Monolithic Power Systems, Inc. (MPWR) Q4 FY2025 Earnings Call Transcript & Summary

Welcome, everyone, to the MPS Fourth Quarter 2025 Earnings Webinar. My name is Arthur Lee, and I'll be the moderator for this webinar. Joining me today are Michael Hsing, CEO and Founder of MPS; Bernie Blegen, EVP and CFO; Rob Dean, Corporate Controller; and Tony Balow, Vice President of Finance. Earlier today, along with our earnings announcement, MPS released a written commentary on the results of our operations. Both documents can be found on our website. Before we begin, I would like to remind everyone that in the course of today's presentation, we may make forward-looking statements and projections within the meaning of the Private Securities Litigation Reform Act of 1995 that involve risks and uncertainties. The risks, uncertainties and other factors that could cause actual results to differ from these forward-looking statements are identified in the safe harbor statements contained in the Q4 2025 earnings release, our Q4 2025 earnings commentary and in our SEC filings, including our Form 10-K and Forms 10-Q, which can be found on our website. Our statements are made as of today, and we assume no obligation to update this information. Now I would like to turn the call over to Tony.

Thanks, Arthur. Good afternoon, and welcome to our Q4 2025 earnings call. Today, we made an announcement that after 15 years at MPS, and 10 as a CFO, Bernie will be retiring. Before we begin our prepared remarks, I'd like to turn the webinar over to him for his thoughts on his time at MPS and the transition ahead. Bernie?

Thanks, Tony. As I was preparing for today, I realized that this is my 40th earnings call as MPS' CFO. That's a nice round number to finish up with. It's been my pleasure and honor to work closely with Michael for so long and to have been a part of MPS' leadership team. I want to thank our investors and analysts for the trust you placed in me. It is natural for all businesses to go through cycles. Your support has been consistent regardless of the circumstances, and I have greatly appreciated it. As I look ahead, MPS' prospects remain bright. With our unique culture, our dedicated employees and a fantastic portfolio of products, MPS is well positioned to sustain the broad-based growth, you've come to expect from us during the last 10 years. I have a lot of confidence in the team. I am transitioning my responsibilities to. Starting with Rob Dean, who will be our interim CFO, while many of you may not know Rob, we have been partners in this enterprise for the last 9 years. And like my predecessor and for myself, Rob continues the MPS tradition of transitioning from MPS' controllership to CFO. This ensures a lot of continuity in the role. Likewise, I will remain with the company to support the successful transition. Rob, would you like to say a few words?

Yes. Thanks, Bernie, and I'm grateful to you and Michael for this opportunity to continue the tradition and to have been part of the MPS finance team where you been CFO. I know I speak for the entire finance team when I thank you, Bernie, for everything you've done over the last 15 years. He's not only helped guide the business to consistent growth and execution, but he's grown a great team around him. I work closely with Michael and the executive team for close to a decade. We've developed a strong relationship, which I expect to continue as we grow the company and take on the opportunities ahead. I appreciate the confidence they have placed in me in this new role. I look forward to meeting all of you in the coming days and weeks. With that, I'll pass it over to Tony.

Thanks, Rob. I'll now move to our prepared remarks before going to Q&A. In 2025, MPS posted its 14th consecutive year growth with a full year revenue of $2.8 billion, up 26.4% from 2024. For Q4 2025, we had a record quarterly revenue of $751.2 million, 1.9% above Q3 2025 and 20.8% higher than the fourth quarter of 2024. This performance reflected our consistent execution, continued innovation and customer focus. Let me call out a few highlights from 2025. Our non-Enterprise Data end markets grew by over 40% year-over-year, showcasing the strength of our diversified business model. We achieved our milestone of securing more than $4 billion of geographically balanced capacity and continue to add additional supply chain partners to support future growth. We had record module revenue and positioned ourselves for a further shift to solutions by sampling our 800 volt power solution for data center. In automotive, we launched solutions for 48 volt and zonal architectures, including the first fully integrated 48 volt e-fuse and a kilowatt level zonal controller that will support growth in 2026 and beyond. We expanded our customer base and data center for power solutions across AI, server, memory, optical modules and switch applications with leading-edge current density, power efficiency and packaging. I am also pleased to announce that our quarterly dividend will increase 28% to $2 per share. For the 3 years ending with December 2025, MPS has returned over 72% of free cash flow to stockholders through share repurchases and dividends. Our proven long-term growth strategy remains intact as MPS focuses on innovation and solving our customers' most challenging problems. We continue to invest in new technology, expand into new markets and to diversify both our end market applications and global supply chain. This will allow us to capture future growth opportunities maintain supply chain stability and quickly adapt to market changes as they occur. I will now open the webinar for questions.

Thank you, Tony. [Operator Instructions] Our first question is from Chris Caso of Wolfe Research.

Thank you. And Bernie, I'm lucky to be the first to congratulate you on your retirement and wish you all the best, it's been a pleasure all these years. So for my first question, I guess, as we look into the March quarter, could you give some color on what you're seeing with respect to the different segments, what do you see within the various market segments?

Sure. Let me start to call this by talking a little bit about Q4 '25. We saw a good step-up in the ordering patterns in the quarter. Our book-to-bill ratio was well in excess of 1. And that's really reflected in our backlog, which is starting to extend out into Q2 and Q3 of '26. We also finished with fairly routine rather -- our channel inventory stayed at the low end of our range. So we feel that we're servicing real demand and that we're seeing a lot of strong ordering trend. So as we look at the fourth quarter. You can see that we saw some pretty good strength, particularly as it relates to enterprise data and also to the communications. We expect that those trends, along with automotive should continue to extend into Q1 and into the remainder of the year.

As a follow-up, and you mentioned Enterprise Data. And obviously, that's been a focus of attention, not just for you, but the whole market. You had made some comments on Enterprise Data for '26 on the last earnings call. And if I just annualize the Q4 numbers, you pretty much get to where that guidance was. So what are your thoughts on that in the year? And perhaps is there a seasonal element to Enterprise Data as we go through the year?

Sure. I'll start off on this one. As I said, in Q4, we saw some fairly pronounced changes in ordering patterns which has given us a fair amount more of confidence as far as what the outlook for Enterprise Data could be in '26. Now I think for those who worked with me for the last 10 years, you know that I like to stay pretty conservatively profiled when I make an estimate. So I'd probably say that whereas last quarter, I talked about a range of between 30% and 40%. Maybe I can increase that to a floor of 50% growth for 2025.

Well, 50%? I thought that we can do a lot more than that.

Conservatively.

Well, I see -- this is what I see here. We won the many design wins, and across the board, not from 1 company, 1 large company, okay, we have multiple customers. They're all very big. They call what [indiscernible] 7 or 8 or whatever, okay, we won all the designs again. And we are proven, we are one of the valuable AI power supply. And Also, I see the other end, we have all the capacities. We can deliver this year to our customers' needs. And I don't see why not, it's not only 50%, we will be a lot more than that.

Maybe just 1 last thing to add. Chris, you've heard us talk about the drivers for growth, which is really around growing existing customers adding new customers, seeing new platforms coming to market and then, of course, just server tailwinds. I think we always knew those were in place and now we're just seeing some of the backlog to go along with it. The only other thing I'd add, right, is we focus a lot on Enterprise Data, but I think we've seen a strong Data Center demand, which is really also pulled through storage growth, optical modules, switches and other areas we've talked about. So I think overall, we've seen strong Data Center demand through the end of the year.

Our next question is from Joe Quatrochi of Wells Fargo.

My congrats to Bernie as well. Thanks for all the help. Maybe just to follow up on that, in the enterprise data and the increase in outlook. How much of that is related to like traditional server CPU demand that it seems like it is accelerating as well.

Well, as I said, we have a lot of new design wins in the -- particularly last 1.5 years. And we see continuously change -- adopt our modules and even from a changing from silicon to modules. And we see the trend. And with our power densities, I mean we're winning the market.

Yes. I think to go back to part of your question, which had to do with the traditional maybe CPU data center. The lines between AI, GPUs and CPU are getting pretty blurry because they're so integral to 1 another these days.

They're using the same kind of a power supply now.

Exactly. But I would say that we've been trending very well in both categories. So I can talk to a trend line, but I can't really give you an absolute figure.

Okay. That's helpful. And then as a follow-up, as I think about just kind of like storage and compute and maybe the exposure to like PCs, are you seeing anything related to just kind of memory prices increasing and just kind of maybe some pressure on some demand destruction around that part of the market?

Oh, you're talking about PC. The PC is a different animal than the data centers.

New question, he is asking about memory and the constraints there, whether we're seeing that affect us?

Oh, memory constraints. Okay. As I said earlier, we don't have constraints on the capacity side.

I think where you're going, Joe as well is also is there are any demand destruction in PCs. I think as you looked at Q4 into Q1, remember, we're coming off a very strong first half of 2025, so we expected that to be down a bit as well as we are participating more selectively in the emerging parts of that business. I don't think we know how it's going to play out through the rest of this year at this point. So I don't think it's possible to say how that market might trend. We hear a lot of the same, but I think it's too early for us to tell.

On the memory constraint on the PC?

Yes.

I don't know. We don't know those markets. Okay. We don't know what our customers do. We deliver what our customers ask, I care less.

Our next question is from Josh Buchalter of Cowen.

Definitely want to echo the congrats and best wishes to Bernie after an incredible run and a long, sometimes strange trip. I very much appreciate the support over the year, and congrats and best of luck to Rob. Maybe just to start the incremental confidence in Enterprise Data is great to see. And it seems like you guys are suggesting, you feel better about visibility there than you have in the past. Is that a fair read? And if so, as part of this, it's just the market is maturing and scaling and also just the capacity needs are so great. I was just hoping you could maybe speak to how visibility compared to maybe a year ago or something?

Yes. I think in prior quarters, I've said that we've been experiencing a turnaround, much of which has been around the enterprise data or more broadly AI markets but that the anomaly had been that we've been seeing very short lead times and that they were not putting a lot of backlog in our books. And I'd say that the fundamental change that also making us more confident right now is that we are seeing longer ordering patterns because some of our customers are concerned about capacity constraints, not necessarily with us, but just in general.

Okay. That's helpful color. And the 40% non-Enterprise Data growth number for 2025 is obviously huge and above your historical algorithm of I think it was 10% to 15% above the analog industry. Is that still the right way we should be thinking about the non-compute exposed verticals into 2026 as well? Congrats again.

I will try to manage your expectation. We have not said '26, we have a 40% growth, okay? And this growth, okay, we are still a small players in the overall markets compared to all the market size, okay? And some growth in some years, we see the opportunities we can grow better than any other years, okay? But the long-term trend or even short-term trend and even '26, we will grow.

Our next question is from Quinn Bolton of Needham.

Bernie, it's been a great run, a great decade. So thank you for all your help along the way. Welcome Rob. I wanted to ask, some of your competitors in the AI power space are talking about their business is doubling in 2026. And I know your business and their businesses don't overlap 100%. There's different compositions. My question is, do you guys think that you're gaining share as you look into 2026 broadly in the AI power segment?

Yes. I'll refuse to get in to pis**** contest, and the whole -- who wear less on the stage, okay? We let the numbers speak itself. As always, as our -- since IPO 21 years history, we never do that.

I do think that we clearly have great products, broad engagement across the customers. We have design wins in a broad swath. So how the market then plays out remains to be seen, I think. But I think we're very confident in our product portfolio and the engagements we have right now.

That's a much better small talk.

Well, as you said, we'll see where the numbers shake out at the end of the year. Michael, I wanted to ask you -- you talked about sampling your 800...

You probably see much early.

Oh, good. We'll stay tuned. I wanted to ask you about the 800 volt solutions for 800 volt racks, some of the participants in the market are suggesting NVIDIA and others are looking for GaN-based solutions. I think you guys are offering a silicon carbide-based solution. And so just wondering if you can talk about what you're seeing in the market. Is there a preference for GaN or silicon carbide? Do you think there will be a mix of compound semiconductor solutions for that 800 volt to 12 or 6-volt stage in those 800 volt racks.

Yes, yes, yes. Okay. Again, okay. And this is not a good venue to talk about technical terms. Actually, I'm happy to be a person who really knows the semiconductor device. And we are developed silicon carbide, okay? And 10 years ago, I was wrong about GaN. But in the last few years, we developed our own GaN devices, okay? And 800 volt since Tony has mentioned about it, and we're entering a pis**** context, okay, that revenue is not for this year, not for even for next year. So I can maybe end of them next year. However, we are the first company to sample it. Now that's a part of a pis**** contest talking.

And again, we've done a good job and shown ourselves to be very adaptive changes to the market. So whether it turns out to be GaN or silicon carbide that is what's demanded I'm sure that we'll be well positioned to take advantage of it..

Our next question is from Rick Schafer of Oppenheimer.

Thanks. And I would just say, Bernie, it's been a genuine pleasure. You're going to be missed. And Michael, I just want to confirm for everybody that you're never retiring, right?

Well, it happened to be today is my 86th earnings call. I'm looking for doubling it, okay. Well, be serious, okay? And I enjoy this MPS, okay. And immensely, okay, we actually created this platform. Everybody can maximize their capability. And the more interesting things to me is we -- company evolve. And we form -- we sell semiconductor, power management. Now in the semiconductor, we're getting to MCU, we're getting in to a data converters, we get to even high speed, okay. These are a few gigahertz of stuff. And you will see the revenue soon. And on the overall marketing -- market segment address, we migrate from silicon to systems to a module to systems, and you will see a lot more. I am enjoying this process a lot and I'm a part of it, again, and I have on product line.

Thanks, Michael. It's reassuring. Also, I just had a quick clarification and then I've got a couple of follow-ups. But the clarification, Michael, you said earlier, that CPU, GPU and server are using the same power supplies now. So does that mean that server CPU is already migrating to 48 volt?

It still -- they use mostly -- okay, if they have some advanced high-priced servers or special servers, a special need. They use still use 12 volts. And but I'm aware of some models in the 48 volts, my guess, it's still small. I'm not very clear on that. But this is -- and the majority is still 12 volts now they -- it's clear the module is the way to go, and if you want to improve efficiency.

So my first question really is on optical transceiver because that's basically a brand-new product line. A little over a year old, I believe and by our count, in our model, it's close to roughly 5% of sales exiting last year. I think now, which is a pretty remarkable ramp. So I guess I'm thinking -- we're asking what are your expectations for that business this year? And what does that imply for comms segment? What are sort of the puts and takes within the comms segment?

I can comment on this because we entered the module journey since 2016 or '17. And -- these happen to be the highest power density product on the market. Then optical modules, they wanted that because they have limited rooms. So okay, in terms of business, I don't know the details and maybe Bernie -- Tony can answer it okay.

Yes. I'll just follow up. I think, Rick, we have obviously seen great growth in optical modules over the past 1.5 years. I think the way we look at the market very typical for MPS is sort of interconnect. So it's not just optical modules, but engagements for CPO, active copper or other things as well because the market will then figure out what interconnect technology is actually going to succeed over the long term. We obviously don't guide by some end market let alone sub-end market, but we would expect optical module to continue to grow as you start to see the 1.6 ramp, as we go through. And for communications, it should be an area of growth for us in '26, both on optical modules and in switches because that's where our data center switches are as well.

And then if I can sneak in 1 on automotive. I mean, obviously, a great year '25. And I'm just curious what are the -- what you see is the top drivers of segment growth this year. I know you highlighted 48 volt Zonal, and I didn't hear you say much about ADAS, but I assume ADAS. I don't know if you can update us on how big ADAS is within that segment now? Or -- and if there's any way to quantify sort of a shift this year that you expect in potential content per vehicle.

Sure. I want to take a victory lap on 2025 where automotive grew 43% year-over-year and...

But that's only the beginning.

That's exactly the point here is that -- what we saw in 2025 and is going to continue is that while ADAS certainly was a strong initial ramp, particularly in '23, '24 and '25. I think we saw a lot more diversification into other content opportunities on the automotive platform. And so as we look ahead here, keep in mind, we're not necessarily driven by the SAAR of the business but we are -- our growth is dependent upon how fast our customers implement these new technologies, particularly as it relates to Zonal and 48 volt.

Or even ADAS.

Or ADAS.

And the majority of cars on the market is nowhere anything close to what Tesla does, okay? And that's the car I drive, okay, you have a full -- I don't drive anymore. And I think a majority of people still drive, okay. And that adoption rate that came in a large automotive company, they do things very slowly and much slower than a Tesla does. And for futures, we are up to this point after the next couple of years, I see our products provide the complete power supply chipsets, and also we have all the firmware, software win very much engaged with all the carmakers. And I don't see why not. That business is going to continue to grow. You actually guys -- okay, you know how many cars shipped with ADAS, which levels. You can count MPS in it.

And Rick, we're a little hesitant to probably call any numbers for the full year just because there is a lot of macro uncertainty still, great design wins, great engagement with Tier 1s and OEMs, but whether it's tariffs, whether it's the end of EV subsidies or whether you can talk about what the impact on the auto market is from the memory shortage, I don't think we know. So I think we're a little hesitant to actually put a growth rate on it for the year.

Our next question is from Gary Mobley of Loop Capital.

Bernie, your retirement is well deserved and look forward to working with you, Rob. I think everybody on the call would share the same sentiment that I have that you're definitely one of my favorite CFOs and for my retirement gift to you. I wanted to throw you a big softball question, but I think it's an important topic. Looking back over the past decade, when you've been CFO, you've outperformed the overall analog chip market, the overall voltage regulator market, consistently every year and seemingly for different reasons each year. But thinking about the outperformance of the market in 2025. Maybe if you can give us a sense of what drove that? Was that just share gains and volt's regulator die? Or was this something more substantial like moving into data converters? Was it tied to the higher content associated with modules and related, can you give us some KPIs as it relates to sort of your module mix right now? Anything you can help us get a better understanding of that consistent market share growth.

Sure, Gary. And thank you for the kind words. They're appreciated. When you look at the overall performance for the company in 2025. We had -- what had been our largest revenue end market in enterprise data, had actually declined 2%, and yet overall, the company grew 26%. And strategically, how we're differentiated from our competition is that we are represented with the best technology, the best services across all of the end markets that we service. And this is really just a reflection of our execution against that strategy over all of these years. It wasn't that, pardon my saying, we pulled the rabbit out of the hat. We actually are able to adapt very quickly to changes in the market. So that's what this is really a reflection of in our performance in 2025.

As my follow-up, I wanted to ask about maybe some nuances in your increased visibility and incontinence regarding that. If I talk to you guys 3 months ago, I think you were thinking maybe the 2026 year was going to be a little more second half weighted just given the stronger bookings that you've seen, the stronger order backlog as you see here today, would you say the shape of the year is a little more linear, less dependent on the second half?

I'd say that the first half for enterprise data in particular, but for the company, it is more secure I think there's still a lot of variables that need to be shaped before we really understand what the second half trajectory is going to look like. But obviously, the initial signs that we saw from the ordering pattern in Q4 and continuing into the year have been exceptionally positive. So now we have more of the high-level issue of trying to figure out what's real demand and what may be some double ordering on the part of our customers as they try to secure capacity. As I said, that's a high-level issue, and we've shown that we can adapt to that as well as anybody by the performance we gave in late 2020 and early 2021.

We work with the customers very -- especially all these large data center customers very closely. And they will give us a very good lead times and forecast. So we have the capacity ready and we just meet their demand.

Our next question is from Tore Svanberg of Stifel. Tore.

Congratulations, Bernie. You're a class act, and I'm going to miss you tremendously. My first question, Michael, I'm going to zoom in on -- in a market where there's perhaps less contests, which is storage and especially SSD power. It seems to be an area that could see quite a bit of upside and growth in data centers this year. So I was hoping you could talk a little bit about the profile of that business? I mean I think historically, it's been more tied to client and edge devices. But again, what's the company's position, SSD for data center going into 2026?

Yes. These are the power management and also the single processors, these are all in the consortium driven by JEDEC. I mean maybe Tony, you are a lot more familiar than I do. And we are part of it. And DDR4, we don't have much business, okay, very little. And DDR5s, we're in -- we're in on the dining table, before DDR4 wasn't, okay? And -- so now this business is ramping, came in memories and that came in all shifted to DDR5 in the game. And we clearly see the volumes now in last year and this year. And we're not stopping there. We'll migrate down to -- we expanded our product lines to the single site. And this is all in the memory modules.

And -- so I think there was a portion of your question as well about the non-DDR5 part of that business, sort of SDD and HDD and we have seen an uptick in that part of the business as well. And I think you're right, it's being pulled through much more by the enterprise than consumer these days. And that's why we generally talk about storage being data center-driven, inside of the storage and compute segment.

Yes, that's what I was trying to get to. And as my follow-up, Michael, congratulations on getting to $4 billion in capacity, but it looks like you're going to need quite a bit more of that. So perhaps you could give us a little bit sense of what you're doing on the capacity front, especially in the next 3 years because you're clearly going to need much more than $4 billion.

Yes. We are very aware about that. Okay. And speaking, of course, and we're continuously expanding our capacity MPS history, okay. the worst things are shutting customers down, okay? And fortunately, we haven't happened in the MPS, okay. And now -- gets a little -- it gets a little more complicated, okay? We established the last -- from the beginning of the last year, we established our supply management. And this is not only for silicon and we -- and not for semiconductors, including silicon carbides and gallium nitrides and the materials and we do our modules and all the module components. And so we established that the supply chain management. So I think it's -- and also quality, don't -- okay, it's not everybody can play the game. Every supplier can play the game. These are -- we go through heavy auditing. And so their standards, okay, they meet our standard ultimately is a reflect into our margins. And so the short answer is, yes, we are expanding very fast.

Our next question is from Kelsey Chia of Citi Research.

Bernie, congratulations on your retirement. Really appreciate the opportunity to work with you over the past year. So I think on my first question is with regards to the updated guidance for Enterprise Data. Is there any market share gains assumption there? Or is it just primarily due to this industry growth? And also MPS has demonstrated strong execution and historically gained share during periods of supply constraints. So is it fair to assume that MPS could navigate any potential supply constraints and see -- and to take share in this environment?

Sure. And I think that I'd be doing a disservice to this conversation if I tried to break it down into a formula that says what share gains or what's new business. And I'd rather sort of respond a little differently that, this is a large market. We talk about the large -- the top 6, 7 customers. And we are fully engaged with them in a strategic manner, where we're developing not just the release of the next generation, but the 1 beyond that. But this also is end market with a long tail, and we're participating in the mid-market and the small size as well. So we're still very, very early in how this market is going to roll out and what our positioning is going to be. And I think that we're as well positioned as anybody to take advantage of the market opportunity but this is a long and very big story.

For the company -- I'm thinking about beyond AI, beyond Enterprise Data centers. And Wei, I'm not retiring, I'm staying 10, 15 years ahead.

That's great. Can continue the outperformance. And MPS outlined a gross margin target of 55% to 60%. Could you provide an update with regards to which end markets are currently above or below the corporate average or outline some of the specific gross margin drivers?

We're in the range, but on the low side, I noticed that.

Okay. Let me add a little bit of color there. So Michael said this earlier, when we look at all of the opportunities, we keep in mind what is the corporate model for gross margin, which is between 55% and 60%. And I've been fairly consistent over the course of the last 4 to 6 quarters when we've been trending it between 55.5% and 55.8%, which as Michael said, is the low end of our model. that in order for us to show improvement, we really need to have a little longer time horizon as far as backlog to be able to manage in it. So we are starting to see a backlog developing, which I don't want to make too much out of with just 1 quarter's of experience but we should be able to resume at some time during the year. The cadence that we've historically shown of incremental sequential improvements at maybe 10 to 20 basis points quarter-over-quarter.

Our next question is from Jack Egan of Charter Research.

Great. I'll echo the congratulations for Bernie and Rob. I had a bit of a technical one. So during last year's Investor Day, you mentioned a packaging innovation that would allow you to basically double the current density of your modules to about 3-amp per millimeter squared. I was just curious, are there any updates on that? Like is that still a work-in-progress, or is there kind of a time line for that milestone?

We started sampling those products. And again, we expect to have a shipping in this -- thi quarter or next quarter? This quarter, yes. Those are already implemented, already qualified, and our customers went through qualification on it.

Got it. Okay. Great to hear. And then kind of from a higher level then. Last quarter, you talked a bit about the gross margin implications of moving from a silicon supplier to a system provider over the long term I was a bit curious about the impact on OpEx as well. I mean is that going to require you to kind of bring on new teams with experience in systems? Or is there enough overlap between the chip design and system design processes that you can accomplish it organically, I guess, just any details on the impact that R&D dollars or SG&A leverage would be nice.

Well, We need to -- first thing first, okay, and we only gain not lose, okay? I never believes I get big dollars, investments and translates to the bigger gain. Okay, that's a bullsh**. And so look at MPS. We're creating a few thousand, 4,000 to 5,000 products, and I lost tracking. We address the multiple segments of market. While we can put up all these 1 plus 1 is equals 3, and again, not just the 2 anymore. We can pull all these products put in the systems and we can provide higher values to users which doesn't mean we're building a refrigerated, we're building a TV. We're building some things to alleviate our customers' design effort, the manufacturing effort. And that's -- give us a high ASP. And I said earlier, there we're a second highest of selling silicon, and, okay, but why we kind of just put all the silicon together and migrates to system levels like modules like systems. You will see more and more. So at least, I can say the net margins, net profit had to increase, and to be a company we're going to be a lot more efficient.

I want to touch on something that this transformation has been occurring now for well over 10 years. If you think about, we were pretty much completely an analog design house 10, 12 years ago. And then we've been able to migrate where we added both digital engineers and software engineers. And now we've had to take on new responsibilities and new skill sets related to packaging, and as Michael said, in testing. And each time we've done this, we've maintained the same level of R&D efficiency of getting the most out of the dollars spent. So just changing to develop new skill sets around the new opportunities we've identified, does not necessarily mean that it's going to get more expensive or it's going to compress our operating margins.

And I'll just add 1 last comment. Even we made the model during Investor Day, Jack, that was fully aware of this transition. And so we talked about growing OpEx slower than revenue, and giving some leverage to the model. We knew this transition was happening when we put that guidance out there.

Our last question is from Sebastien Naji of William Blair.

I'll just echo the best wishes for you, Bernie on your retirement. My first question is really on the shift to vertical power solutions in the data center. As we move to 2026, what are your expectations for adoption of vertical power? And has that changed at all from your view in previous quarters?

Vertical power. That's a long -- that's everybody is going to vertical power, so I guess, okay. those ones don't, sooner or later they will.

This is just the direction of the market. It's the only energy-efficient solution you can put in place if you're going to operate in these high-voltage -- high-current. So that's just a natural evolution of the marketplace.

Got it. So you think that starts to drive revenue in 2026 then? Is that fair to say?

Yes, yes, yes.

Great. Great. That's helpful. And then maybe just as a follow-up on your optical module business. I think you talked about this a little bit in a previous question. But as we think about the shift to co-packaged optics that's getting a lot more attention these days, how does that potentially change your revenue opportunity in optics. Is it more revenue per port or the ASPs significantly higher? Any thoughts on that?

Higher current high density is always good for us. And we're smart in the increased power, and there's a lot more opportunity for us.

And a higher level of integration.

Yes, high-level integrations. And widen the competition gap.

Yes, I think the only thing I'd add there is, again, I think that's, again, very long term. I don't think that necessarily moves the needle on '26, just to be sure, but it's certainly something we're engaged in over the longer term.

This concludes our Q&A session. I would now like to turn the webinar back over to Tony.

I'd like to thank all of you for joining us today on this conference call. Our first quarter 2026 conference call will likely be held in late April. Thank you, and have a great day.