Home / Transcripts / Navitas Semiconductor Corporation (NVTS) · February 24, 2026

Navitas Semiconductor Corporation (NVTS) Earnings Call Transcript

February 24, 2026

NASDAQ US Information Technology Semiconductors and Semiconductor Equipment earnings 56 min

Earnings Call Speaker Segments

Brett Perry executive
#1

Thanks, operator. Hello everyone, and good afternoon. We're glad you could join us for Navitas Semiconductor's conference call covering the financial results of our fourth quarter of 2025. On the line with us today are Chris Allexandre, who serves as Navitas' President and CEO, along with our CFO, Todd Glickman. I want to point out to everyone listening that what we're reporting today has not been finalized. These figures could still change because the company has yet to complete its closing procedures, and our independent registered public accounting firm still needs to perform its customary quarterly review. That means the numbers we're sharing are unaudited and could be revised before the company submits its Form 10-K covering the year that ended December 31, 2025. On top of that, the remarks management has prepared include forward-looking statements, and those come with risks and uncertainties. Management might also offer further forward-looking statements when answering your questions. With that in mind, the company is claiming the safe harbor protection for forward-looking statements provided under the Private Securities Litigation Reform Act of 1995. What actually happens could end up being different from what we talk about today, so we encourage you to read a fuller discussion of the risks and uncertainties found in the company's filings with the Securities and Exchange Commission, which include Form 10-K and Form 10-Q. Also, any projections regarding how the company will perform in the future reflect what management estimates as of today, February 24, 2026. Navitas has no obligation to revise these projections down the road as market conditions may or may not shift, except where applicable law requires it. Beyond that, in the company's press release and in what management says during this conference call, we will discuss certain measures and financial information presented on both a GAAP and a non-GAAP basis. The company's press release contains definitions and reconciliations between GAAP and non-GAAP items, and those offer additional detail. If you aren't able to listen to the whole call right now, a recording can be accessed through webcast for 90 days in the Investor Relations section of Navitas' website at www.navitassemi.com. At this point, I'm happy to hand the call over to Navitas' President and CEO. Chris, the floor is yours.

Chris Allexandre executive
#2

Good afternoon, everyone, and thank you for joining us. This marks my second quarterly conference call since becoming Navitas CEO, and I'm glad to be hosting it. We closed out the year with a strong fourth quarter, and we keep accelerating the move toward Navitas 2.0 as we bring the entire organization into alignment around high-power markets. Honestly, the five months since I joined have been invigorating, and my conviction in our industry-leading GaN and our high-voltage SiC offering has only deepened, along with my assurance that our strategic pivot is on course to scale the company successfully to its next level. Before I get into quarter-specific remarks and updates, I want to quickly go back over a handful of important things regarding the strategic transformation we've already shared, as well as our vision for what we refer to as Navitas 2.0. To start, we are accelerating the shift away from what has historically been the company's mobile and low-end consumer business, and moving toward high-power markets instead — places where our GaN and high-voltage SiC products can offer real differentiation and value through greater density, efficiency and reliability. Our focus is squarely on 4 high-growth, high-value market segments: AI data center, energy and grid infrastructure, performance computing and industrial electrification. Put all this together, and what we're looking at is a serviceable addressable market worth $3.5 billion come 2030. That splits about evenly — roughly 50 to 50 — between GaN and high-voltage SiC, and the two combined are growing at a CAGR above 60%. Now, the largest portion of that $3.5 billion SAM does fall inside AI data centers along with grid and energy infrastructure. But I want to stress something: AI acts as a common catalyst underneath all 4 of our target markets, and it's pushing a fast acceleration when it comes to re-architecting infrastructure, what customers expect, and how quickly this new high-voltage technology gets adopted. We have a decade-long demonstrated history as an early mover in GaN at scale — more than 300 million GaN units delivered — and that's paired with strong system and application know-how plus our high-voltage SiC leadership via GeneSiC. What Navitas 2.0 aims to accomplish is clear: move fast into markets, lock in deeper relationships with customers, and reach scale, which should deliver growth that's more durable, steadier, and profitable going forward for Navitas. Now a quick look back at our Q4 numbers. As an early step in shifting toward Navitas 2.0, we finished reorganizing the whole company — covering both capabilities and locations — so we're centered on serving high-power markets. That covers shifting our people, our road map, and where we're putting our attention, all the way through. For Q4, sales landed at the top of the range we'd guided to, hitting $7.3 million, and it was also the first quarter in which the high-power market accounted for more than half of what we brought in overall. Our view that Q4 marked the low point hasn't changed. One thing worth calling out: Mobile fell quarter over quarter, going from more than half of revenue in Q3 to under 25% of the total in Q4. We see Mobile shrinking further as a share of quarterly revenue, and by the close of '26 it should no longer matter much at all. And then, in line with what we said on the previous quarter's call, our guidance for Q1 is quarter-over-quarter growth, and we expect that sequential growth to keep going through '26, powered by better sales traction in the high-power market. During the past several months, through the expanded meetings I've been holding with customers and partners, I've come across many proof points showing that adoption of new technology is speeding up. AI acts as a catalyst, transforming the game across markets. What exists today in technology and architecture simply isn't enough anymore. The pace of technology adoption in the industry is faster than at any point before, with customers clearly stepping in to capitalize on GaN and high-voltage SiC technology. As I said before, AI serves as the leading catalyst creating momentum and widening adoption of high-power solutions across all 4 of the high-power markets we target. Each customer interaction has verified that the market is going through secular change and that AI is the sparkling revolution we're concentrating on. This compelling inflection point in architecture, design and technology adoption strongly favors GaN and high-voltage SiC, placing Navitas 2.0 at the heart of this revolution. As we laid out on the prior call, the Navitas 2.0 transformation into a high-power company rests on decisive actions and is built on 4 pillars: market focus, technology leadership, operational efficiency and financial discipline. Let me now take you through the key strides we've made in each of those areas since we last held an earnings call. Starting with where we're aiming in the market. As I mentioned before, we're keenly concentrated on the high-power segments: AI data centers, energy and grid infrastructure, performance computing, and industrial electrification. Looking specifically at AI data centers, Navitas occupies a distinctive spot as a front-runner in GaN and high-voltage SiC, backing all the leading AI data center architectures. The density of compute power is going to require better efficiency along with higher power density, and that is precisely what is driving the faster adoption of GaN in the data centers of the next generation. During this quarter, we accelerated the pace at which we sample products and deliver solutions, focusing on our 100-volt GaN and 650-volt GaN, which target AI data centers, 800-volt HVDC, and 48-volt IBC HV buck architecture. Samples are available right now across multiple package sites, and over a dozen customers are in the process of evaluating them. More recently, on February 9, we introduced our breakthrough 10-kilowatt DC-DC design platform. This is an all-GaN 10-kilowatt 800-volt to 50-volt DC-DC platform, and it employs advanced 650-volt and 100-volt GaNFast FETs within a 3-level half-bridge architecture featuring synchronous electrification. This platform has achieved a 98.5% peak efficiency, which we believe represents the industry's best result to date. This full-brick package design platform reached leading power density and supports plus or minus 400-volt VDC standard for AI data centers. Here's a great illustration of how Navitas can leverage our 10 years of GaN and system expertise. We're establishing the benchmark for scalable, high-performance AI infrastructure. Our product lineup made it possible to achieve power density never seen before, which helped enable quick, large-scale growth of AI data centers, while giving hyperscalers and OEMs the means to get the most out of compute density and cut energy loss as they roll out next-generation AI workload. On the SiC side, we're working closely with customers on AC-DC PSU designs for the AI data center architecture of today, leveraging our latest 1.2 kV SiC devices and building on the fifth-generation GeneSiC technology that we introduced earlier this month. Better figures of merit and best-in-class thermal performance come with this product, plus the topside cooling Q-DPAK package, and the response from customers has been positive. Now, looking at the grid and energy infrastructure market: a major transformation and modernization is underway in the energy grid, fueled by the AI catalyst and by wider growth in energy demand. That's not a short cycle — it's a secular, sustainable, multi-decade trend that will reshape grid and energy infrastructure. As a result, we are seeing faster design cycles as well. We're leading this effort with our recently introduced 2.3 kV and 3.3 kV SiC modules, which operate at ultra-high voltage, along with a plan to reach even higher voltages; at present, we are in the evaluation stage with over 15 OEMs globally, primarily located in the U.S. and Europe, and the U.S. in particular is showing clear acceleration. When it comes to performance computing, GaN continues to gain more ground in high-power chargers and power units designed for high-end computing and AI notebooks, taking the place of silicon.

Chris Allexandre executive
#3

We've got over 15 projects that are in production, and roughly double that count in the design stage, spanning 170-watt, 200-watt, 250-watt, 240-watt, and all the way up to 360 watts, working alongside leading global computing companies. Through '26, we anticipate continued momentum building in the performance computing market. Then, moving to industrial electrification, GaN and high-voltage SiC activity is beginning to show up in high-performance applications that range from industrial pumps to heavy equipment electrification, such as DC-DC converters and megawatt chargers. Now to our second pillar, technology leadership. Innovation remains our priority across GaN and high-voltage SiC technology, covering both products and solutions, backed by growing customer engagement and co-development projects. Our breakthrough 10-kilowatt DC-DC platform, which I mentioned earlier, stands as one example of this innovation and system expertise. A further highlight came when, during the last quarter, we announced our 2,300-volt and 3,300-volt ultra-high-voltage SiC module portfolio, and we've sped up sampling to additional customers. These modules carry proprietary Trench-Assisted Planar technology for AI scalability, delivering advanced robustness and performance in mission-critical applications spanning grid-tied infrastructure, energy storage, and megawatt scale fast charging; they come in SiCPAK G-plus power modules, discrete packages, and known good die format, with extended AEC-plus reliability testing. As I noted earlier, last week we announced our Gen 5 technology along with an upcoming new 1.2 kV SiC Q-DPAK product aimed at PSU AC-DC for AI data centers. Our new Gen 5 SiC technology keeps enhancing the figure of merits of our leading GeneSiC technology. It draws on our Trench-Assisted Planar TAAP architecture, best-in-class thermal behaviors, and topside cooling in Q-DPAK. We're sampling our first new 1.2 kV Gen 5 SiC product right now to multiple OEM and ODMs designing high-power PSUs and AC-DC for AI data center; on our third pillar, operational efficiency, we've taken actionable steps toward a more streamlined and rebalanced geographically deployed organization. Employee buy-in has been strong, and we're seeing tangible benefits from these efforts. Additionally, on November 20, we were pleased to announce a long-term strategic technology and manufacturing partnership with GlobalFoundries to accelerate GaN technology design and manufacturing in the 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 gives Navitas the chance to manufacture our solution in critical and national security applications in U.S. Development kicked off a few weeks ago, and both companies are collaborating deeply, with production expected to begin later in the year and accelerate in 2027. Over time, we expect to transition to 8-inch in order to lower product costs and increase scale. Also, during the quarter, we executed actions to restructure and optimize our go-to-market strategy. That involved significant consolidation of distribution channel partners, going from approximately 40 down to less than 10 distributors. 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 on resource realignment in support of our focus on high-power market. Included in this is a highly focused cut of 19% to our workforce during Q4, which was balanced out by realignment steps supporting the Navitas 2.0 transition, such as bringing aboard fresh staff members well suited to high-power markets, especially in the United States. Our Q4 revenue mix makes clear that we've achieved enormous progress. Beyond that, we added new leaders possessing expertise in sales and marketing, R&D and operations, all aimed at driving stronger execution. Through these combined moves, the whole company is now concentrated on the high-power market, and they lay a groundwork for efficient and effective execution moving forward. Despite a bigger market opportunity, our resource realignment lets us direct our quarterly spend efficiently toward the high-power market. Consequently, we aim to keep operating expenses level across the next year. We further anticipate achieving gradual margin expansion during '26 via better scale and mix in the high-power business. Finally, to bolster our balance sheet even more and finance future operations, we closed a private placement of common stock during November, yielding net proceeds of roughly $96 million, which added to a quarter-end cash balance of $237 million. These funds bolster our Navitas 2.0 strategy even further, speeding up our transformation and providing working capital for scalable growth and long-term value creation. To wrap up, I'm extremely happy with the overall progress we made in a fairly brief span. Speed is a financial component of our corporate culture, and it's plainly delivering results; we are establishing Navitas 2.0 as a high-power company, honing our execution focus to facilitate scalable growth. Going forward, we expect a return to sequential top line growth beginning in Q1, driven by higher revenue from high-power markets. Paired with the advantages of our optimized cost structure, streamlined go-to-market approach and accelerated product roadmap, we're likewise set up to deliver gradual improvement in gross margin and bottom line results across the next year. I take immense pride in the team's commitment, effort and agility in shifting toward the Navitas 2.0 vision. I'd also like to express gratitude to our customers for backing our new strategic direction and for their continued role in mutually beneficial collaboration and partnership. With that, I'll hand the call to Todd so he can go over our fourth quarter and full year results, along with our first quarter guidance.

Todd Glickman executive
#4

Thanks, Chris. In what I'll share now, I'll cover the financial results for our fourth quarter and for the full year 2025. After that, I'll go over several key Q4 accomplishments and market dynamics, along with what we expect for the first quarter of 2026. Then I'll hand things back to Chris for closing comments before we open it up for questions. Fourth quarter 2025 revenue came in above the top end of our guidance, at $7.3 million, versus $10.1 million during the third quarter of 2025. As we anticipated, this quarterly revenue mirrors the deliberate choice to pull back from our low-power, less profitable China mobile and consumer business, together with our work to simplify the distribution network so that our attention stays on high-power markets. As Chris noted, high-power markets accounted for most of our quarterly revenue for the first time since the company began, while mobile fell to under 25%. That marks a highly significant milestone and shows what our strategic shift is about. As we've said previously, we view Q4 as the low point for revenue, since our strategic moves help drive growing contribution from the high-power business ahead. Before I get into gross profit and expenses, please look at the GAAP to non-GAAP reconciliations included in our press release. For the remainder of what I say, I'll be discussing non-GAAP measures. I'd also note that our GAAP results for the fourth quarter contained a $16.6 million restructuring and impairment charge, made up of roughly $10 million for distribution contract terminations, $4 million for fixed asset impairments, and $2 million for workforce reduction expenses tied to realigning the whole organization and distribution channel around high-power markets. Within that $16.6 million restructuring and impairment charge for the quarter, $3.8 million related to noncash items. Gross margin for the fourth quarter was $38.7 million, unchanged from the previous quarter on a sequential basis, which shows we could hold our margin profile even with lower quarterly revenue. At these revenue levels, we don't yet have the leverage to get past our fixed costs, though we expect that to get better as revenue from high-power markets continues to grow. As we said on the prior earnings call, we expect margins to expand as we shift our mix toward higher power markets and away from mobile and low-end consumer. In the fourth quarter, we carried out a 19% workforce reduction, largely in mobile and consumer, plus an organizational realignment toward U.S. high-power customers and markets, which brought operating expenses down sequentially from $15.4 million to $14.9 million. That's part of our strategic plan to refocus the company's resources on Navitas 2.0. Operating expenses consisted of $6.8 million in SG&A expenses and $8.1 million in R&D expenses; these levels are in line with our cost reduction targets. For the fourth quarter of 2025, our loss from operations came in at $12.1 million, versus $11.5 million during the third quarter of 2025, since lower operating expenses weren't enough to make up for the revenue decline. Weighted average shares outstanding in Q4 were roughly 222 million; across all of 2025, revenue totaled $45.9 million, down from $83.3 million in 2024. For the full year, gross margin stood at 38.4%, compared with 40.4% the prior year. Operating expenses for 2025 reached $63.6 million, versus $83.4 million in 2024. Our full-year operating loss was $46 million, against $49.7 million a year earlier. As Chris noted, quarterly revenue hit its low point in Q4, and we anticipate a return to sequential top-line growth across 2026 as our shift toward high-power markets continues. Now on to the balance sheet. Accounts receivable fell to $3.6 million from $9.8 million in Q3, which brought our DSOs down to 45 days. Inventory declined to $13.3 million from $14.7 million in the prior quarter. At quarter's end, cash and cash equivalents totaled roughly $237 million, which reflects net proceeds of about $96 million from the private placement of common stock we completed in November 2025. The company still has no debt on its books. Our balance sheet remains very strong, and we finish the year with substantial liquidity and a better working capital position. Now for first quarter 2026 guidance: we forecast revenue rising sequentially to a range of $8 million to $8.5 million. That would mark the first quarter-over-quarter increase since the company's pivot. As I said earlier, we expect sequential growth to persist through the year, fueled by growing revenue contribution from high-power markets. We project gross margin of 38.7%, plus or minus 25 basis points, for the first quarter. We still expect that the technological innovations we bring to high-power, high-growth markets will drive progressive expansion of gross margins going forward. Now on operating expenses. We expect operating expenses to stay at roughly $15 million in the first quarter. We anticipate continuing to allocate resources and expenses as we redeploy company resources toward higher power customer end markets, especially in the U.S; this redeployment is expected to offset the strategic downsizing of our facilities, resulting in flat operating expenses; for the first quarter, we expect weighted average shares outstanding of approximately 230 million. To wrap up, we're pleased with our early progress and accelerated pivot to Navitas 2.0, which is demonstrated by high-power products accounting for the majority of our quarterly revenue for the first time. We expect to benefit increasingly from broader adoption of our GaN and high-voltage SiC products in targeted high-power markets. Along with our recent steps to reallocate resources, optimize operational efficiencies and restructure distribution channels, we believe Navitas is on a path to deliver improving margins and bottom line results. I'd now like to hand the call back to Chris for some final comments before we open the call to questions.

Chris Allexandre executive
#5

Before we wrap up this session, there's one more item I'd like to cover. Following a remarkable decade of committed work, Todd has chosen to resign from his role as CFO in order to chase other pursuits. To each person at this company, he has been an incredibly valuable collaborator, contributing financial rigor, strategic vision and weathering integrity that were instrumental in guiding us through stretches of expansion as well as difficulty. Todd has likewise been a tremendous ally throughout the past 6 months, aiding in the pivot and shift of this business toward Navitas 2.0. Speaking for the whole Board alongside the executive team, I wish to offer our thanks for everything he has given across the last 10 years; a robust financial structure is already established here, and Todd remains entirely devoted to supporting a smooth handover until a replacement is appointed. In the weeks ahead, we anticipate sharing news about who will take over for Todd and become Navitas' next CFO. We step into this fresh chapter feeling assured about our approach, our trajectory and our capacity to keep generating lasting worth for those who hold our shares. Once more, thank you for being with us today. Operator, let's go ahead and take questions now.

Kevin Garrigan analyst
#6

Chris and Todd, congratulations on those results. Todd, I also want to wish you the best as you move on to whatever comes next. Could you both take us through how the high-power end markets each did in Q4? And what should our expectations be for how each of those markets trends in Q1?

Todd Glickman executive
#7

Sure. Clearly, the revenue increase we saw from one quarter to the next came from the high-power markets. So those markets are doing well. At this point, we aren't going to break those high-power markets out individually, but as mobile fades in importance over the course of the year, we do anticipate that every one of them will keep performing going forward.

Kevin Garrigan analyst
#8

All right, understood. Then, to build on that, could you bring us up to date regarding where things stand with the 800-volt architecture opportunity? Also, could you give some indication of when customers might make their decisions?

Chris Allexandre executive
#9

Kevin, it's Chris here. Like we discussed previously, quite a bit of effort is underway involving us and the hyperscalers — not just a single one, but several — regarding adoption of the 800-volt HVDC; we sampled, as noted in the press release and the script, a number of the new products set to be used in that kind of architecture. We additionally unveiled a leading-edge 800-volt to 50-volt DC-DC brick, which showcases the performance achievable using those products; so collaboration keeps moving forward. Telling you when this gets confirmed is a bit too early at this point, but I believe we're getting closer and closer with our customers.

Kevin Garrigan analyst
#10

Alright, excellent. Congratulations once more on those results.

Chris Allexandre executive
#11

One more point I'd make, Kevin — when people talk about the 800-volt HVDC, they describe it as a step function for power content inside the AI data center, and that's accurate, particularly as GaN gets adopted and replaces silicon during the shift to the 800-volt HVDC, correct? However, I'd point out that within AI DC — and you've heard this from a number of vendors — demand is also accelerating around the use of the traditional architecture, meaning AC-DC, right, with SiC; and we're seeing growth across the year prior to that step function involving GaN in HVDC.

Kevin Cassidy analyst
#12

Nice work on the momentum. You said you're engaged with the hyperscalers. Are those direct relationships? Will they make their own power supplies? Or does that become a demand pull from today's power supply vendors toward the hyperscalers?

Chris Allexandre executive
#13

Kevin, every one of those paths applies, all right? So the hyperscalers, in part 2, are shaping the new architecture, correct, in terms of what they want on density and power level in AC-DC PSUs, and also the 800-volt, either 50-volt or even lower voltage HVDC architecture, correct? Now, we don't engage solely with the hyperscalers. Consider PSU, which is plainly designed into OEMs and ODMs that serve those hyperscalers, and consider HVDC — many of these classic merchant power companies that serve the hyperscalers are likewise doing designs around this new architecture. So we engage with all of them. I'd say the force behind the architecture shift originates in the U.S. and the hyperscalers, yet plenty of the OEM and ODM in Taiwan, in China, and also in the U.S. are pushing that. And as you're aware, we recently announced this board, correct, which I referenced, that was essentially a co-development with the customer; and that's essentially to demonstrate the efficiency level achievable by using GaN on the primary side and GaN on the secondary side in an 800-volt to 50-volt DC-DC brick; and you'll see additional reference implementations like those going forward as well.

Kevin Cassidy analyst
#14

All right, great. And do any of your customers or the hyperscalers give you a sense of when that inflection point might arrive, when they'd begin the installations?

Chris Allexandre executive
#15

Well, what I'd say is this — as I mentioned a bit earlier in response to Kevin's question — there are 2 streams, if you will, behind the AI data center growth. First, more data centers, more power, and that pushes more PSUs, higher-power PSUs, and that in turn fuels the growth in SiC, which we're observing all through '26. Now with regard to 800-volt HVDC, which I believe is what you're asking about, when a discontinuity appears and silicon can no longer be used on the primary side, since you're at 800-volt and must shift to high-voltage GaN. What really drives this isn't the GPU change, it's the rack architecture change. As you pack more GPUs into one rack and reach a megawatt rack, silicon can't deliver the power density and the efficiency; that's the discontinuity. As we've said previously, I'd say this is really a '27 story. Might GaN be adopted a little sooner? It's possible. There's a scenario where GaN can be used in the 48-volt IBC to replace silicon, as I noted in the script, where GaN delivers higher efficiency. You can accomplish it with silicon, but you get higher efficiency. And that could be the first instance of GaN appearing in data centers. But the genuine step function really stems from the 800-volt DC, which is really tied to the Kyber rack, which is the greater integration of GPUs in '27.

Kevin Cassidy analyst
#16

Kevin, I'd also like to express my sincere good wishes to Todd, and it has truly been a pleasure collaborating with you.

Todd Glickman executive
#17

Todd, thank you, Kevin.

Quinn Bolton analyst
#18

Congrats on the strides made with the Navitas 2.0 transformation, and all the best to you, Todd. Chris, I suppose I'd like to revisit the 800 HVDC solution, particularly when you consider the primary side of that 800-volt rail. Plenty of people across the industry, I believe, are still discussing the use of silicon carbide in that 800-volt conversion step. Clearly, you're advocating for the GaN approach. That said, I guess, could you share — what are you hearing from the top GPU and hyperscaler suppliers who are evaluating the 800-volt or the 400 plus/minus rack architectures? Are they advocating more for GaN? Do they accept both GaN and silicon carbide options? How do you anticipate this unfolding from a technology standpoint when comparing GaN and silicon carbide?

Chris Allexandre executive
#19

Thanks, Quinn; that's actually a really solid question, because I believe there's a bit of misunderstanding out there. To start, I'd say we aren't advocating for anything; we offer both SiC and GaN, and we're fine with SiC being utilized on the primary side when necessary and GaN being utilized on the primary side when necessary. So we aren't advocating for anything. We're being pulled; we haven't observed any notable instance of board implementation in customer assessment involving SiC on the primary side. SiC sees broad use at 1.2 kilovolt, as I noted, in the standard AC-DC, correct, which essentially comes before the 800-volt DC; however, regarding 800-volt DC, customers have been pulling us. And I'm referring to hyperscalers, as Kevin asked about, that are truly fueling GaN uptake since it's more efficient and delivers greater density.

Quinn Bolton analyst
#20

All right. Then, Chris, you mentioned your 10-kilowatt all-GaN brick solution as well. Could you help us understand whether that's primarily a reference platform? Or is it something Navitas would aim to source as a complete brick-level product? Since my guess is it contains quite a bit of extra componentry. So if you're selling that whole brick, my guess is the dollar content could be quite high; so could you discuss whether you'd truly just sell the GaN portions within that brick? Would the whole brick be sold? And assuming the whole brick were sold, how would margins be affected?

Chris Allexandre executive
#21

Well, certainly, that's an excellent question. Thanks, Quinn. For our customers, we see this as an enabling solution and an enabling technology. So to start, like I said, this is work we've completed alongside the customer. We didn't do it alone in a vacuum, correct? This was co-developed with a leading customer — that's point one. Point two: we don't compete with customers, all right? Right now, we don't foresee a route where we'd be selling the modules. That design, though, gets shared with our customers and the hyperscalers, plus the ODM and OEM, who want to understand how we managed to reach that efficiency level, correct — 98.5% — which, judging by what we've observed and feedback from some competitors along with customers, we consider among the industry's best, right? So what I'd say is, for us, this is — this is what we've historically accomplished with GaN. We led the way with GaN in mobile by showing customers and assisting them in reaching maximum efficiency, minimum EMI, and maximum density. The same thing is happening in AI data centers. That's what I mean when I bring up 2.0 — we're drawing on both the advantages and the capabilities from 1.0, correct? And this kind of system expertise matters. Ultimately, our business involves selling GaN and silicon carbide and helping our customers. In fact, on that board, we rely on a few of our competitors for silicon and for other technologies and products we lack. Yet the aim is demonstrating to customers and helping them speed up GaN adoption in HVDC.

Jon Tanwanteng analyst
#22

I'd also like to add my good wishes for Todd on his next chapter. To get started, could you perhaps share some thoughts on how the competition looks when it comes to providing the 800-volt data center? When you're going after these sockets, who exactly are you up against — are they beating you on price, or outperforming you technologically? And beyond that, how is the collaboration with Infineon developing in this area?

Chris Allexandre executive
#23

Thanks for asking, Jon. Let me take the last part first; our work with Infineon is ongoing. There's a cross license between us, as you're aware, and our vision is aligned — we both want to speed up the adoption of GaN and silicon carbide across the AI DC, correct, so SiC for the traditional architecture and GaN for the 800-volt DC, right? So the 2 companies are engaged in a great deal of discussion on that subject, right? Second, all of you have observed that a number of vendors have appeared on the 800-volt AI factory ecosystem, so to speak. In fact, I believe the count stands at 13 vendors as of today. However, we don't encounter every one of them in each socket we pursue. So my suggestion would be to examine how many among those 13 truly offer high-voltage GaN — meaning how many provide a 650-volt GaN in the proper package capable of supporting 800-volt HVDC — and how many offer mid-voltage GaN for the 50-volt secondary side. A number of them appear as silicon vendors, okay? We appear as a GaN vendor. Another point: since we discussed SiC being applied to the AC-DC as well, there's a natural draw toward more SiC as voltages rise, and also beyond the data center — to achieve 800-volt HVDC, you must facilitate a shift in the grid architecture. That is purely a high-voltage, ultra-voltage SiC endeavor. So here's what I'd say: competition is substantial, yet not everyone is contending over the same thing. And few among the listed vendors possess either high-voltage SiC or ultra-voltage SiC — whether competing in the AC-DC at 1.2 kV or in the grid at 2 kV and above — while also offering high-voltage and mid-voltage GaN. So that competitive pool is in fact shrinking; that's why we're extremely clear on our focus. We don't participate in silicon. We participate in GaN — high-voltage, mid-voltage — and in high-voltage and high-voltage SiC.

Jon Tanwanteng analyst
#24

Understood. Then, for the second item, can you bring us up to date regarding the incremental margin on the 800-volt data center offerings, or on high-power products broadly, particularly while you bring on additional suppliers?

Chris Allexandre executive
#25

To begin with, as Todd mentioned, correct, we anticipate ongoing gross margin improvement; keep in mind that this year's growth is driven by every high-power market while mobile declines, correct, accounting for under 25%. What I'd say is that greater scale will contribute more gross margin. As revenue increases, a portion of our fixed costs get absorbed, and that produces margin improvement. Second, within the high-power market, high-power products carry a higher margin than mobile did, all right? And that mix will shift. And third, we are quite busy ramping up new suppliers, especially in packaging, which will assist us in lowering costs, correct? So there are several dimensions behind our confidence in gross margin improvement, just as we laid out plainly for the remainder of '26, correct? And as we gain further scale in '27, we anticipate that continuing.

Jack Egan analyst
#26

I wanted to ask something that builds on the gross margin topic that came up just before me. With mobile continuing to shrink, what I'm wondering about is how things look further out. Will your gross margins mostly be shaped by mix — meaning the split between data center and non-data center end markets? Or does it come more from the kind of technological innovation Todd brought up, which sounds like it points to new products carrying higher ASPs? Basically, I'm trying to figure out what's actually driving that margin expansion — is it end market mix, or is it improved product margins?

Todd Glickman executive
#27

Honestly, it will be some of each. End product mix certainly plays a role, no question. When mobile goes down, the high-power markets will deliver better margin for us. Those tend to hinge more on reliability and performance. Beyond that, though, as we scale up and as those new products enter the high-power markets, we anticipate additional expansion from things like optimized process, yields and packaging costs, and that brings our product cost lower, which in turn pushes our margins higher too.

Chris Allexandre executive
#28

One more point to make, Jack — sure, scaling, lower costs, and products with better margins, as Todd mentioned. But what I'd say is this: regarding this year's growth, I believe we've made it quite clear that Q4 marked the lowest point, we're steering Q1 upward, and we stated that quarter by quarter we'll expand throughout the year while margins widen. Let me stress once more that this growth stems from the entire high-power sector. Absolutely, AI data centers represent a significant chunk of what lies ahead; the SAM we presented only weeks back shows it accounts for close to 50% of the SAM we anticipate for ourselves by 2030. Still, I'd point out that performance computing is expanding now, correct? It will keep expanding and additionally benefit the margin blend. Grid infrastructure is genuinely speeding up. And I believe you'll witness products with greater ASPs and better margins entering the picture there, where reliability and performance carry more weight than price. Naturally, price does matter. Then, as we discussed regarding AI data centers — a market where cost sensitivity runs high — efficiency is everything at this moment, right? And I believe you'll observe every one of those sectors driving the growth in gross margin expansion, correct? So I merely wanted to adjust your question somewhat to ensure we don't view gross margin expansion as originating solely from AI data centers.

Jack Egan analyst
#29

Absolutely, Chris. That's really useful. Then, stepping back a bit, I realize you aren't providing as much to certain automotive and industrial markets, but silicon carbide has been through — generally put, it has been through a stretch of quite major oversupply. So I was just wondering, what do you expect regarding when supply and demand across some of those other markets could come back into balance, either for Navitas or for the sector overall? I realize you have some big volume wins, so maybe it affects you less, but any thoughts on that would help.

Chris Allexandre executive
#30

Honestly, Jack, I'd say, first, as someone who's been in this industry a long time, it will take a while, but I think that question belongs with the vendors selling into EV. We don't. We aren't in that same SiC arena, all right? You've heard me say very plainly that we compete and concentrate on 1.2 kV for the PSUs for AI data center and 2 kV and above up to 5 kV and even higher for the grid; here — this isn't a matter of supply scale, it's a matter of how dependable and efficient and high performance your technology is; so for certain SiC vendors working at 450-volt, 650-volt, 800-volt with an EV focus, that concern makes sense. As for us, it's about scaling with the ultra-high voltage, all right, which has nothing really to do with supply right now.

Richard Shannon analyst
#31

Sorry about the background sound — I literally just got off the plane and caught only part of this call. So hopefully I'm not asking anything that's already come up. But Chris, the thing I really want to ask about is the AI data center opportunity. How much of what you're seeing there comes through the Infineon partnership — riding their coattails, so to speak — versus through other routes? And separately, is there any cross-fertilization of wins inside the rack between point of load and the 800-volt down? Do those two cross-fertilize and hand you any extra benefit?

Chris Allexandre executive
#32

That's a really good question. Like I said, Infineon is our partner. There's a cross-license between us. That was put in place a couple of years ago. We keep pushing the collaboration forward to drive GaN adoption, at both high voltage and mid voltage. I'd say we don't leverage Infineon, and we don't benefit from them in that way. What you'll see is — I think Kevin asked me earlier, who do you run into the most? My answer would probably be Infineon, and definitely Infineon, since their vision matches ours; their technology matches ours — high-voltage GaN, mid-voltage GaN, SiC, and ultra-high-voltage SiC too, for grid infrastructure. So the similarities are very strong. So we run into each other; we trail each other, but I wouldn't describe it as leveraging Infineon, right? Now, what's interesting is that when we launched the package, we realized that — since we're talking to the same customers and thinking the same way — we wind up viewing the technology identically. That's how I'd put it. On your question about expanding the portfolio, that's something we're evaluating. My priority at this moment was pivoting the company, okay, and putting all our eggs, every engineer we've got, all our focus behind the 4 high-power markets plus high-voltage GaN and GaN in high-voltage SiC; even so, we are reviewing opportunities to broaden the portfolio. As voltage rises in the data center, circuit protection becomes necessary, and that's something we'll examine down the road, right? But at this point, we're laser-focused on executing with the product we've got and the one we just released.

Richard Shannon analyst
#33

All right, that's reasonable. Then the next thing, and this is likely directed at Todd, regarding gross margins. If I heard the close of his prepared comments correctly, he mentioned that the business doesn't yet have sufficient scale to generate real leverage on gross margins. Is there some revenue threshold at which that starts to occur? And what sort of fall-through margin should we expect once that path becomes visible?

Todd Glickman executive
#34

Yeah, excellent question. As the composition of what we sell shifts, clearly margins are going to rise. However, at this moment, that scale problem persists. We do anticipate margins beginning to widen once more in Q2 and afterwards. So at present, that's essentially the point where things tip.

Chris Allexandre executive
#35

Here's what I'd say to you, Jack: within the high-power sector, the high-power offering is arriving with better margins. As for mobile, it's declining — during Q4 we put that figure at 25%, and we stated it would keep falling and become negligible by the close of '26, correct? Therefore, my view is that we have strong conviction that this shift in mix — the growing share of higher-power product within the high-power market as a portion of the whole company — combined with the new product and the cost reductions we're implementing, will produce gross margin expansion. So you'll observe this plainly and unmistakably, alright, beginning not terribly far off from today.

Richard Shannon analyst
#36

All right. Thanks for that information. And, Todd, best of luck with whatever you do next.

Quinn Bolton analyst
#37

Chris, you devoted quite a bit of time to discussing the high — the 800-volt data center possibility. Yet you also mentioned that re-architecting the grid would be necessary. So I was just wondering whether you might take a moment to discuss what opportunity exists for high-voltage silicon carbide and solid-state transformers — like, where do things stand in the design process for a few of those solid-state transformers? And could you give a rough estimate of what the dollar content opportunity might be? I'm not sure whether that's measured per megawatt or per unit, but is there some way to size how much high-voltage SiC goes into a solid-state transformer as those begin to be deployed while the grid gets re-architected?

Chris Allexandre executive
#38

Quinn, thank you for that follow-up. In fact, I'm happy you raised it, since everyone has been concentrating on AI DC data centers and on the 800-volt HVDC architecture — which matters, because that's a discontinuity in the architecture, a substitution of silicon with GaN, correct, or with high-voltage technology — yet none of that can happen unless the grid transforms, and that isn't merely about achieving a more efficient grid; it's an architectural shift. You brought up SSTs, which essentially means stepping from 35,000-volt AC super-high-power, high-voltage lines all the way down to 800-volt DC at the greatest level of reliability; that's what propels the change. And I can tell you, I have never witnessed grid infrastructure evolving this quickly. So, as you asked — and I believe we mentioned this in the script and in the press release — we are speeding up sampling of our 2.3 kV and 3.2 kV. The applications cover any grid type application, SST obviously, but also battery energy systems, megawatt chargers, solar farms at the grid type level. All of those applications are in accelerated designs, and we are extremely busy speaking with all those customers. That's the reason I described it as 2 legs, all right? We have 4 high-power market. But looking ahead, those 2 legs — the AI DC and the grid — matter equally; and this is purely a high-voltage SiC play. As for the earlier question regarding EV, it isn't the same technology, because what must be delivered is high reliability technology, highly reliable modules. So it's a different play, correct? So I'm glad you raised it. We observe an accelerated design momentum. Naturally, it will take time. This design cycle is longer than computing. It's longer than AI DC. Still, I think you'll begin to see meaningful revenue growth starting '27. Regarding what you asked on content, during our most recent investor meeting, we essentially cited $25,000 to $35,000 for each megawatt of total content for Navitas as a SAM, and that figure rests once more on ultra-high voltage, high-voltage SiC — meaning 1.2 kV and 2 kV and above — plus GaN, while roughly $10,000 to $12,000 actually sits outside of data center. So when you consider $25,000 to $35,000, then $10,000 to $12,000 of that falls outside of data center, being purely SST, BESS and every one of those applications, correct? So — and once more, within data center, there's a split between SiC for PSUs and GaN as you shift to 800-volt DC. Yet we shouldn't underestimate how important the grid infrastructure is, nor its potential. In fact, a couple of weeks ago we put out a detailed SAM analysis showing that GaN and SiC each account for 50%-50%, all right, by 2030 when it comes to what Navitas could capture of that $3.5 billion we mentioned earlier. And you're able to see that the grid isn't far off the SAM of data center. Another point I'd make is that this is only the start. So when you consider grid, it's a multiple-decade transformation that will keep pushing voltage higher. We began at 2 kV. We're reaching 3 kV; we will reach 5 kV and above. And that is going to fuel transformation across the coming multiple decades.

Chris Allexandre executive
#39

We appreciate all of you being here on this call. It should be clear that the advances we're achieving make us quite proud. During the initial 5 months and 6 months after my arrival, the focus was on redirecting the business and getting clarity on our direction. That, I believe, has been accomplished; our attention now is on speeding up demonstrations of what our technology can do; there are 4 pillars to this transformation, as I noted earlier: concentrating on markets, leading in technology, running efficiently and maintaining financial discipline. We will keep bringing you updates about our headway; and in my view, a promising future lies before us.

Operator operator
#40

Everyone, that brings us to the end of today's session. Thanks to each of you for being here. You're free to leave the line now.

Read the full transcript via the API

You're viewing the first half of this call. Get the complete Navitas Semiconductor Corporation transcript - plus 132,000+ transcripts from 7,000+ companies, speaker segments and full-text search - through the EarningsAPI REST API or hosted MCP server.

Get an API key View API docs →

For developers and AI pipelines

Programmatic access to Navitas Semiconductor Corporation earnings transcripts and 132,000+ others is available through the EarningsAPI REST API and the hosted MCP server. Quarterly plans from $162 - full transcripts, speaker segments, full-text search, and the /api/v1/transcripts/recent polling endpoint for ETL pipelines.