If you have to change the grid. So I think we see an acceleration of investment development of battery energy system, which is going to be a large market, SST, kind of moving those huge transformers, weighing 2 tons and made of coil to electronification of the grid for lack of better terms. We see solar -- grid-level solar farms like huge stadium level type of thing. We see megawatt chargers and [ DC-DC ] converters, see many applications and they need super high voltage SiC, okay? This is -- you don't go there with silicon. You don't even go there with your 650-volt SiC that you use in an EV, right? That doesn't make the cut. But what is interesting is those guys are moving super fast.

And you're going to see interim step. You're going to see first couple of megawatt SSTs, bundled with battery energy systems. Then you see high single-digit megawatt type of SSTs coming up. And as I said, it's going to take 2, 3, 4 decades before every transformer that we hear blowing up every other day in San Francisco is going to be replaced.

Yes. Okay. Great. Third category, industrial electrification, I guess, more of a catch-all around robotics and things like that. But what opportunities are you excited about there?

Yes. It's -- catchall is a bit of -- I know it's not dismissive, but...

I don't intend to be dismissed.

I don't know. I always use it with my team and my team say, don't say it -- it's -- the reason why we call it industrial, it's a mixed bag of many applications, okay? And the one thing I would tell people, which is also a surprise to me is -- we've talked about power efficiency for the last 20 years. Everybody always said we're going to be more efficient, but nobody cared. Now this is table stakes. So we have customers doing pumps. They want to get the next level of efficiency. We have customers doing high-power converters. They want to be more efficient. So it's almost like the energy has become the currency like one of our famous leader in the base said, and I woke up everybody that says, you cannot continue to waste power. And how do you waste power is heat.

How do you basically get it out by being inefficient. So the table stake has moved -- being 90% efficient is not good enough. So we have customers coming out of the woodworks, doing, as I said, high-power pumps, not something I would have guessed GaN would go, right? And they come and say, oh, we need to use 650-volt GaN in this super high power couple of kilowatts type of pumps, right? So it's a catch-all, but it's -- the message is, I think we see the beginning of a transition to high-power technology. Customers have been reluctant because it's reliable, it's difficult to implement. They're not ready. They could get done with silicon. And now I think this is changing, right? And I think that's what's happening.

Okay. Great. And then the fourth market, performance compute. Is there a risk that you see some of the same stresses that you saw on the mobility side in that market?

Everybody asked me that question, okay? So first of all, I would say you're right, but we'll deal with that in the future. So in 2030, the SAM that we talked about, $3.5 billion, computing is $400 million. It doesn't make or break the company. But in the short term, and I think in the next 3 to 4 years, what we see is -- and I'm talking about the high-end computing. I'm not talking about your $300 notebook. I'm talking about high-end portable workstation, high-end game consoles, AI notebooks, okay, that are like $1,500, $2,000 and above. They need high power because unless you have a big backpack, you're not going to carry a 250-watt or 300-watt charger made of silicon. I mean we all got used to carry those big bricks, but it's painful.

So -- and those are 100, 200 watt, 150-watt chargers. It's not going to work at 250. customer we talk to, they want 250, 300, 350 watts. That's just enough to drive the move to GaN. So what drove the commodity decision of mobile? Mobile was good up to a point where the customers say, okay, I have 100-watt charger. It's small enough. I don't want it smaller. I don't want it to be higher power, so it's going to be cheaper.

Computing is not at that place. Today, when we do a 250-watt charger, they say, okay, can you do 350 in the same footprint? And now they say, can you make it smaller? So go back to what mobile had seen in the past. It's much higher content. It's higher margin, too. I think it's in 3, 4 years when every charger out there is tiny and then we'll see commoditization, but I don't think it's relevant today. And they do, it was basically a call I made. I said we're not just going to talk about how big things are going to be in the future. As long as we can make good money at good margin and we serve customer innovation with GaN, why not?

Okay. Great. So your long-term growth rate, you've talked about sort of a 60% to 75% SAM growth. If I have the numbers right, I know they've changed a little bit. But is that kind of in the right ballpark? And how are you framing that opportunity?

Yes. So it's the right ballpark. As I said, some of my competitors, I talked to one big competitor on Monday, say, why do you under call the SAM? I'm like, okay, $3.5 billion is big. So for me, it doesn't matter, okay? It's going to grow. What matters is what makes or break the SAM, right? The one thing I would call out to everybody is if you look at the SAM, people ask me, why do you call SAM and a TAM? I said, the reason why I exclude TAM is exclude EV, I exclude mobile. I don't want to confuse anybody. I mean the SiC TAM today is in the $6-plus billion. Remove EV, low-voltage industrial for SiC, you're down to the hundreds of millions. So for me, I call about the super high voltage, ultra-high-voltage SiC in the context of what we do. That's the $1.7 billion to 2030. Same with GaN.

I want -- we're working away from mobile. I'm not going to include mobile in the SAM. That was misleading. We don't want to support EV with GaN. I mean GaN will have a life in EV, mostly in OBC. No offense to everybody in EV. We have customers in EV today that we support, and we're going to continue to support them. But it's not where I put the R&D because it's a long to revenue. And what I would say is the pain to revenue ratio is very high, too high. So I think we have our hands full, okay? So we just removed that from the SAM, and that gives team the clarity of what we go after.

Okay. Okay. Great. You mentioned on the supply side, partnerships with Powerchip, GlobalFoundries, TSMC's decision not to build GaN. Just how do you think about that supply chain? How robust do you think that is?

So we've been -- and we're still in business with TSMC on the GaN side. They've been our partner. We pioneered GaN with them, a great partnership. It was my first partner meeting when I flew -- I took the job a week later, I was in Taipei, winning TSMC. And they are helping us in the transition. So I give them great partners. Actually, as a matter of fact, I'm sure you've seen that they licensed their technology to GF, okay, as part of the GF deal that we've done, right? My predecessor and the team had started a partnership with TSMC. We sampled the first mid-voltage GaN with TSMC. It's in the end customer. We'll continue that, okay? We don't want to create more disruption for the customers. But this is a big business we go after. It requires scale. It requires 8-inch. It's not a nice to have to have a fab in the U.S. in national security application like grid and AI data center.

So we quickly met with Tim and his team. They have GaN capabilities. They have GaN technology. They invested in GaN. We have long experience in process, in device, in application, in systems. So combining both was for us, the best thing of accelerating GaN production in the U.S. That's why we called it a strategic partnership. It's not a buzz word, and we call it a technology partnership more than a manufacturing partnership, right? So there's no direct flight between my site and Burlington, but there's a lot of back and forth between the -- between the 2 sides and the team are working around the clock. We'll sample to customers at the end of the year. We are open book with the customers, large customers on when they're going to get samples, what they're going to get. We're going to be in production next year.

And you always look at what are the best decisions you've made, right? You asked me about mobile. This was probably one of the best decisions I've made. This is one of the best decisions I've made. I feel like it's going to be giving U.S. foundry [ COUS ] like we have in SiC, okay, for the -- and Tim and his team have a great team and a great fab.

Well, Tim will be here on this exact stage tomorrow, so we'll have a conversation. Yes, I will. Maybe we could pause there and see if there's questions from the audience.

If not, the move that you made to consolidate your distribution around Avnet and WT, how are you thinking about that? How do you pursue -- you have this narrower focus now. How do you pursue those opportunities?

So I mean, Avnet and [ Wintech ] are great partners, okay? I worked with Phil and the team in the past. I worked with Eric Chang and his team in the past. it's kind of not the most important. What is the symbol of this is the pivot we've done from a go-to-market point of view. So if you think about 70% of the company was mobile. What does it mean? Well, that means your company center of gravity is in China and Korea, okay? Because I don't consider the other large company in the U.S. to be a mobile company. They are a computing company. So we work with them. We continue to work with them, but they are a computing company.

So distributors kind of tend to gravitate where the business is. When you move to serving the hyperscalers, the Deltas, the Eatons, the big ODM in Taiwan or across the world, this is a different go-to-market, right? You need more people. The center of application is to be in the U.S. and less, okay, in China. You don't need one distributor per customer, okay? I think in mobile, you tend to have one distributor per customer. I mean, we've said it publicly, when I came on board, we had 40 distributors, 4-0. For the size of our company, this is one distributor per customer. So that was a bit too big.

And no offense against the choice that we made before, but it made sense in the context of many ODM, OEM mobile guys. But when you go to a few hyperscalers, a few OEM, a few ODM and the grid guys, they also kind of different play, right? So it's a go-to-market change that we simplified. And I talk about speed all the time, which is, I believe, the foundation of our culture. As a small company, we have to be more innovative and we have to be faster. If we come after the big guys, it's not acceptable, right? And for me, Michael and the team were an aspiration. I was in TI when we look at those guys coming out of nowhere. And everywhere was going, they were first, okay? And we looked at them still with a form of they are too small. But when you're small, you have to be fast, okay? And to be fast, you have to have less baggage, okay? So having less distributors, a simpler go-to-market is a way to simplify what you carry and get faster.

Great. Well, I don't know if you have any final remarks, but we can wrap it up there. I mean, a very impressive view of the things you've done with the company and the decisiveness that you've approached this...

No, the one thing I would say is I talk about the 4 pillars, okay? And this was what I wake up in the morning, I look at those 4 pillars, okay? And we're going to continue to update how we make progress. It's a marathon at the speed of a sprint, okay? It's a marathon because it's -- it's a multiyear transition. What the AI guys and the grid guys are trying to do is a complete revolution, but they are trying to do it at consumer speed.

So I'm very clear to the team, let's walk before we run. Let's be clear, let's execute one step after the other. And quarter in, quarter out, we make progress, and that's why we're going to keep telling investors and you guys and everybody to show that we are walking the talk. And that's what we paid for.

Great. Well, thank you so much. It's great conversation.

Thank you.

Good afternoon, and welcome to Navitas Semiconductor's Fourth Quarter 2025 Financial Results Conference Call. [Operator Instructions] As a reminder, this conference call is being recorded today, Tuesday, February 24, 2026.

I would now like to turn the conference over to Brett Perry of Shelton Group Investor Relations. Brett, please go ahead.

Thank you, operator. Good afternoon, and welcome to Novitas Semiconductor's Fourth Quarter 2025 Financial Results Conference Call. Joining us on today's call are Navitas President and CEO, Chris Alexandre, CFO, Todd Glickman. I'd like to remind our listeners that the results announced today are preliminary as they are subject to the company's finalizing its closing procedures and customary quarterly review by the company's independent registered public firm. As such, these results are unaudited and subject to revision until the company files its Form 10-K for the year ended December 31, 2025. In addition, management's prepared remarks contain forward-looking statements, which are subject to risks and uncertainties, and management may make additional forward-looking statements in response to your questions.

Therefore, the company claims the protection of the safe harbor for forward-looking statements that is contained in the Private Securities Litigation Reform Act of 1995. Actual results may differ from those discussed today, and therefore, we refer you to a more detailed discussion of the risks and uncertainties in the company's filings with the Securities and Exchange commission, including Form 10-K and Form 10-Q. In addition, any projections as to the company's future performance represent management's estimates as of today, February 21, 2026.

Navitas assumes no obligated to update these projections in the future as market conditions may or may not change, except to the extent required by applicable law, Additionally, in the company's press release and management statements during this conference call will include discussions of certain measures and financial information in both GAAP and non-GAAP terms, included in the company's press release of definitions and reconciliations of GAAP to non-GAAP items, will provide additional details. For those of you unable to listen to the entire call at this time, a recording will be available via webcast for 90 days in the Investor Relations section of Navitas' website at itassemi.com. And now it's my pleasure to turn the call over to Navitas President and CEO. Chris, go ahead.

Good afternoon, and we appreciate you joining us today. I'm pleased to be hosting my second quarterly conference call as Navitas CEO. We closed out the year with a positive fourth quarter as we continue to accelerate our [indiscernible] to Navitas 2.0 and align the entire organization to focus on addressing high power markets. In fact, it has been energizing 5 months as I joined the company, and my conviction in our industry-leading GaN and high-voltage 6 solution has only grown stronger and our strategic pivot is on the right path to [indiscernible] the company to the next level. Before providing comments and update specific to the quarter, I will briefly reiterate several key events on our previously communicated strategic transformation and our vision to what we call Navitas 2.0. First, we're accelerating our pivot away from the company's historical mobile and lower consumer business to focus on high for markets, where GaN and high-voltage IC products can deliver real penetration and value through higher at least [indiscernible], efficiency and availability.

We are laser-focused on 4 high-growth, high-value market segments, AI data center, energy and green infrastructure performance computing and industrial electrification. Collectively, this segment represents a service addressable market of EUR 3.5 billion by 2030, split roughly 50-50 between GaN and high voltage 6 with a combined CAGR of more than 60%. although the largest portion of this $3.5 billion time are within AI data centers and grid and energy infrastructure, I want to emphasize that AI is a shared underlying catalyst across our 4 target markets. driving a rapid acceleration in terms of reactor infrastructure, customer expectation and the adoption of the new high-voltage technology.

While leveraging our proven 10-year track record as a pioneer of Ghana scale, having shipped over 300 million net gain devices, coupled with the deep expertise in system and application as well as our leadership in high reality, high-voltage sick to our genetic technology. The end goal of Navitas strategic transformation is straightforward to rapidly penetrate secure expanded customer engagement and achieve scale, resulting in a more sustainable, consistent and future profitable growth for Navitas.

Turning to a brief recap of our fourth quarter results. As initial progress of our pivot to Navitas 2.0, we've completed a realignment of the entire organization, both in terms of skills and geography to focus on addressing high power markets. This includes fully redeployed organizational resources, or roadmap and focus accordingly. Revenue in the fourth quarter and at the high end of our guidance range at $7.3 million, coupled with the fourth quarter being the first time that high power market represents the majority of our total revenue.

We remain confident that the fourth quarter was the bottom. Notably, our mobile business declined sequentially from the majority of revenue in Q3 to less than 25% of total revenue in Q4. We expect mobile to continue going down as a percentage of quarterly revenue and become insignificant by the end of '26. Also consistent with our comment last quarter, guiding to quarter-over-quarter growth for Q1 and anticipate continued sequential growth throughout 2016, driven by increasing sales traction in the high role market. Over the last several months, as part of expanded needs with customers and partners, have seen numerous proof points that the new technology adoption is [indiscernible]. AI is a catalyst changing the game across markets. Existing technologies and architectures are no longer sufficient. The industry is moving faster than it ever has in terms of technology adoption with customers clearly moving to take advantage of gun and high-voltage technology.

As previously mentioned, AI is a primary catalyst that's driving momentum and borrowing the adoption of high-power solutions across all 4 of our target high power markets. Every interaction with customers has confirmed the market is undoing secular change, and that AI is sparkling revolution we're focused on. This impelling inflection point in architecture design and technology adoption is highly favorable to GaN and high-voltage Tech, putting Navitas 2.0 at the center of this revolution. As outlined in our last call, the Navitas to transformation to a hyper company is being backed by decisive actions and grounded in 4 pillars that include market focus, technology leadership, operational efficiency and financial discipline. Let me now review with you the measure of progress that we've made in each of these areas since our last earnings call. Starting with market focus. As I mentioned earlier, we're sharply focused on the high-power market of AI as center, energy and green infrastructure, performance computing and industrial extrication. In AI as centers specifically, Navitas is uniquely positioned as one of the leaders in GaN and high-voltage SEC, supporting all major AI data center architectures. The density of compute power which required a higher efficiency and poor obesity.

It's driving the acceleration of CAM in next-generation data center. During the quarter, we've accelerated samplings of product and solution delivery with our on Volcan and 6 Valgan targeted at AI glass center, onefold HVDC and 48-volt IBC HD box architecture. Samples are currently available in different package sites, and are being evaluated by more than a dozen customers. More recently, on February 9, we announced our breakthrough 10-kilowatt DCDC design platform. This is an all game 10-kilowatt, 800-volt to 50-volt DCDC platform, which employs advanced 650-volt and onboard a fast in a 3-level average architecture with synchronous actification. This platform has delivered a 98.5% peak efficiency, which we believe is the best in the industry so far. This full [indiscernible] package design platform achieved leading power density and support plus or minus 400-volt BDC standard for AI [indiscernible], this is a great example on how Navitas is able to leverage our 10 years of GaN and system expertise. We're setting the benchmark for scalable, high-performance AI infrastructure. Our product portfolio enables unprecedented power density to support rapid large-scale expansion of Aida center, while also allowing hyperscalers an OEM, the ability to maximize compute density and reduce energy loss in the top of the [indiscernible] AIU club.

On the 6 front, we are very active customers in their ACDC PSU designs for current AI dacite architecture with our latest 2.2 KV [indiscernible] devices, leveraging our latest fifth generation genetic technology announced earlier this month. This product brings improved filed merit and best-in-class thermal behavior, the [indiscernible] QD pack packages that are being well released by customers. In the grid and energy and paper market, the energy grid is in the process of a major transformation and modernization to support the AI catalyst, but also overall growth in segment. This is not a short cycle, but rather a multi-decade secular and sustainable trend that will transform grid and energy infrastructure. As a result, we are seeing an acceleration in the design cycle here as well.

We are leading this effort with our new ultra-high voltage 2.3 kV and 3.3 kbits module and road map to even higher voltage. We are not in evaluation with over 15 OEMs globally, mostly in U.S. and Europe with notable acceleration in the U.S. In percent computing, we continue to see increased gun adoption in high-power chargers and 4 units for high-end computing and AI notebooks. The placing silicon. We have more than 15 projects in production, and [indiscernible] twice that in designing across 1200-watt 2041 and up to 360-watt with leading global contain companies. We expect to continue gaining momentum in the performance complete market through opens. And lastly, in industrial explication we're starting to see GaN and high-voltage secation in high-performance applications, spanning industrial pumps and heavy equipment electrification like DDC converts and megawatt orders. Turning to our second pillar, technology leadership. We continue to prioritize ovation across GaN and high-voltage fit technology. including both product and solutions, supported by expanding customer engagement and core development project.

One example of this innovation and system expertise with our breakthrough 10-kilowatt DC/AC platform that I just discussed previously. Another highlight was our announcement during the last quarter of our and 3,300 volt ultra-high-voltage SiC module portfolio, which we have accelerated simply to more customers. These modules feature providence assisted planar technology for [indiscernible] balance robustness and performance in mission clinical application across grid-tied infrastructure, energy storage and the [indiscernible] scale fast charging.

This structure available in six-pack G+ power modules, discrete packages and nongood [indiscernible] by point with extended AACs lalability testing. As mentioned earlier, we announced last week our Gen 5 technology and upcoming new 1.2 kV security pack product again PSU CDC for AI as centers. Our new Gen 5 technology continues to improve the figure of merit of our leading genetic technology. It leverages our trench assisted plan or TAP architecture best-in-class thermal behaviors and top 5 cooling unpack.

We're now somnour first new 1.2 Ken product to multiple OEM and ODM designing high-power and AC/DC or AI [indiscernible]. On our third pillar, operational efficiency. We have taken actionable steps to create a more streamlined and rebalance erotic deploy organization. We have been leasing strong employee buying and resin tangible benefits from this [indiscernible]. Also, on November 20, we were pleased to announce long-term strategies technogy and manufacturing partnership with Global Foundries to accelerate an technology design and manufacturing in United States. This partnership enables secure scalable solution for our target high power market and ensures that Navitas can deliver the performance efficiency and scale our customer demand. It also provides Navitas the opportunity to manufacture our solution in critical and national security applications in U.S. Development began a few weeks ago, and both companies are duly collaborating with production expected to begin later in the year and accelerate in 2027.

Over time, we expect to transition to [indiscernible] in order to lower products and increased scale. Also, during the quarter, we executed actions to restructure and optimize our go-to-market rate. This included significant consolidation of distribution channel partners from approximately 40 to less than 10 [indiscernible] we have the ability to scale and are well suited for serving high-power market while removing previously mobile-centric distributors. And our fourth pillar, financial discipline centers and resource realignment in support of our focus on high power markets. This includes very targeted 90% reduction in head count in the fourth quarter. offset by alignment actions to support the Navitas 2.0 shift, including hiring new employees well equipped for high-power markets, in particular within the United States. As evidenced by our fourth quarter revenue mix we have made tremendous progress. We also brought in new additional leaders with skills in sales and marketing, R&D and operations with a focus on enabling stronger execution. These collective actions focused the entire company on high power market and provide a foundation for efficient and effective execution going forward. Even with a larger market opportunity, our resource alignment allows us to efficiently focus our quarterly spend on the high-power market. As a result, we're targeting to maintain operating expenses flat throughout the coming year. We also expect to drive gradual margin expansion throughout '26 through improving scale and mix of high power business.

Lastly, to further strengthen our balance sheet and fund future operations, we completed a private placement of common stock in November with net proceeds of approximately $96 million. contributing to a quarterly end cash balance of $237 million. These proceeds further support our Navitas 2.0 strategy axing our transformation and funding working capital for scalable growth and long-term value creation. In closing, I am very pleased with the overall progress we achieved in a relatively short period of time. Speed is a financial element of our company's culture and it's clearly working. We are positioning Navitas 2.0 as a high power company, sharpening our focus on execution to enable scalable growth. Looking ahead, we anticipate a return to top line sequential growth starting in the first quarter, fueled by increased revenue from high-power markets. When combined with the benefit of our optimized cost structure, streamline go-to-market approach, an accelerated product road map, we also positioned to achieve gradual improvement in gross margin and bottom line results over the coming year. I'm incredibly proud of the team's dedication, hard work and agility in platination vision. I also want to thank our customers for their support to our new strategic dilation as well as ongoing contribution to mature beneficial collaboration partnership.

With that, I'll turn the call over to Todd to review our fourth quarter and full year results as well as our first quarter guidance.

Thank you, Chris. In my comments today, I will take you through our fourth quarter and full year 2025 financial results. And then I'll walk you through some of the important Q4 achievements and market dynamics as well as our outlook for the first quarter 2026. I will then return it to Chris for final remarks before we take questions. Revenue in the fourth quarter of 2025 exceeded the high end of guidance at $7.3 million compared to $10.1 million in the third quarter of 2025. As expected, revenue for the quarter reflects our strategic decision to deprioritize our low-power, lower-profit China mobile and consumer business, as well as our efforts to streamline our distribution network to align our focus on high power markets. .

As Chris mentioned, our high-power markets represented a majority of our quarterly revenue for the first time in the company's history, with mobile declining to less than 25%. This is a very important milestone and representative of our strategic shift. As mentioned before, we believe that Q4 represented the bottom for revenue as our strategic actions support driving increased contribution from our high-power business going forward. Before addressing gross profit and expenses, I'd like to refer you to the GAAP to non-GAAP reconciliations in our press release. In the rest of my commentary, I will refer to non-GAAP measures. I would also like to point out that our GAAP results for the fourth quarter included a $16.6 million restructuring and impairment charge that consisted of approximately $10 million of distribution contract terminations, $4 million of fixed asset impairments and $2 million of workforce reduction expenses associated with realigning the entire organization and distribution channel to focus on addressing high power markets.

Of the $16.6 million restructuring and impairment charge in the quarter, $3.8 million was noncash related items. Gross margin in the fourth quarter was $38.7 million, which was flat sequentially with the prior quarter, reflecting the ability to maintain our margin profile despite the lower quarterly revenue. At these revenue levels, we do not yet have the leverage to overcome our fixed costs. but we expect this to improve as we further grow revenue from high-power markets. As mentioned in our last earnings call, we expect to deliver expanded margins as we pursue a mix change towards higher power markets and away from mobile and low-end consumer. During the fourth quarter, we executed on a 19% workforce reduction, mostly deployed to mobile and consumer, and an organizational realignment towards U.S., high-power customers and markets, thereby reducing operating expenses sequentially from $15.4 million to $14.9 million. This is part of our strategic plan to realign company's resources to the Navitas 2.0 focus. Operating expenses were comprised of SG&A expenses of $6.8 million and R&D expenses of $8.1 mine.

These expense levels align with our cost reduction targets. The fourth quarter of 2025 loss from operations was $12.1 million compared to $11.5 million in the third quarter of 2025 as the reduction in operating expenses did not fully offset the decrease in revenue. Our weighted average share count for the fourth quarter was approximately 222 million shares. For the full year 2025, revenue was $45.9 million compared to $83.3 million in 2024. Gross margin for the full year was 38.4% and compared to 40.4% last year. 2025 operating expenses were $63.6 million compared to $83.4 million in 2024. The full year loss from operation was $46 million versus $49.7 million last year. As Chris mentioned, the fourth quarter represented the bottom in quarterly revenue, and we expect to return to top line sequential growth throughout 2026 as we continue our transition to high power markets.

Turning to the balance sheet. Accounts receivable was down to $3.6 million from $9.8 million in the third quarter, reducing our DSOs to 45 days. Inventory decreased to $13.3 million from $14.7 million last quarter. Cash and cash equivalents at quarter end were approximately $237 million, reflecting net proceeds of approximately $96 million from our completed private placement of common stock in November 2025. The company continues to carry no debt. Our balance sheet remains very strong as we exit the year with a high level of liquidity and improved working capital position. Moving to guidance for the first quarter of 2026. We expect revenue to increase sequentially to between $8 million and $8.5 million.

This represents the first quarter-over-quarter growth since the company's pivot. As I just mentioned, we expect sequential growth to continue throughout the year, driven by increasing revenue contribution from high power markets. Gross margin for the first quarter is expected to be 38.7%, plus or minus 25 basis points. We continue to anticipate the technological innovations to bring to high-power high-growth markets will result in progressive expansion of future gross margins. Turning to operating expenses. We anticipate operating expenses to remain approximately $15 million for the first quarter. We expect to continue to allocate resources and expenses as we redeploy company resources towards higher power customer and markets, particularly within the U.S. This redeployment of resources is expected to offset the strategic downsizing of our facilities to result in flat operating expenses. For the first quarter, we expect our weighted average share count to be approximately 230 million shares.

In closing, we are pleased with our initial progress and accelerate Pivot to non-top 2.0. As evidenced by high-power products representing the majority of our quarterly revenue for the first time, we expect to increasingly benefit from the broadening adoption of our GaN and high-voltage tick products. in targeted high-power markets. Together with our recent actions to reallocate resources, optimize operational efficiencies and and restructure distribution channels, we believe that Navitas is on a path to deliver improving margins and bottom line results. I'd now like to turn the call back to Chris for some final comments before opening the call to questions.

As we close today's call, I want to address 1 additional matter. After an extraordinary 10 years of dedicated service, Todd has decided to step down as CFO to push other opportunities. He has been an invaluable partner to every 1 company. bringing financial discipline, strategic insight and weathering integrity that help steer us through pay of both growth and challenges. Todd has been a great partner over the last 6 months, helping to pivot and furnish the company to agitate 2.0. On behalf of the entire Board and executive team, I want to extend our grategic for all of this solution over the past decade. We have true financial organization in place and [indiscernible] fully committed to assisting in a seamless transition until a successor has been made. We expect to communicate in the coming weeks regarding [indiscernible] replacement and Navitas new CFO. We enter in this chapter with confidence in our strategy, our momentum and our ability to continue delivering long-term value for our shareholders. Thank you again for joining us today. Operator, we might now open the call to questions.

[Operator Instructions] And our first question comes from the line of Kevin Garrigan with Jefferies.

Congrats on the results. Can you guys just walk us through how each of the high-power end markets performed in Q4? And how we should think about the trajectory for each of those markets in Q1?

Yes. Well, our quarter-on-quarter growth in revenue was due to the high power markets. So they are performing well. We're not going to sort of break out the high power market at this time. But we do expect all of them to be performing on a go-forward basis as mobile becomes immaterial as we move through the year. .

Okay. Got it. And then as a follow-up, can you just update us on the progress of the 800-volt architecture opportunity? And can you give us a sense of a time line on customer decisions?

Kevin, this is Chris. As we talked last time, there's a lot of work going on between us and the hyperscalers not only one, but multiple of them on the adoption of the 800-volt AC/DC. We sampled, as we mentioned in the press release and in the script, some of the new products that will be used in this type of architecture. We also announced a leading-edge 800-volt to 50-volt AC/DC brick that demonstrates the performance we can get with those products. So there's a contingent of collaboration. It's a bit too early to kind of tell you when this will be confirmed, but I think we are getting closer and closer with our customers. .

Congrats again on the results.

And the one thing I would add, given is everybody refers to the 800-volt HVDC as a step function for power content in the AI data center, and this is true, especially with the adoption of GaN replacing silicon as you move to the [indiscernible] HVDC, right? But I would outline that in IDC, and you've heard it from multiple vendors, is that there is an acceleration of demand also in using the classic architecture, which is ACDC, right, using SiC and we see a growth throughout the year ahead of the step function with GaN in HVDC. .

Our next question comes from the line of Kevin Cassidy with Rosenblatt.

Congratulations on the progress. Just as you mentioned, you're working with the hyperscalers. Is this -- are you working directly with them? Would they be building their own power supplies or is that going to be a pull from the current power suppliers to the hyperscalers.

Kevin, it's actually all of you above, okay? So the hyperscalers, partially are driving the new architecture, right, both in terms of what they expect in terms of density and power level in the AC/DCs as well as the 800-volt and the 50-volt or even lower voltage HVDC architecture. Now we don't work only with the hyperscalers. If you think about PSU, which is clearly designed in OEMs and ODMs that are serving those hyperscalers. And if you look at the HBDC is that our classic merchant power company serving the scales also doing designs on these new architectures. So we work with everybody. I would tell you that the driver of the change of the acitecture comes from the U.S. and hyperscalers. But a lot of the OEM and ODM in Taiwan, in China, but also in the U.S., are driving that. And as you know, we just announced this Board, right, which I mentioned, which was basically a co-development with the customer. And that's basically to showcase the level of efficiency you can get by using GaN on the primary side and gain on the secondary side in 800-volt to 50-volt AC/DC brick. And you're going to see more of those reference implementation in the future as well.

Okay. Great. And many of your customers are the hyperscalers giving you an idea of when that inflection point would be when they start doing the installations.

I mean the thing I would say, as I said just earlier on Kevin's question, there are 2 stream, if you prefer, of the AI [indiscernible] growth. Number one is more data centers, more power, and that drives more PSUs, higher power PSUs and that drives the growth in SIC, which we are seeing throughout '26. When it comes to the 800-volt HVDC, which I think is your question, when there is a discontinuity and you cannot use silicon anymore on the reside because you're 800-volt and you have to move to high-voltage GaN. This is really driven by not the GPU change, but the rack architecture change. As you compact more GPUs into a seal rack, and you get to a megawatt rack, you cannot get the power density and the efficiency with silicon.

And that's this discontinue. I would say, as we said before, this is really about '27. Can be used slightly earlier it could. There is a case where you can use GaN in the 48-volt IBC replacing silicon. As I mentioned in the script, where GaN brings higher efficiency, you can do it with silicon, but you get higher efficiency. And this might be the first time you see GaN in data center [indiscernible] the real step function is really coming from the 800-volt DC, which is really kind of linked to the [indiscernible], okay, which is the higher integration of GPUs in 7.

Next question comes from the Nathaniel Quinn Bolton with Needham.

Congratulations on the progress on the transformation to Navitas to and gestations to you, Todd. I guess, Chris, I wanted to come back on the 800 HBDC solution, especially if you think about the primary side of that 800-volt rail, there still a lot of folks in the industry that I think are talking about using silicon carbide in that 100-volt conversion step you guys are obviously pushing the GaN solution. But I guess can you say what are you seeing from the leading GDU and hyperscaler vendors that are looking at the 800-volt for the 400 plus/minus rack architecture. Are they pushing more for GaN? Are they open to begin and silicon carbide solutions. Just how do you see this playing out from a technology perspective between GaN and silicon carbide?

It's a very good question, actually, because I think there is some level of confusion. First of all, I would tell you that we are not pushing anything both SiC and GaN and we welcome stick being used on the primary side if it's needed and GaN being used on the primary if that's needed. So we're not pushing anything. We are being pulled. We've not seen any significant use case our Board implementation of customer evaluation using SIC on the primary side. SiC is being used widely at 1 to 2-kilovolt I mentioned in the classic AC/DC, right, which is basically prior to the and. But when it comes to Eden DC, we've been pulled by customers, and I'm talking about hyperscalers to Kevin's question here that are driving the adoption because it's more efficient and more driving higher density.

And then, Chris, you also talked about your 10-kilowatt all GaN BRIC solution. Can you give us a sense, is that more of a reference platform? Or would that be a solution that Novitas would look to source that entire brick level product? Because I imagine it includes a fair amount of additional componentry. And so just thinking about to the extent you're selling the full brick solution, I imagine that might be pretty high [indiscernible]. So could you just talk about whether you sell the GaN solutions [indiscernible] Portet brick would you sell the entire brick? And if you did sell the entire brick, what would the margin inflation stay?

It's a very good question. Thank you, Quinn. We view this as enabling solution enabling technology for the customer. So first of all, as I mentioned, this is something we've done with [indiscernible]. We have not done that in a vacuum on our own, right? This is something that we could develop with a leading customer. Number one. Number two is we don't compete with customers, okay? At this stage, we don't see a path where we're going to sell the modules. Now that design is shared with our customers and the hyperscalers as well as the ODM and OEM that are looking at how we've been able to achieve that high level of efficiency, right, 98.5%, which we believe, based on what we've seen and some of our competitors' feedback on top of customers are 1 of the best in the industry, right? So I would tell you, this is for us. This is what we've done in GaN historically. We pioneered GaN in mobile by demonstrating and helping customers to get to a higher level of efficiency, lower EMI, highest level of density. And we are doing the same in AI data center.

And this is what I talk about when we talk about -- we're leveraging the benefit and the skills of 1.0, right? And this system expertise makes a difference. At the end of the day, we are in the business of selling GaN and silicon carbide and enabling our customers. As a matter of fact, on that board, we're using some of our competitors in our silicon and other technologies and products that we don't have. But the focus is how to show and help the customers to accelerate the adoption of GaN in HVDC.

Next comes from the line of John Tanwanteng with CJS Securities.

if you could start, maybe talk a little bit about the competitive landscape in supplying the 800-volt data center. What are you seeing just in terms of who you're bidding against in these sockets if they're outpricing you or we're doing better in technology. And on top of that, how is your partnership with [indiscernible] and evolving in that space as well?

So thank you for your question, John. So I'll start with the end. We continue our partnership with Infineon. We have a cross license, as you know, and we share the same vision, which is to enable the accelerated adoption of GaN and silicon carbide in the IDC, right? So sick as the traditional accenture and GaN in the 80s, right? So there's a lot of dialogue between the 2 companies on that front, right? Number two, you will have seen that there are multiple vendors having been listed on the 800-volt AI factory kind of ecosystem.

As a matter of fact, I think it's up to 13 vendors relay. But we don't see all of them in each of the socket we target. So I would recommend that you look at how many of those 14 are actually in the high-voltage game. So how many of them have a 650-volt gain in the right package to be able to enable donor HVDC. How many of them have been voltage GaN to enable the 50-volt secondary side. Some of them are listed as a silicon vendor, okay? We are listed as a gun vendor. The other thing is, as we talked about seeing used on the AC/DC as well, there is a natural pool and more sick as we get to a higher voltage and also outside of data center.

To do the 800-volt HVDC, you need to enable a change of the grid architecture. This is a pure, high voltage, ultrahigh-voltage sticky. So I will tell you is there's a lot of competition, but not everybody is feeling on the same thing. And there are not many of the vendors being listed that have both high voltage SiC or Ultra voltages and the competing in the AC/DC with 12 kV or in the grid with 2 KV and above and adding high-voltage and [indiscernible]. So this competition pool is actually being reduced. That's why we are very clear about what we do, we play in the silicon, we play in the again, high-voltage mid-voltage and in the high voltage and high voltage [indiscernible]

Got it. And then second, could you us on the incremental margin of either this 800-volt data center products or high-power products in general, especially as you roll out new suppliers?

So first of all, as Todd said, right, we expect continued gross margin expansion. So remember that the growth this year is coming from all high power markets and basically mobile going down, right, being less than the 5%. What I would tell you is the scale is going to help more gross margin as we grow revenues for our fixed costs, so that and that drives margin expansion. Number two is the HIBOR product in the high-power market are coming at higher margin than mobile work, okay? And that mix is going to change. And the third one is we are very active in ramping new suppliers, partially on the package side that will help us to reduce cost, right? So there is multiple aspects of our we are confident to see gross margin expansion as we clearly outlined for the rest of '26, right?

And as we scale further in '27, we expect to continue.

[Operator Instructions] Next question comes from the line of Jack Egan with Charter Equity Research.

I had kind of a follow-up on the gross margin question before me. So as mobile is getting smaller and smaller, I'm kind of curious of the longer-term outlook. Are your gross engines more so going to be driven by mix as in data center or nondata center end market mix? Or is it more kind of the technological innovation, I guess, that Todd mentioned that -- it sounds like it's referring to new products with higher ASPs. I guess I'm trying to look at what the driver of the margins -- that margin expansion is whether it's end market mix or better product margins?

So it's actually going to be a combination of both. So definitely end product mix, right, as mobile decreases, the high power markets are going to give us higher margin. Those are more reliability and performance. But then on -- as we sort of scale and have these new products come into the high-power markets, we do expect like further expansions through like optimized process yields and packaging costs, which will help drive our product cost down, thereby driving our margins up as well.

The one thing I would add, Jack, is scale cost reduction and basically higher margin product, right, as Todd said. But the one thing I would say is that, again, the growth this year, and I think we've been very clear that we are very confident that Q4 at the bottom, we are guiding up for Q1, and we said we're going to grow quarter-over-quarter throughout the year with margin expansion. I would reemphasize again that the growth is coming from whole high-power market. .

Yes, AI data center is a big part of the future outlook. And if you look at the SAM that we shared just a few weeks ago. It's nearly half of the SAM that we see for us in 2030. But I would outline that performance computing is growing this year, okay, and will continue to grow and also help on the margin mix. Grid infrastructure is really accelerating. And I think you're going to see higher ASP product, higher-margin product coming in play there, where it's more about reliability and performance and less about cost, of course, cost matter. And then as we talked about in [indiscernible] AI Center, which is a cost-sensitive market, it's all about efficiency right now, okay? And I think you're going to see all those markets contributing to the growth expansion in gross margins, right? So I just wanted to kind of calibrate a little bit your question to make sure that we don't see the gross margin expansion only coming from AI center.

Sure, Chris. No, that's super helpful then. And then I guess kind of from a higher level, I know you're not supplying as much into some of the automotive and industrial type markets. But silicon carbide has gone through just broadly speaking, has gone through a period of pretty significant oversupply. And so I was just kind of curious, what are your expectations on when that supply and demand in some of the other markets might balance out whether for Navitas or whether the industry as a whole. I know that you're dealing with some large volume wins that, so it might not apply to you as much, but just any commentary there would be helpful.

I mean, to be honest with you, John, I think, first of all, as an industry -- I would say it's going to take some time, but I think you should that to the vendors that are supplying to EV, we don't. We don't play in the same league of SIC, okay? You've seen me being very clear, but the fact that we compete and focus on 2 kV for the PSUs for AI data center and 2 KV and above up to 5 kV and even more for the grid. This is where -- this is not about scale of supply, this is about how reliable and efficient and high performance is your technology. So I think for some of the SI vendors operating at 450-volt 650-volt 80-volt focusing on EV, that's a valid concern. For us, it's about scaling with the Ultra voltage, okay, which is nothing really related to supply at this stage. .

Next question comes from the line of Richard Shannon with Craig-Hallum.

Apologies, the ambient noise just jumped off a plane here and I missed a bit of the call here. So I hope I don't repeat your question here. But Chris, one thing I'd love to ask you is in the data center opportunity here. To what degree are your opportunities coming from your partnership and kind of drafting behind, if you will, from Infineon versus other ways? And then also, are there any -- is there any cross-fertilization of wins within the rack between point of load and the 800-volt down, did those kind of cross-fertilize and give you additional benefit at all? .

So a very good question. As I mentioned, we partner with Infineon. We have a cross license. We've done that a couple of years ago. We continue to drive collaboration to enable GaN adoption, both high voltage and mid-voltage I would say that we don't leverage or benefit from [indiscernible] what you will see is, I think I got a question earlier from Kevin, who are you bumping into most of the time, I would say, it's fully Infineon and surely Infineon because they have the same vision. They have the same technology, high-voltage GaN, [indiscernible] voltage GaN and SiC and Ultra voltage as well for the green infrastructure. So very similar [indiscernible]. So we bump into each other. We follow each other, but I would not say that we leverage Infineon, right? Now interestingly enough, when we release the package, we found that we -- because we talk to the same customers, we think the same way is we end up seeing the technology the same way. That's what I would say. On your question about expansion of portfolio, that's something we are looking at. My focus right now was to pivot the company, okay? And put every eggs we have, every engineer we have, every focus we have on the 4 high power markets and high voltage GaN in hybrid SiC, but we are looking at opportunity to expand the portfolio as you get higher voltage in the data center, you're going to need circuit protection, and that's something we're going to look at in the future, right? But for now, we are laser-focused on execution with the product we have and we just released.

Okay. Fair enough. And the second question, probably for Todd. Just on the gross margins. If I caught the end of his prepared remarks talked about not having enough scale to really drive leverage on gross margins quite yet here. Is there a revenue level by which that happens here? And kind of what's that fall-through margin when you start to see that trajectory?

Yes. That's a great question. As our mix changes, obviously, our margins will grow. But right now, we have that scale issue. We do see margins starting to expand again in Q2 and beyond. So that's sort of the tripping point right now. What I would tell you, Jack, is the high power markets and the high-power product in the high -- are coming at higher margins. The mobile is going down. We said that in Q4, it was 25%, and we said it's going to continue to go down and get insignificant by the tone, right? So I think we are very confident that the mix of the mix of the mix change, the higher power product in the high-power market increase as a percentage of the company and the new product and the cost reduction we have will yield to gross margin expansion. So you'll see it light and clear, okay, starting not very far from now.

Our last question comes from the line Quinn Bolton with Needham.

Chris, you spent a lot of time talking about the 800-volt data center opportunity, but you also talked about needing to rearchitect the grid. And I just wondering if you could spend a second talking about the opportunity for the high-voltage silicon carbide and the solid-state transformers from where are you in the design process for some of those solid state transformers. And is there a way you can ballpark like what's the dollar content opportunity? I don't know if it's on a per megabasis or a per unit basis. But is there a way to size what the amount high-voltage sick that goes into a solid-state transformers that start to be deployed as the grade is rearchitected.

Thank you, Quinn, for this follow-up on. Actually, I'm glad you asked that question because everybody focus on AI data center. [indiscernible] focusing on the 800-volt HVDC architecture, which is important because it's a discontinuity in the architecture. It's a replacement of silicon by GaN or by high-voltage technology. But none of this is possible if the grid is not changing. And this is not just about getting a more efficient grid. It's a change of the architecture. And you refer to SSEs, which is basically getting from 35,000 volt A, super high-power, high-voltage lines down to 800-volt DC at the highest level of renability. That drives the change.

And I tell you, I've never seen the grid structure changing that fast. So you asked me, and I think we in the script and in the press release, we are accelerating the sampling of our 2.3 kV and 3.2 applications are any greed side application. SST, of course, but battery energy system, megawatt chargers, solar farm at the grid tide level. Any of those applications are in accelerated designs. We are very busy talking to all those customers. That's why I said it's 2 legs, okay? We have 4 high-power market. But if I look at the future, it was 2 legs, the IDC and the grid are equally important. And this is a pure high-voltage sick play. And to the earlier question about EV, this is not the technology because what you have to deliver is high-reliability technology, high reliable modules, so it's a different play. So I'm glad you asked the question. We see an accelerated design momentum. Of course, it's going to take time. This is a longer design cycle than computing. It's longer design cycle than AI DC, but I think you'll start to see a significant revenue growth starting with 7. To your question about content. In the last investor meeting we had, we basically referred to $25,000, $35,000 per megawatt of total content for Navitas as a SAM, which is based again on ultra-high-voltage, high-voltage 6, 12 kV and 2 KV and above and GaN. And about 10 to 12 is actually outside of data center. So if you think about 25 to 45, 10 to 12 is outside of the center, which is purely SST, BSS and all those applications, right? So -- and again, inside data center, you have a share between sick for PSUs and GaN as you move to it on a DC, but we should not underestimate the importance and the potential of the grid infrastructure. As a matter of fact, we released a couple of weeks ago, details some analysis that shows that both gain and sick of 50%, 50%, okay, in 2030, have of the potential for -- in terms of the [indiscernible] for Navitas to the [ $3.5 billion ] that we referred earlier. And you can see that the grid is not far off the SAM of data center. And the other thing I would say is this is just the beginning. So if you think about grid, this is a multiple decade transformation that will drive higher voltage continuously.

We start with 2 KV, we get into 3 KV. We're going to get to 5 KV and above, and that's going to drive transformation for the next multiple decades.

That concludes the question-and-answer session. I would now like to turn the call back over to the management team for closing remarks.

Thank you, everybody, for attending this call. As you could tell, we are very proud of the progress we are making. The first 5 months and 6 months since I joined was about pivoting the company and being clear about where we are going. I think we've done that. We are focusing on accelerating samples of our technology. We have 4 pillars of the transformation, which I mentioned, market focus, technology leadership operational efficiency and financial discipline. And we'll continue to update you on how we make progress. And I think we have a bright future ahead of us.

Ladies and gentlemen, that concludes today's call. Thank you all for joining us. You may now disconnect.

Thank you for standing by. My name is Jordan, and I'll be your conference operator today. At this time, I'd like to welcome everyone to the Navitas Semiconductor Third Quarter 2025 Earnings Call. [Operator Instructions] I'd now like to turn the call over to Lori Barker, Investor Relations. You may begin.

Good afternoon, everyone. I'm Lori Barker, Investor Relations for Navitas. Thank you for joining Navitas Semiconductor's Third Quarter 2025 Results Conference Call. I'm joined today by Chris Alexandra, AI President and CEO; and Todd Glickman, CFO. A replay of this webcast will be available on our website approximately 1 hour following this conference call and available for approximately 30 days. Additional information related to our business is also posted on the Investor Relations section of our website.

Our earnings release includes non-GAAP financial measures. Reconciliation of these non-GAAP financial measures with the most directly comparable GAAP measures are included in our third quarter earnings release and also posted on our website in the Investor Relations section. Non-GAAP expenses and operating margin excludes stock-based compensation, amortization of intangible assets and other nonrecurring items. In this conference call, we will make forward-looking statements about future events, our future strategy or the future financial performance of Navitas.

We may make predictions or describe trends in our industry and markets. You can identify some of these statements by words like we expect or we believe or similar items. We wish to caution you that all such forward-looking statements are subject to assumptions, risks and uncertainties that could cause actual events or results to differ materially from expectations expressed in our forward-looking statements. Important factors that can affect Navitas' business include factors that could cause actual results to differ from our forward-looking statements are described in our earnings release. Please also refer to the Risk Factors section in our most recent 10-K and 10-Q. Our actual performance may differ from our projections and our estimates, assumptions and strategies may change.

Navitas assumes no obligation to update forward-looking statements to reflect actual results, changed circumstances or other events that may occur except as required by law. And now over to Chris Allexandre, CEO.

Good afternoon, and thank you for joining us. I am very excited to have the opportunity to address you today. I have now been in my role for approximately 60 days, and I would like to take this opportunity to share with you my vision for the future of Navitas. I will start by saying that I feel incredibly proud to be leading the Navitas team, a world-class team that has been at the forefront of both gallium nitride or GaN and high-voltage silicon carbide or SiC since the very beginning of their development. Today marks a crucial moment for Navitas as we enter a transformation of our company. Over the last 60 days, I've been on the road meeting and collaborating with customers, employees, suppliers and partners.

That strategic tour gave me a clear view of our strengths and challenges and most importantly, the opportunity ahead. The conclusion is straightforward. Navitas is a company with enormous potential, underpinned by strong foundational elements already in place in both GaN and high-voltage SiC. And we have a tremendous opportunity to win in high-power, high-growth markets such as AI data centers, performance computing, energy and grid infrastructure and industrial electrification. Customers are eager to adopt those technologies into their application, and we have the experience and track record of delivering those technologies in scale and volume, and they want to collaborate.

Simply put, we are in the right markets with the right technologies, and we can win with focus on strong execution. We will accelerate, pivot and double down on those high-power markets and customers as we move away from consumer and mobile. I call this Navitas 2.0, a transformation to a high-power sharp focused company serving grid to the GPU to drive more consistent, profitable and sustainable results. Before we dig into what we are doing, let's quickly cover what is happening in the market right now. Across the entire market, electrification is accelerating and moving up in power demand. AI data centers necessitate accessing power distribution to achieve higher efficiency and density, and they are doing so exponentially.

In parallel, the energy grid is transforming with storage, solid-state transformers, utility scale renewables and megawatt charging to support the AI catalyst, but also the overall growing energy demand. This is not a short cycle. It is a durable, multi-decade sustainable trend that will reshape power architecture at rack, system and grid type levels. This requires fundamental change in customer system architecture, design and technologies and simply means the total market size Navitas is addressing has increased multiple folds. It opens immense opportunity for high-power players such as Navitas 2.0.

I spoke to a variety of customers over the last 60 days. Every single customer I met in those segments from leading U.S. hyperscalers and AI GPU vendors performance computing OEM and ODM to the many and very innovative customers [ in having a complete architecture ] in the energy grid told me the same message. GaN and high-voltage SiC technologies are the solution to the problem they are trying to solve and the revolution they are driving, and they view Navitas at the center of this transformation given our long history and track record in shipping those at scale. We have heard that message loud and clear, and therefore, will immediately accelerate accordingly.

We are one of the few companies with a complete high-power portfolio, GaN, GaN integrated with IC and high-voltage SiC. This combination, along with very strong system expertise built over the last many years, offers more value to customers. 10 years ago, Navitas was looking ahead at global energy demand, which was projected to grow by 200% to 300%, doubling to tripling over the next decade. EV, wind turbine, cloud computing, data centers, solar power, climate change, quantum computing, all those energy platform requiring significant reformats to consume efficient high-voltage technologies, which did not exist at that time. This enormous forecasted demand made it clear that power electronics would need a transformative leap in efficiency and power level.

Standing back, it was obvious even 10 years ago that existing silicon-based chip and technology would simply not get us there. And we didn't even consider AI and its power hungry implication back then. Navitas led the introduction of GaN into power electronics, leapfrogging established silicon players. Then we acquired and merged with GeneSiC, the leading advanced technology in high-voltage SiC. We've shipped over 300 million GaN units with proven quality and reliability over the last seven years, and GeneSiC brought leading-edge high-voltage SiC technology, both together enabling power architecture evolution in AI data centers, performance computing, energy and grid infrastructure and industrial electrification. GaN is now mainstream for AI data centers, performance computing and industrial electrification.

The NVIDIA 800-volt DC AI factory ecosystem announcement is the first proof point, and we expect it to be adopted across many other players. High-voltage SiC is also supporting and enabling the energy grid transformation necessary to enable AI and the associated demand for more power. Both markets are intertwined. Underneath our portfolio are long-standing partnership with leading fabs, back-end and module partners, a deep co-design with customers and a very advanced solution and system architecture understanding with more than 300 patents issued or pending. Our team has been on the forefront of GaN and high-voltage SiC from the early days, now reaching over 2 decades combined, and that experience matter when customers move fast and execution is critical.

Going back to Navitas 2.0 and the transformation from a mobile and consumer-focused foundation to a high-power company, that pivot is backed by decisive actions that we have begun to take. Number one, resource realignment. We are reallocating engineering, commercial and application support and R&D program towards high-power platforms and customers. We're ensuring we have the right people in the right markets and the right geographies, led by a renewed global high-performance leadership team. Number two, road map acceleration. We are accelerating the release of new products tailored to high-power markets targeted at rack, system and grid type nodes. We expand medium-voltage GaN devices, high-voltage GaN devices and IC and high-voltage fixed module. Number three, go-to-market restructuring.

We are focusing on hyperscalers, GPU vendors, Tier 1 OEM and ODM and leaders in our focus markets, AI data centers, performance computing, energy and grid infrastructure and industrial electrification. We intend to streamline our distribution network to align with those high-power focus markets. This also means a change in geographical resource deployment, including creating a stronger presence in the U.S., where we have growing and promising engagement. Number four, portfolio and customer pruning. We are deprioritizing lower margin, short life cycle projects, transactional markets and customers such as mobile and selected China-based segment to redeploy capacity and attention to durable high-power program. Our focus is on the long-term engagement where technological innovation makes a difference. We believe this will ultimately drive high-quality business with greater predictability, consistency and higher margin. Overall, this change will impact our business model. High-power engagement are indeed deeper, longer-lasting and multigenerational. We may often engage across multiple subsystems within the same customer, some served by GaN, others served by high-voltage SiC.

That breadth is expecting to increase win rates, raise the blended margin and produce more predictable, repeatable revenue compared with transactional, lower-margin segments such as mobile. This is the foundation for Navitas 2.0, a scalable, profitable and sustainable enterprise. At the OCP Global Summit, NVIDIA named Navitas a power selector partner for its next-generation 800-volt DC AI factory power architecture. That validates our ability to serve the entire power path from the grid to the GPU. In support of this ecosystem, we announced our first 100-volt GaNFast alongside our portfolio of 650 GaN discrete Fast and our GaN Safe IC and expanded high-voltage SiC products. This is our first formal entry into medium voltage GaN, the critical range for AI server power stages and rack level distribution. We're sampling now 2.3 kilovolt and 3.3 kilovolt high-voltage SC module to leaders in battery energy storage system, solid-state transformer program and megawatt charging. The strategy and opportunity are clear.

To get there, our plan is grounded in 4 pillars: number one, market focus, AI data centers, performance computing, energy and grid infrastructure and industrial electrification. We will stay sharply focused on those high-power markets only. Number two, technology and manufacturing leadership, continuous innovation in GaN, GaN IC and high-voltage SiC informed by customer requirements and co-design. We have a strong foundation in technological innovation, and we'll continue to lead the industry. We will also expand our manufacturing footprint to better serve our high-power customers. You may also see us doing more partnership to enable faster adoption. Number three, operational efficiency, a streamlined and rebalanced geographically deployed organization, a scalable foundry and packaging and module partnerships.

Number four, financial discipline, prioritized investment, leverageable OpEx and a shift towards high-margin program. In the near term, this transformation will have an impact, including a reduction in guidance before returning to growth. We expect Q4 to mark the bottom as we take decisive actions, including reducing channel inventory, consolidating distribution channel and adjusting our inventory to better align with our new high-power markets and customers. By deprioritizing lower-margin revenue and redirecting our road map and investment away from non-high power businesses, we believe we will accelerate our transformation and gradually improve the overall quality and profitability of our business throughout 2026.

This is expected to yield more consistent growth and margin expansion. We'll continue to provide transparent updates on our progress throughout this transition to ensure accountability at every step. AI data centers and performance computing are already shaping product [ requirement and designing ] and design wins. On the AI data center front, we expect material P&L contribution starting 2027. Our work with NVIDIA and other hyperscalers, GPU vendors, OEM and ODM, however, established already in 2026, a durable design win foundation for long-term growth. On performance computing, we continue to make progress in engagement as GaN technology is gaining rapid adoption in higher power, and we expect this to drive growth already in 2026.

In parallel, energy and green infrastructure are multibillion-dollar markets with multi-decade opportunity where high-voltage SiC is exceptionally well suited to our customer road map. As we embark in this transformation and execute this transition, we will share information for you to track our progress through transparent and clear update in the following area. First, a sharper focus on high-power account and program, which will be seen in growing importance and weight in our revenue, driving a change in mix. Second, operating expense, financial discipline and return on investment-driven road map decision solely focused on Navitas 2.0 North Star. Third, gradual gross margin improvements as we reduce lower-value shipments, grow more higher power engagement and overall improve the mix.

In conclusion, our GaN and GaN IC built a strong presence in mobile fast charging, and we are very proud of the 300 million units shipped. This gives us an in-depth understanding and [indiscernible], and this is the business that brought us to where we are today. We have complementary high-voltage SiC technology, products and modules, enabling us to cover more of this high power chain. High power markets are different and more rewarding. Engagements are deeper, the past cycles are longer and the value we deliver is measured in system-level performance and efficiency over multiple generations. This is where Navitas 2.0 will focus on.

We're executing a clear pivot to high-growth, high-power markets focusing on AI and data centers, performance computing, energy and grid infrastructure and industrial electrification, anchored by a complete GaN plus high-voltage SC portfolio with long-standing customer relationship and disciplined operation. We're aligning our organization, resource allocation, road map and channels to the markets that matter. We firmly believe that the change we are making will improve the quality of our business and position us for sustained growth and margin expansion throughout '26 and beyond. From the grid to the GPU, Navitas 2.0 is a high-power company built for scale and profitability.

Thank you to our employees, customers, suppliers and partners for their support during this transition. I look forward to deepening our partnership with stockholders, analysts and investors with transparent updates as we execute. With that, I'll turn the call over to Todd to review our third quarter results and our guidance. After Todd's remarks, I'll return for Q&A, including detailing our high-power Navitas 2.0 strategy, the actions behind the transformation and what it means for our business. Thank you.

Thank you, Chris. In my comments today, I will take you through our third quarter 2025 financial results, and then I'll discuss the financial implications of Navitas 2.0 with our accelerated transition to a high-power company with focus on AI data centers, performance computing, energy and grid infrastructure and industrial electrification markets. Also, I will outline how we plan to reallocate our resources with our new, more focused approach designed to grow revenue and seek profitability. Revenue in the third quarter of 2025 was at the midpoint of guidance at $10.1 million. While the industry environment remained relatively static compared to the second quarter of 2025, the expected revenue reduction reflects both adverse impacts from the China tariff risk for our silicon carbide business and pricing pressure in our mobile business, particularly in China.

Before addressing gross profit and expenses, I'd like to refer you to the GAAP to non-GAAP reconciliation in our press release. In the rest of my commentary, I will refer to non-GAAP measures. Gross margin in the third quarter was 38.7%, which was up sequentially compared to 38.5% in the second quarter, primarily due to a slight favorable change in end market mix. In the third quarter, we executed on further operational efficiencies, and we reduced operating expenses sequentially from $16.1 million to $15.4 million. Operating expenses were comprised of SG&A expenses of $7.1 million and R&D expenses of $8.3 million. These expenses align with our cost reduction target. Adding all this together, the third quarter 2025 loss from operations increased sequentially to $11.5 million from $10.6 million in the second quarter of 2025 as cost reductions did not fully offset the sequential decline in revenue.

Our weighted average share count for the third quarter was 213 million shares. Turning to the balance sheet. Accounts receivable was down to $9.8 million from $12.5 million in the second quarter. Inventory was relatively flat since last quarter at $14.7 million. Our balance sheet remains very strong as we exit Q3 2025 with high levels of liquidity and an improved working capital position. Cash and cash equivalents at quarter end were $151 million, and we continue to carry no debt. Moving on to guidance for the fourth quarter. We currently expect revenues at $7 million, plus or minus $250,000. This expected revenue reduction reflects our strategic decision to deprioritize our low-power, lower-profit China mobile business as well as our efforts to level set channel inventory and streamline distribution network to align ourselves with our high-power directive.

We believe that Q4 will represent the bottom for revenue as these actions will allow us to move faster to concentrate on the high-power business and customers that will, in turn, enable consistent gradual revenue growth throughout 2026. Gross margin for the fourth quarter is expected to be relatively flat compared to the third quarter with our guidance at 38.5%, plus or minus 50 basis points. However, we anticipate the technological innovation we bring to high-power, high-growth markets will result in a progressive increase in gross margins going forward.

Turning to operating expenses. We anticipate continuing to trim expenses to $15 million in the fourth quarter, reflecting a 24% year-over-year reduction. We expect to continue to reallocate resources and expenses as we redeploy the company towards higher power customers and markets, notably U.S. customers. The redeployment and an appropriate downsizing of our facilities will result in a lower quarterly operating expense level, and we believe we will be well positioned with our personnel and resources to execute on our pivot to higher power, resulting in quarter-over-quarter quality sales and margin growth and route to profitability.

For the fourth quarter of 2025, we expect our weighted average share count to be approximately 214 million shares. In closing, it is an exciting time at Navitas 2.0 as we leverage our leadership in GaN and high-voltage SiC to pivot to a high-power company and capture the exponential growth expected to come from AI data center, performance computing, energy and grid infrastructure and industrial electrification. We are moving fast to transition from consumer and mobile markets to more sustainable, higher power segments where Navitas is well positioned as the leader in GaN device shipped and high-voltage SiC to deliver a high-quality, scalable business. We are confident these strategic moves position the company for its next wave of more profitable growth. Operator, let's begin the Q&A session.

[Operator Instructions] Your first question comes from Kevin Cassidy from Rosenblatt Securities.

Welcome, Chris. Looking forward to working with you. Maybe just to understand a little better the shape of this transition. How long of a tail is the mobile market? Do you have some higher voltage applications in the mobile market that could continue on and then cross over to the power supplies, high-voltage power supplies. When do you expect that would be more than 50% of the business?

Yes. No, that's a great question. Maybe I could start there and Chris can add color. So when we're looking at the business today, right, I think as we look at Q3, mobile represent the vast majority of our business. And as we move into Q4, it's going to actually represent less than 50%. And all the growth going forward in our company as we go quarter-on-quarter, as we discussed, the gradual growth is going to come from these new markets of AI data center, performance computing and grid infrastructure.

[ Ralph ], nice to meet you and a very good question. As we move towards '26, what we see is as mobile continue to decline, we see an acceleration of the high-power market that Todd just listed. And we believe this is going to enable us to drive the quarter-over-quarter gradual growth that we talked about. When it comes down to mobile, I think you have to differentiate the high end of the mobile to the low end of the mobile. The low-end side has quickly commoditized over the last year or 2. The high end, however, we've reached a plateau. And if you think about 100-watt chargers today, there is less and less differentiation, and we anticipate an acceleration of the commoditization. That's why we decided along with the speed and the opportunity in the hyper market that Todd mentioned, we decided to accelerate that pivot that we have planned for a while.

Okay. Great. And just as a follow-up, just to understand the AI data center, I see it as 2 stages. One, you have the power supply companies that tend to be more conservative, we'll say, in making transitions and compared to the AI data centers that are starving for more power and would probably want to switch over more aggressively. So what is the strategy there? Do you work with the end users and pull the power supply customers along with them? Or do you work behind the power supply customers?

It's a very good question, [ Ralph ]. It's actually -- we do both. But as we pivot, the orientation of our engagement is definitely more towards the OEM and the hyperscalers. Let's start by AI because everybody talks about AI. AI is really a catalyst that drive change across all the markets we talked about, data centers, of course, but we'll talk about performance computing that is also disrupted by AI as well as the grid energy infrastructure. So as we ship today to AI, we ship to the, as you said, the power companies, mostly sitting in Taiwan. However, as the hyperscalers are taking control and driving the disruption of the architecture, we see engagement is pivoting towards the U.S. And the announcement made by NVIDIA with the 800-volt DC AI factory is really just an amplified example of how the hyperscalers are now trying to drive from the grid to the GPU and driving the architecture change at all stages. So we talk to those hyperscalers on all level of the stages.

We, of course, work with their partners to implement those solutions, but the system-level discussion that we have and that we had had over the last few weeks and months are really kind of enabling this transition that NVIDIA talked about, both for GaN as well as high-voltage SiC.

Sorry, I called you -- I didn't use the proper name Kevin.

Your next question comes from the line of Ross Seymore from Deutsche Bank.

Chris, welcome aboard, and you can call me, Kevin, if you want. So I guess a bigger picture question just to start. So when you guys were added to the collaboration list for the 800-volt data center stuff at NVIDIA, there were 10 names. The line at the time was you were the only one with GaN and silicon carbide. Now there's 14 names and the incremental 4, a bunch of them do have the GaN side, maybe not as much silicon carbide. So I guess the question is, when you look at 14 potential competitors or whatever subset you align to, what do you think the true competitive differentiation is for Navitas? And is it more on the silicon carbide or the GaN side?

Thank you. Very good question. I would say -- I would start by saying that the fact that we have both high-voltage SiC and GaN has not changed. And this is truly, as you said, a differentiator and not many of our competitors are having that. The other thing, and I've spent over the last 8 weeks, a lot of time with those customers, right? They all told me the same. When it is about enabling that pivot that technology disruption, that adoption of GaN track record matters. And we have through the pioneering of GaN into the first market that got GaN at scale, which is mobile charger, developed expertise, deep technology understanding and experience that really matter. So I would say beyond the technology understanding, the fact that we have both speed and track record is going to be a key differentiator.

And as I talked to NVIDIA and a few other of the hyperscalers, they all told me the same. It's about speed and support to have them enabling that transition with safety and execution. And that's where we're going to differentiate against our competitors.

And I guess whether it be you or Todd, it's nice that the fourth quarter is the bottom. When you think about growth going forward, I thought I heard, and forgive me if I missed out a little bit on this, but I thought I heard a significant portion of the data center side would be more in 2027. So just not putting absolute numbers around it, but the tailwinds sequentially in 1Q, 2Q, et cetera, of '26, what gives you the confidence this is the bottom? And just kind of conceptually, what are the areas that are going to grow off of the $7 million in the fourth quarter?

Yes. I think what gives us confidence that this is the bottom is we actually proactively are walking away from revenue in mobile. We want to make sure that there's not a distraction in the business and we can concentrate on the long-term goal here, which is data center performance computing and grid infrastructure. So that's how it gives us the confidence. And going forward, the growth is going to come out of those markets and not mobile. And with today, mobile represent a large majority going forward, that's going to continuously go down as we move through 2026. So that gives us confidence not only the better revenue, but it will be more sustainable and more profitable.

Let me add just one comment. First of all, as Todd said, by walking away from short term, less long term, less innovation-based engagement with customers and applications, we are pivoting our resource a lot faster to drive this new growth. Number two, there's a clear acceleration in those markets. Should it be the growth in data centers itself, even in the form of the traditional, should I call it this way, traditional power implementation before we move to the 800-volt DC, and we benefit from that through the power vendors. If you look at the impact of AI in performance computing with those notebooks and other high-end computers getting more powerful, they also require a lot more power. And we see an acceleration of that demand to higher power in '26 is going to drive a lot of growth.

Last but not least, we also see an acceleration of demand in any form of energy and green infrastructure. So we have a lot of pull from midsized to large-sized customers really trying to accelerate how they can enable the AI revolution. So long story short, the AI is a catalyst across multiple markets. And yes, data centers itself through the 800-volt DC will drive a lot of accelerated growth, but we see that catalyst already [ yielding ] the fruit in '26.

Your next question comes from the line of Quinn Bolton from Needham & Company.

This is Shadi Mitwalli on for Quinn. Obviously, the move away from China Mobile is having a bigger impact than expected. But I just want to get some more color on the puts and takes here. And just overall, has anything new changed over the last 90 days?

You know, nothing's new changed besides the fact that Chris has been on the road talking to customers, and they've asked us to go faster. And so instead of focusing and sort of allowing mobile to still represent a large majority of our business going forward, we did a more proactive approach to walk away from that. And so that's really the key difference in the last 90 days that's happened just because the growth in these new sectors is extraordinary, and we want to make sure we're in the best position to take advantage of that.

[ Sean ], let me add some color because I joined 8 weeks ago. So if any change, I'm probably the one behind the change, right? I would say the market itself has not changed, okay? There is a clear acceleration, and it is faster than it was just a few months ago. But what has changed is the clarity that we have that we can't continue to transition smoothly. I come from a background where if you double down on the greatest opportunities, you have more chance to capture it. Again, as I said, I have met a lot of customers, all of them in data centers, computing, infrastructure, all told me the same. They want to move faster, they want to enable the transition. So it was very clear for me after a couple of weeks on the road that we cannot transition this keeping some of our resource maintaining the past. So we have to double down and move our entire resource or most of our resource into enabling the future so that we can capitalize on the opportunity we have and maximize our chance of winning.

So I would say the one thing that has changed is the clarity and the pivot. I think historically, we've been trying to manage the past and the future. And with the amount of resource we have, we just have to accelerate that transition, in my opinion.

Great. That was helpful. And then my follow-up is on the solar end market. I believe in Q1, you guys talked about ramping a solar microinverter win in the second half of this year. And I was just curious if this is still on track? And if so, how is the ramp going?

Yes, nothing has changed. I mean you're referring to the GaN BDS, which will ramp in '26 with our lead customer, which is part of the energy and green infrastructure segment that we talked about.

Your next question comes from the line of Jack Egan from Charter Equity Research.

Chris, congrats on being appointed CEO. So Navitas, I mean, I guess kind of a high-level question. Navitas certainly has some exciting opportunities ahead of it. You came from Renesas, you had experience with quite a few large companies in the analog and power space. And so obviously, one big difference is going to be the culture of moving from a company with 10,000 employees to a few hundred. So yes, just from a high level, what are some of maybe the cultural or operational characteristics, I guess, from your prior experiences that you might plan to install at Navitas?

Thank you for the question. Of course, we are transitioning the culture as we speak. I come from a culture of strong execution and pivot with clarity, and that's what we are bringing here. What I would say as well is have learned very well some of the pitfalls of big companies that do not always move the fastest. So the culture that we are building in Navitas and what you are seeing in the choice we've made walking away from our historical business is, number one, clarity and being solely sharply focused on what's going to make us more profitable and more sustainable from a result point of view in the future. Number two, speed -- speed so we can transition faster and execute the plan for our customers in a faster way. And number three, execution, which I think was probably missing in the culture that we had here so far.

Great. Okay. That's helpful. And then for the mobile business in China, you are sticking with the longer term, some of the higher voltage, higher value opportunities. Is there any reason why the competitive and pricing pressures won't eventually kind of move up to that side of the business, like kind of like what's happening with the lower value chargers now?

I think it's a very good question. What I would say is what's happening in mobile today is the fact that innovation has stopped bringing any value. We've been in the last few years about how do we basically get more power into a smaller form factor. Today, we've reached a plateau. 100-watt charger is fairly small, and it's all about getting it cheaper. When you look at data centers, even if you look at performance computing, we are not at that point. At this point, this is about how do we basically accelerate the efficiency? How do we accelerate the transition, the adoption of high-voltage SiC, the adoption of GaN. And we are at complete different power level. So if I give you an example about high-performance compute with AI going to the client, we are talking about how do we get those super high-end computers into multiple hundred watts. And this is a complete different ball game compared to what we've seen in mobile, where content is much higher, expertise and ability to execute is much harder.

And I think this is going to benefit companies like Navitas who have mastered GaN for the last few years into the first market that took it to scale. So I think the answer to your question is yes. Ultimately, when power level will plateau, and innovation will not make a difference. You're going to get there. But what I see is we are very far away from this in any of the market that we're talking about, in particularly anything that touch AI. We're talking about 10x, 100x every generation after generation [ in power ] delivery.

Your next question comes from the line of John Tanwanteng from CJS Securities.

My first one is, could you just talk about the data center prospects in '26 before you ramp the 800-volt products? Just what kind of growth are you expecting in the next year with the legacy servers and ecosystems that are out there currently?

Yes. That's a great question. So Jon, I mean, today, we do -- we are shipping into AI data centers today. It's not material. As we look into '26, we'll continue to grow that revenue. But at the end of the day, the materiality when this is going to become a larger portion of our business is going to wait until 2027 when 800-volt becomes prime time. So that's when we really expect the exponential growth happening from data centers.

Sorry, do you mind if I just add a few color? The way I view it is, again, there is a disruption in the adoption of GaN and high-voltage SiC coming with this new concept of basically breaking the traditional way the data centers were built, okay, moving from the super high voltage going through AC/DC and then converting it down to lower power. With the NVIDIA announcement, we are talking about volt down to GPU. So that will drive a lot more content and the rapid adoption of GaN. In the interim, as I said, AI is a catalyst. There's more server and AI deployment out there that drive ultimately growth for anybody serving power, right? So that's what Todd said. If you look at we are shipping today in data centers, that revenue will grow. It's not material, and it will not represent a material impact. But in '27, as we pivot to the new architecture, this is where the content in GaN, mid-voltage GaN and high-voltage GaN is going to drive to a complete different level.

Great. And then I was wondering if you could talk about your cash position and your burn rate, and if that leaves enough for you to ramp growth to meet that next-gen data center demand as you enter '27?

Yes, absolutely. So we finished the quarter at $151 million, and it's a very healthy balance sheet with no debt. Right now, we're burning around $10 million to $11 million a quarter, and that should be plenty to give us ongoing -- for ongoing operations. So that's where we stand today.

Your next question comes from the line of Tristan Gerra from Baird.

This is Tyler Bomba on for Tristan. Navitas was initially planning on starting to ramp silicon carbide [ epitaxy ] internally mid-2024. How much of your silicon carbide output is currently in-sourced? What is your target for that either in terms of dollars or percent of revenue?

Yes, that's a great question. So yes, we did initially when the market was tight, and we wanted to generate some epitaxy in-house. But however, given the market has loosened, we no longer do that. So we never initiated that project. And we are -- all of our substrates and epi as it relates to high-voltage silicon carbide is outsourced today.

The next question comes from the line of Richard Shannon from Craig-Hallum.

Chris, welcome to Navitas. I look forward to working with you as well. Let me ask my first question on data center here and talk about the engagement and how things are going here and ultimately when you expect to get to wins here. I think the kind of one of the part of the engagement model, which we didn't hear about any in your prepared remarks is working with Infineon as a second source into this area. Wondering if that's still part of the strategy here. And then ultimately, when do you expect to be able to talk about getting wins and having visibility on when they ramp?

Thank you for the question. So first of all, I'll answer on the Infineon. We continue to communicate with Infineon. You know that we have a cross-licensing going on. We continue to be aligned on the vision that we want to enable the adoption of GaN and high-voltage SiC into those hyper markets. When it comes to the engagement with NVIDIA, and I would say broadly with the hyperscalers because we don't only spend time with NVIDIA, but we also spend time with a lot of the other hyperscalers, OEM as well as ODM, right? For me, '26 is about enabling the transition. First of all, we are not focusing only on the mid-voltage and closer to GPU. We are focusing on the whole chain.

We talked about energy and grid infrastructure. If you think about the first stage of the 800-volt DC, it requires to basically move from transformers to solid-state transformers where high-voltage SiC is a prime technology. Then you have to go to multiple stage of power conversion. So -- if I -- your question is when can you basically give us color on the engagement. We'll continue to give you in a transparent way how we expand our partnership. We can't be specific on any form of engagement. But what I can tell you is the number of interaction at system level, product level, technology level with some of the hyperscalers that you mentioned is all-time high.

Okay. Great. Second question, I really just kind of looking at calendar '26 here, and it's kind of a multipart question here. Please bear with me here. wonder if you could give us some sense of the relative contribution of your lower voltage GaN here and also silicon carbide? And then ultimately, I think the question we're going to get asked a lot tomorrow morning is, can you -- how do we think about calendar '26 sales? It seems like it's almost no chance it's going to grow year-on-year based on the dynamics you've discussed here. But if you could make any comments about how that looks, that would be great.

So I'll start by answering your question about GaN versus SiC, right, and the growth towards '26, and I'll let Todd add some comment with a bit more detail. First of all, we don't look at the business at GaN versus SiC. As I mentioned early on, we are focusing on enabling this low power to high power. And we look at revenue really from a segment point of view, okay? That's number one. Now if you want a bit more color, and we talked about the fact that starting Q4, which is the bottom, we're going to grow quarter-over-quarter gradually as the high-power markets are offsetting the decline in mobile. Both GaN and SiC will grow in high power markets. That's the fact for '26.

Now if you look at segment level, mobile will be down. Mobile consumer will be down. But high-power computing and energy grid infrastructure will be up. And data centers will grow even though not material throughout '26 and really kind of accelerate the growth in '27.

Great. And I think one of the other questions was on when was low voltage going to go into revenue? Is that's correct, Richard?

Yes. and some characterization of what kind of contributions you'll see for next year.

Yes. So low voltage, we went to market with low voltage, as you saw, PSMC. That's designed right now for the 800-volt data center. So when that ramps, you're going to see material revenue from low voltage taking place.

Yes. We just announced a couple of weeks ago, our first -- it's actually a mid-voltage GaN 100 volt, which was the first outcome partnership with TSMC that we had for a couple of years. That's really tailored towards the last stage of power conversion into the AI 800-volt DC, and we expect that to start to ramp in 2027.

Your final question comes from the line of John Tanwanteng from CJS Securities.

I was just wondering if you could talk about the incremental margins in the high-voltage data centers and the great opportunities, if they're any different from the existing high-voltage businesses that you have for high-power businesses?

Yes. I think both those margins, we expect those markets to have higher margins and more sustainable than our current mobile and consumer business, which is what is contributing to our guide to basically go quarter-on-quarter growth as margins as revenue picks up in 2026.

And Jon, what I would say is the high-power markets are very different in nature. First of all, as we talked about, they are at the beginning of the adoption of those high-voltage technologies, so be GaN and high-voltage SiC. Number two, it's about innovation. It's about speed, it's about execution. Yes, all those customers want us to be cost competitive. Yes, cost is part of the criteria that they make decision. What I can tell you, having met many of them over the last 60 days is that they don't lead with that, very different from mobile, where we've plateaued from an innovation standpoint. And now it's about how can you make my charger cheaper. When we talk to the OEM, ODM, hyperscalers, they are asking us, how do you do what you do faster. So I think margin and value that we bring and value that we extract will help us to uplift the margin as the high-power portion of our revenue increase over time.

Got it. And actually, the second question is a follow-up to that. How are you planning to accelerate that development? Is it -- what are the remaining things that you need to do to produce products for these markets that work and do you have -- you can scale?

Well, this is why we've decided to basically move faster away from mobile, right? It's application engineering, it's support to customers. It's system engineering. It's R&D. It's basically moving most of the R&D, if not the entire R&D towards making road map for those markets instead of maintaining generation over generation presence in the low-end mobile market. So it is basically pivoting -- and I think I've heard the word pivot and transformation. It is pivoting the entire resource level to capture on that growing opportunity moving forward.

Okay. Great. Last one for me. Can you just talk about the capacity to ramp for these customers? I know you have that new agreement with Powerchip. Just help me understand what the capacity there is to meet the demand if you do get a significant amount of share there.

So what I would say is, first of all, TSMC has been a great partner, and they are helping us to transition for the next few years. We expect to continue to ship from TSMC for the next multiple years. Number two, we are ramping TSMC. We are in transition. We announced the first product from mid-voltage GaN. We're going to transition the high-voltage GaN in '27. And last but not least, we continue to look for new foundry partners to be able to capitalize on the opportunity that we have, either from a geographical supply chain point of view or a cost point of view as well as volume point of view.

That concludes the Q&A session for today's meeting. You are now able to disconnect. Everybody, have a great day.