Navitas Semiconductor Corporation (NVTS) Earnings Call Transcript & Summary

I think we can start...

Sure. I've got the PowerPoint downloaded, right? So I might as well control it on my side.

Sure.

Are we jumping straight into a presentation? Or what's our format, Edward?

I think so, yes. I'll just do a quick opener, and you can start. I'm sure attendees will file in, 14 on the call now. So thank you, everyone, for joining yet again our call for Q2 earnings review with Navitas. So thank you very much, Gene and Steve, for being on this call. I'm always excited to hear from you guys, and I'm sure the investors here will have a lot of questions. So let's get started.

Sounds great. Thanks, Edward, and thanks to all of you for joining and look forward to sharing some further updates here. So let me just get this set up for everybody. Okay. So if you happen to be new, I'll be brief here, but we are a pure-play next-generation power semiconductor company. By that, we mean we're when 100% focused on power next-generation technology. And there's two key ones here, gallium nitride and silicon carbide, they're both similar in that they provide next-generation performance compared to traditional power silicon, which has been used for the last 3 decades. Gallium nitride formed together to create a stronger, faster, more efficient bond than silicon for low to medium power applications and silicon carbide does something similar for higher voltage and higher power applications. And we're going to talk a lot about both of these technologies that Navitas is bringing to the power electronics market. Here's a good overview, just a picture on the Y axis, on the left is the power level, going from very low power applications, tens of watts, hundreds of watts, kilowatts, to very high power, tens of watts, 100 kilowatts, even a megawatt. On the X axis is the voltage. You'll notice, in general, the entire world of electrical products need power. So the market opportunity for us is huge, $22 billion. It's predominantly silicon today, but it's widely recognized that most of these markets will be transitioning to GaN or silicon carbide over time depending upon their voltage and power, 650 volts and below, 10 kilowatts and below, prime spot for gallium nitride. You go above 1,000 volts in voltage or above 100 kilowatts in power prime real estate, so to speak, for silicon carbide. And then there's a section in the middle that's actually not clear. And what we love about it at Navitas is we have the best gallium nitride. We have the best silicon carbide. And we have the best application experts to help our customers figure out which technology is best in each of these markets, including that middle section, where it's unclear for even industry experts on what's going to be the winner over time. It's a huge market, but it's also very important that we are focused. We're a small company. We cannot serve the entire $22 billion. So a lot about what I'm going to talk to you today is about our specific focuses in some of these key areas. Just on last year, we had started -- well, we started our company 10 years ago as a gallium nitride company, went 100% focused in that low-power application, specifically around mobile chargers, that continues to be a key benchmark market for us, a key driver for our business, but we've expanded rapidly in the last 2 years into higher power applications, in part through our acquisition of GeneSiC for the leading-edge silicon carbide technology, I'm going to address that later. But also we've spent the last 2 years developing higher power GaN IC technology for the higher power more industrial markets. This technology we call GaNSafe because it's the most safest, most reliable, most protected GaN in the world because we integrate safety, protection, reliability circuits right into that GaN chip. Now where are we deploying GaN? We're actually deploying a combination of GaNSafe and GeneSiC technology together in one of those overlap markets, one of those medium voltage, medium power markets, which is actually the AI data center market. This was already a very good opportunity for us to push more power, more efficiently into the server rack power supplies of data centers. Even 2 years ago, this is a market we were targeting developing our silicon carbide and our GaNSafe to go into this market. This market got a whole lot more interesting in the last 12 months with NVIDIA and the AI processor road map. It went from a big power challenge to an extraordinary power challenge. It went from people talking about doing 20, 30, 40 kilowatts per server rack to now talking about 120 kilowatts per rack, even 240, 480 kilowatts per rack. That's in the same fixed form factor to pack 10x, 20x, even 100x more power. It's very ambitious. It won't happen overnight. This market is still developing. It's very early days in the world of AI, as we all know. But there's a lot of exciting developments and a lot of exciting technical challenge for Navitas to pursue this market. We just started sampling our GaNSafe late last year and early this year. It takes about 12 to 18 months for new designs to come to production. So we'll see our first GaNSafe and GeneSiC-based data center products coming into production later this year, actually, a little bit this quarter and more in Q4. That's mostly focused on Hopper. But of course, we're working on a lot of Blackwell designs that will be going into production, some in Q4, most of them coming next year. And our technology is all about pushing up that energy efficiency, pushing from 95%, 96% efficiency with silicon and 97%, 1% or 2% may not sound like a lot, but this is 1% of virtually all the power running through the data center. And the cost of operating a data center, the #1 cost of operating a data center, is the cost of energy, the cost of efficiency. The second is the cost of cooling the data center. And if you have inefficient power supplies, you're creating heat that then cost you more money in thermal management. But in addition, you have power density challenges. As I said, they're trying to pack more and more power into the same size fixed server racks. We highlighted on our Q2 earnings call seven new design wins. We have 60 projects in development, so to convert more than 10 of them in 1 quarter is pretty good rates. We're happy with that. But again, these projects take time. We expect just a few million dollars in the second half of this year and then ramping from there throughout next year, a key growth market for us. At the same time, we spent a lot of time qualifying our latest technology, especially the GeneSiC technology to an automotive grade so we can be even more aggressive into the electric vehicle market. A year ago, we launched the latest generation, Generation 3 Fast silicon carbide MOSFETs, now we're announcing the automotive qualification of that technology. This technology can address many markets, but with the automotive qualification grade now complete, we can be much more aggressive into the EV space, namely onboard chargers but over time could go into traction inverters, could also go into the roadside chargers themselves. And let me give you a little overview here. Here again, it's a great -- we call it overlap market where I can use GaN or silicon carbide. In this case, it depends on the battery voltage, 400-volt battery, which is probably the most common today, gallium nitride over time can be the best fit. You can also use silicon carbide because it can handle even higher voltages. But if you go to an 800-volt battery, which is a growing trend, gallium nitride cannot handle the 800 volts, you need to move to silicon carbide. So here again, we've got the combination of both gallium nitride and silicon carbide that, depending upon that battery voltage, we can cater that to what the customer needs for that onboard charger to fast-charge the battery or fast-discharge that battery. And that's our main focus is on the OBC and the DC-to-DC converter. And that's what you're showing there in the upper right. We've actually done complete system reference designs with our customers and for our customers to accelerate their time to market. Ultimately, we are packing up to 3x more powerful or faster charging. We put a 22-kilowatt design in about the same size and weight as a 6.6 kilowatt design while improving the energy savings and again, reducing the cooling costs. Here, a lot of customers are in the pipeline. This takes even longer. Most developments are 2 or 3 years in development time. We started on most of them last year and some of them this year, a total of 200 projects in development, a total pipeline revenue of $600 million. But again, that would be ramping most of them, '25, '26, and some in '27. Last quarter alone, we announced 15 new design wins, which we're pleased with, but a lot more work to convert more of those 200 projects over time to drive sort of our medium to long-term revenue for the company. And you can see some highlights here of customers we've already announced. The silicon carbide is designed in the Zeekr car, one of the most popular and fastest-growing brands in China. We're designed into BRUSA. It's not a big known name, but they're actually one of the leaders in onboard charger design in Europe from many of the European brands. And we announced a joint lab of [indiscernible], one of the OBC leaders in China as well. And then the third area I wanted to highlight is our mobile space where we started with GaN, and we continue to innovate on this level as well. Last year, we announced our Generation 4 technology. This year, we've taken Generation 4 and put it in an all-new kind of form factor or a technology platform we call GaNSlim, very simple, highly integrated and actually lowers the manufacturing cost of using GaN for more price-sensitive applications. This has opened the doors to a lot more customers. We're now designed into 10 out of the top 10 mobile players in the world. That's smartphone, laptop and tablet chargers. That includes Xiaomi and OPPO, which are now producing 30% of their -- of all of their charters are using GaN this year. We're not the only supplier to those companies, but we do have a significant share and are benefiting a lot from that growth and our top line revenue this year. Samsung also expanded the use of our GaN in their lineup, not only from the Galaxy where we started with the S24 and S25, but now the Galaxy A, the fold and the flip line, so almost all of the Samsung smartphones are now recommending GaN chargers, which has Navitas GaN inside. And as I mentioned, we just launched the GaNSlim product. That has expanded rapidly. I think last quarter, we announced 25 or 30 projects. That's now up to 50. Most of those are launching next year. So we've got a lot of work ahead of us in supporting those customers. Those are not wins, those are active customer projects, and we're working very hard to convert as many of those wins and ramp them in production next year. And you can see a lot of great names that are already customers today, virtually the who's who in mobile charging, Lenovo, Vivo, Redmi, Xiaomi, Motorola, of course, Dell, Anchor and so many others. Solar is another interesting area for us. We have 100 customer projects here, a lot of the same products, again, silicon carbide with our GeneSiC acquisition and the GaNFast technology. Here, I wanted to highlight, in particular, bidirectional GaN. We announced this two quarters ago as a technology announcement earlier in the year. That's a very nice fit into certain solar applications. Silicon carbide is already shipping into solar customers today. This is a market that was hit quite hard by the interest rates a year ago, continues to be very slow. So we're feeling slower growth, slower revenue outlook here, but we're encouraged as we see GaN coming for the first time into the solar, into microinverters next year, with one of the leaders in microinverters announcing their plans to go aggressively again next year and to use a bidirectional GaN. And Navitas is a close partner to that need customer. 6 new wins in the quarter as well beyond that microinverter customer that I mentioned. So this is a promising area even though the market is quite soft in the near term. And finally, appliance and industrial doesn't get as much attention sometimes for big growth or for big demands in power. But here, again, silicon is dominantly used and step-by-step, this market is moving from silicon to both GaN on the lower power applications and a silicon carbide on the higher power applications. This market also is dominated by a long tail of customers. So there's a lot of small- to medium-sized customers out there that's creating a nice pipeline size for us. The pipeline is over -- well over $380 million per year, but it is a long design cycle. It tends to be 3, 4 years, sometimes 5 years. One particularly interesting area for me is the heat pumps. That's an area that is really eliminating gas heating for so many homes and buildings around the world. That's been using silicon traditionally. But by moving to GaN or silicon carbide, we can make the heat pumps both more cost-effective and more energy efficiency efficient to improve their consumer value prop and hopefully accelerate that market. But it does take some time, 2 or 3 years in development. And finally, on the financials, sort of rolling it all together, we look at our revenue profile over the last few years, really extraordinary growth from our production start in 2019 over the last 4 years. Obviously, in the last year, we, like the rest of the semiconductor industry, unless you're NVIDIA, right in the middle of shipping AI chips, we're in a semiconductor downturn, and we're feeling that. So obviously, the interest rates have slowed things down. Our growth rate has slowed. We had, for the first time, 2 quarters of sequential reduced revenue in Q1 and Q2. We haven't experienced that before. But with that said, our first half revenue is actually up 40% from the prior year, Q2 up 13% from the prior year. And we are cautiously optimistic looking at the second half of the year that we resume growth, but I'd say modest growth, and we're really in a wait-and-see to see how the rest of the market recovers, while we work very hard at converting as much of that $1.6 billion into further revenue growth in '25, '26 and '27. So I went pretty quickly here through a few of the highlights, key markets, key technologies and some of the key wins and pipeline progress we've had in the last quarter, and I'd love to open it up for some questions.

Great. Thank you, Gene. [Operator Instructions].

I should also introduce Janet Chou, our CFO.

Janet, good to see you.

Good to see you, everyone. My pleasure.

Okay. I guess while folks are getting their questions together, I'll have a few questions on my own. I think the first one is, what is the cash planning strategy for the company? I do see, of course, you guys have $112 million in cash on hand. Looking forward into 2025, it seems like you will have a pretty strong year in 2025. But other consideration here as far as where operating expenses and cash at the road.

Let me address the question. I think we made good progress in second quarter to drive cash burn rate down from $23 million in Q1 to less than [ $18 million ] in Q2. We will continue that trend. That is a collective effort across different functions and through improvement of working capital, negotiation terms with suppliers and spending control. As we guided, Q3 operating expense will be pretty much flattish to Q2. So this is a good demonstration of spending discipline. We're not cutting expenses at the cost of revenue growth. It's more about sharpening the focus and not to jeopardize the future growth. There's just so much we can do to optimize our cost structure. I'm pretty confident with all the efforts we put in place. We can continue to drive cash burn rate down. But of course, our top 1 priority is to gain market share, to grow revenue, profitable growth.

And is the expectation that the cash on hand is enough to get you to that point?

Yes, I'm pretty sure. Now the market is pretty much, third and fourth quarter, into the downward cycle, right? And once there's an inflection point of market recovery, with all the action plan in place, it can put us in a good track to achieve profitability and to accelerate the past profitability.

Got it. Thank you. So filling the first question here, when can we expect data center revenues to get in to roll in?

Yes. As I mentioned, second half of this year. So actually, this quarter, we're starting first production ramps, a little bit more in Q4, low single-digit millions this year and ramping into next year.

Got it. Thank you. And a follow-on question for myself for this space, obviously, we are seeing rack power increase tremendously. Where are you seeing customers finding their biggest challenges and then making -- getting to that realization, they have to adopt the GaN solution? And when they do look at the GaN landscape for power supplies, why do they end up using you? I suppose, in these customer conversations that you're having, this pipeline that you have for data center, what are some of the main drivers for customers who used to select gallium nitride that -- gallium nitride/silicon carbide, and then choose to go with Navitas?

Yes. If you break down the individual server power supplies and you could have multiple of them slide into each of the trays or the rack power, each of them, the average power in each of these server power supplies is probably only about 1 kilowatt, which is actually not pushing the limit. Silicon can do this quite easily. So where we see the opportunity for GaN and silicon carbide -- and by the way, we're finding the best power density and efficiency comes when you use silicon carbide on the front end called the PFC and gallium nitride in the second stage. So that combination of silicon carbide and GaN is getting us really extraordinary efficiency in density. But it's really when you start to push that 1 kilowatt per server power spike up to around 3 kilowatts and higher. We announced 3.2 kilowatts earlier this year, then we pushed to 4.5. Now we're trying to push that to 5.5. Now NVIDIA is asking us for 10 kilowatts and 12 kilowatts. So you really want to be on that leading edge. But the mainstream of the biggest volume are actually at 1 kilowatt. That's why it's not like the entire data center is going to convert overnight from silicon to GaN or silicon carbide, but that leading edge of power density, energy efficiency, where you're pushing to 3, 4, 5, 6 kilowatts and gone up. That's where we're seeing the tipping point. It becomes almost impossible to do those power densities with just plain look. And even if you can fit it in the size, the energy efficiency wouldn't be good, creating more thermal challenges. So that's where we see it. And then why do you pick Navitas? It's really two reasons. The GeneSiC technology is actually the best in-circuit energy efficiency you can get compared to any of the players out there. And GaN from others is just discrete GaN. We're the only GaN company making something like GaNSafe. Why do you care about ICs? We're building in protection, reliability and performance features that other people don't have at a very small modest price premium compared to the discrete GaN. But then it's the combination of both. Most of my competitors are either silicon carbide suppliers or gallium nitride suppliers, actually not offering both. And then we put them together in a complete server reference power supply design to show that when you combine GaN and silicon carbide together, you can push to all new levels of power density, energy efficiency. So it's a combination of all of those technical and system-level benefits that put us in a very strong position. So I think we're going to win more than our fair share of designs, but it's still early days, as I said, to really seeing the whole of this market.

Got it. And maybe just for clarification to the audience. These customers that are winning, are they data center, sort of cloud providers who are procuring these systems? Or is it the ODM that's coming to you? Is it the enterprise customer? What is the customer type? Maybe there's a mix.

It's a great question. It's very typical in most of our industries that we generally work with the middleman, the ODMs. In this case, the power supply makers. Even in mobile chargers and adapters, we're often not dealing directly with the mobile player themselves or their division that does the power to group or a separate ODM that does the power supply. In this case, it's power supply companies that have traditionally been doing the silicon designs for a long time. Compuware, [ Chicony ], Delta, Acbelt and others, heavily in China and Taiwan. So those are our main customers. That's where we spend most of our time with the engineering, working out the technical details, the commercial details. At the same time, we always love to talk to the NVIDIAs and the AWSs to try to understand future power requirements, what's coming, so we can do our best to be ready with our technology, our reference designs and use that to work again back to the ODMs to make sure we're ready for those future requirements.

Got it. Thank you. So next question here, probably can't answer, but I'll ask it anyway. So are you guys guiding to any revenue for 2025, '26? And are you guiding towards any breakeven point?

Yes. We're just -- of course, we guide 1 quarter at a time. We'll often guide full year soft guidance at the start of the year, so probably give an update on '25, in early '25, especially in this market environment, which is pretty choppy little clouded. So yes, no specific guidance for '25 and '26. Janet could comment certainly on the breakeven. But we've told people -- our target is $50 million a quarter to be breakeven. How fast can we get there? Depends on the market and how well we convert that pipeline into revenue in coming quarters.

Yes. As Gene we said, we guide 1 quarter at a time. Of course, we're going to focus on what we can do to return value to shareholders. We don't want to create any misperception or mislead investors but we just want the investors to have confidence in the management team to drive all efforts to drive profitable growth and accelerate the past profitability. Yes, we repeated multiple times, $50 million quarterly revenue will give us breakeven, but definitely we're even trying to work towards a lower revenue. That's our goal.

Got it. Thank you. And earlier, you mentioned this is certainly a function of cost reduction, certainly the reduction of OpEx here, but does that also include margin expansion efforts? I suppose this will factor in the programs that you have ramping in 2025, especially if it's in some of these new sectors that are a relatively new revenue streams for Navitas.

Definitely, we are trying to sell products that can differentiate Navitas from the rest, right? The one key differentiator that investors are looking at our gross margins. We're trying to diversify away from mobile. You see we have a strong pipeline in solar, in EV. That's definitely a margin uplift. And on top of that, we're also looking at options of expanding our supply base and continuing to drive down product cost to do yield improvement, to do die shrink or to do 6- even to 8-inch transfer.

Got it. Has any of the CapEx investments you guys made in the quarters ahead started to make a difference in how you guys manage supply?

Let me first comment on that. I think right now, it's too early to call it any meaningful impact, but all this CapEx investment we actually have deployed to some of the suppliers and foundry partners used for different purposes. And eventually, I'm pretty sure this is going to generate the return that we would like to see.

I see. There was a pretty significant investment in epi machines some quarters back. Are you saying some of these machines are being distributed out to your suppliers? Or are they still being held in your facility?

Yes. We announced a 3-year phased plan to invest in epi. We just -- the first year was actually 1 tool, but we think it was cool but with the slowdown in silicon carbide demand, with EV and [indiscernible], we put that on pause on hold, and then we're evaluating what's the right location and timing to bring that online where it's kind of give us the cost savings anticipating.

Got it. Thank you. Next question here is, will Navitas consider the Samsung GaN foundry, which will be introduced in 2025?

Yes. We're always looking at options on different GaN and silicon carbide foundries. Of course, today, we're exclusive with TSMC on GaN. We're working with X-FAB on silicon carbide. Of course it make sense to work with multiple suppliers over time. We have to make really carefully. GaN or silicon carbide is very hard to bring up. It's not quick. It's a very strategic decision on who you want to partner with, what time it takes. So obviously, we're not going to announce something before we're ready to announce it, but I would say we keep our options open, and we're always exploring possibilities to bring on other foundries over time.

Got it. And I think there is a lot of information in those slides talking about new industries that were ramping '25, '26. Can you comment at all on what the expected mix in revenue will be?

Yes, it's pretty broad-based right now. But if I were to pick the three that are the strongest from my view, mobile continues to be our core business, but it's going to decline as a percentage just because we're ramping into new markets. I think data center is obviously very promising for sort of -- I call it medium-term growth. The design is going to take 12, 24 months. So we're starting to see those ramp early this year or next year, of course. And then EV is the -- it's probably the biggest. It's the biggest overall demand when you consider GaN and silicon carbide, but also takes the longest. I would pick any one of those that's going to dominate. We see a nice multi-market growth, but I would point to those three as probably the bigger growth drivers for us going forward.

Got it. Thank you. The next question is any comment on the competitive landscape and if you expect consolidation in the supply chain by way of mergers or other M&A activity?

Yes, there's already been a decent amount. Of course, some of the GaN and you have Infineon, who acquired an GaN Systems, Renesas acquired Transphorm. There's not many players out there. So there's a decent amount of scarcity value, I would say, in the world of GaN. It's the newer technology. Silicon carbide commercially started developing 10 years or earlier. While we haven't seen much change in the competitive landscape in the world of GaN. We continue to be the pioneer in only real GaN IC company monolithically integrating a lot of these circuits, the driver, the sensing, the protection, the reliability. So that continues to be our key differentiator along with our system design capability. I haven't seen any new players enter, hasn't seen any major changes. So I don't think competitive landscape has changed much there. Silicon carbide, too, I don't think there's any big announcements. Of course, we're a smaller player but with really differentiated discrete. Now we don't have a silicon carbide IC, it's a vertical structure, so it would actually be very difficult, if not impossible. Not impossible, but not economical to integrate the other circuits. But within a world of silicon carbide discretes, the GeneSiC technology offers a really nice advantage for in-circuit high-efficiency differentiation under switching applications and the most robust applications. But back to competitive landscape, I haven't seen much change there. Consolidation is, of course, it's -- I don't want to speculate on what might happen over time. But I certainly don't see anything imminent or we haven't heard of anything imminent.

Thank you. Next question is, will Infineon provide GaN solutions for AI servers as well? And if you guys go head to head, what is the advantage here?

Yes. Infineon is already selling GaN today. Of course, they bought GaN Systems, as I mentioned, but whether it's internally developed again at Infineon or through the GaN Systems acquisition, it's still discrete gain. So we come back to our fundamental advantage of these GAN ICs. In the case of data centers, it's the GaNSafe technology. We offer a 20-year warranty to kind of back that up based upon the high reliability of the integrated protection and reliability circuits, plus the protection or the performance advantages of integrating drive and other features. That gives us a lot of component level and system-level differentiation compared to discrete players like Infineon.

Okay. Thank you. Another question is, you mentioned that the high interest rate had a negative impact on solar. Just some comments on the other segments, what other sectors that you guys are pursuing have push out or delays?

Yes. We saw it first -- I mean, I think the industry did. I don't think it was unique to Navitas. I think the industry -- solar industry felt it first last summer. And so we're sort of a year in of not seeing much recovery, but it does seem stable. I think companies like Enphase and SolarEdge seems to stabilize their inventory channel that they're reporting seems to be stabilized and reducing. So I think we see stabilization but not significant short-term growth. Around Q1 or -- yes, around Q1 of this year, we saw some slowdown of industrial and EV as many others reported late last year. So I think all of these had some dominant effect. Even consumer and mobile has been pretty flat or even down a little bit last year, but we benefited from companies like Xiaomi and OPPO converting their existing silicon designs over at GaN. So even in a flat market, we saw nice upside which helped to keep our revenue growing late last year and even stabilizing things a bit in the first half of the year. So I don't think there's any markets that are really immune to the economic slowdown and the interest rate changes that we've seen.

Got it. Thank you. The next question I think I got here is why gallium nitride and silicon carbide makers all around the world seem to be having a hard time making money today?

Yes. Well, maybe analogists, too. Industry is in their early days. I mean look at Tesla, right? Tesla spent, what, a decade building amazing electric vehicle cars, in my opinion, not making a penny losing an incredible amount of money. And in some ways, profitability is a choice. You could just stop investing in developing the technology to fast pace. You could stop investing or slow down investing new markets, helping people to learn how to adopt this technology. So you could drive off ability even faster with lower investment. But in the end, it's going to slow down the top line. So it's a balancing act. I think Janet summarized it really well that you can't cost-reduce your way to success and leadership. So it's a balancing act. And I think GaN and silicon carbide are still early days. We're on the fourth generation, fifth generation of gallium nitride. I try to think the silicon carbide guys also third, fourth, fifth generation whereas silicon after 3 decades, probably on the 30th generation, mature and will monetize. So we're in the early growth phase. It doesn't feel like it's sometimes when it's a little bumpy in these market slowdowns, it's not growing as fast as everybody thought. We feel it. Everybody feels it. But we want to be disciplined. We want to be thoughtful but we don't want to lose sight of the opportunity to really lead and expand in this market in the early days when we're defining the future leaders of next-generation power semiconductors.

I think really important guiding principle or guiding thesis at MESH when we first invested a long time ago was the market for power electronics is very large, and it's almost all silicon, right? And that conversion from silicon to gallium nitride and silicon carbide is going to take time, but eventually, the economics of that market will push that conversion to happen, one way or another. And so perhaps it will be helpful to just gain that perspective again. If we look at power electronics market today, how much how big is that? And how much of it has actually started to convert.

Yes. And if you look at the $22 billion, which is not the entire market, but that we think is likely to convert over time to GaN and silicon carbide, the total conversion today is in the range of $4 billion. So you're [ 20%, 18% ] there across GaN and silicon carbide. It's analogous to me when I started my career in the '80s, we were in the middle of converting from bipolar transistors and linear regulators to silicon power MOSFETs and switching power supplies. If you look, these sort of transitions don't happen that often. Every few decades maybe seems to be the pattern we're seeing now. And that transition took over a decade. But you went from single-digit adoption to tens of percentage, 20%, 30% adoption. That 10, maybe 15 years, it actually changed the market order of things. All new leaders were created, all new leaders were created both at the power system level and at the chip and technology level. And I think when you're over that 10 or 15 years, when you're middle of it, it can be a little bumpy. It's semiconductors. It's going to have up and down cycles. But when you kind of zoom out, as you're referring to, if you're missing the right guys at the right time on the front end of that conversion, you can really end up in a good place after that decade. So we're certainly taking long-term view on being that leader over time but also trying to be disciplined in short term to drive our company towards profitability.

Got it. Fantastic. Thank you, Gene. Are there any final questions from the audience? Well, it seems from Gene, the strategy is buy and hold. We don't give it as an advice. That was not investment advice. But I think this is -- at least what we at MESH has always believed in, is that this is going to be a journey, but it is very inevitable, right? This change has to happen. The efficiency has to be there. There are several industries that need to convert. It's not a single industry that needs to convert, but it certainly will take time and it will certainly be subject to market forces. But what is, I think, reassuring to look at here is, well, really that landscape, just like Janet alluded to earlier, really isn't that complex. There really are not that many players who have the ability to not only have the underlying technology but also the design teams around the world to service their customers. And so while this may take time, what I've learned in semiconductor investing is, these design wins now evolve very quickly and these design wins snowball well horizontally within the customers really quickly, right? So once a Xiaomi version 1 phone, as they usually use it for the more expensive phone, you start to get it in the other phones, and that comes with fairly minimal additional sales effort because they've already proven the economics to themselves. We believe that will be the case with a lot of the design wins that you guys have in the solar automotive data center sectors. And so it will some time. But at least from MESH's side, the conviction is still very high.

Totally agree, well said, Edward.

All right. Well, thank you, everyone. Thanks, Gene. Thanks, Janet. Thanks, Steve.

Thanks to all of you.

Bye, everybody.

Take care. Bye.

Bye-bye.