Q3 2025 Earnings Call Transcript

Welcome, everyone, to the MPS Third Quarter 2025 Earnings Webinar. My name is Arthur Lee, and I will be the moderator for this webinar. Joining me today are Michael Hsing, CEO and Founder of MPS; Bernie Blegen, EVP and CFO, and Tony Balow, Vice President of Finance. Earlier today, along with our earnings announcement, MPS release a written commentary on the results of our operations.

Both documents can be found on our website. Before we begin, I'd 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 Q3 2025 earnings release, our Q3 2025 earnings commentary and in our SEC filings, including our Form 10-K, 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 Bernie Blegen.

Thanks, Arthur. Good afternoon, and welcome to our Q3 2025 earnings call. In Q3, MPS achieved record quarterly revenue of $737.2 million, 10.9% higher than the second quarter of 2025 and 18.9% higher than Q3 of 2024. This performance reflected the ongoing strength of our diversified market strategy, consistent execution, continued innovation and relentless customer focus. Let me call out a few highlights from the third quarter.

Our diversified market strategy drove year-over-year revenue growth in all of our end markets. We continue to expand our automotive customer base with another major Tier 1 supplier adopting MPS for its next-generation ADAS solution. Additionally, we secured our first design win for a full BMS solution on a robotics platform, which further supports our transformation for being a chip-only semiconductor supplier to a full-service silicon-based solutions provider. Overall, we continue to demonstrate our ability to grow and swiftly adapt to all aspects -- all aspects of our business to the fluid geopolitical and macroeconomic environment.

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 up for questions.

[Operator Instructions] Our first question is from Josh Buchalter of Cowen.

Congrats on another beat and raise. I guess to start, maybe you're guiding up about 1%. Can you give us the puts and takes of which end markets you expect to grow more or less? And of note, I think earlier, you mentioned enterprise data was expected to be flat to down 20%. Any updates to the guidance for that segment in particular?

Sure, Josh. When you look at Q3, I think that we saw a little bit better than anticipated performance in both our enterprise data and industrial markets. We had pretty much every other group as we anticipated. Looking ahead, and I know that you're interested in enterprise data, we're seeing a layering of additional customers that began this layering effect in Q4 and is providing this good momentum as we look ahead into the early part of next year.

Well, I should add all these in script readout. Bernie mentioned a list of our market segment, they started growing. And we look back in the last few quarters, all these growth come from greenfield products that were released 2, 3 years ago. And now we see the result. And so in the near future in the next few quarters, we will see these will continue to enhance our revenues.

Maybe a follow-up on that, Michael. I think you've been pretty clear, like philosophically, you have some reservations and even frustrations with the AI market because of the concentration and visibility. Given all the massive announcements over the last quarter even in that space, maybe you could spend a couple of minutes talking to us about just big picture philosophically, how you're approaching these huge forecasts. And I'm sure competitive sockets with massive scale.

How do you guys figure out which opportunities to go after and service and your thoughts on the market just given -- I think you've had frustrations about it distracting from your diversified growth.

Yes. As you said, yes, it is kind of a distraction, okay? But business is business, good money is good money, okay? We don't want to take bad money, okay?

And -- but overall, the bottom line is MPS want to demonstrate in any segment of the market, we are the best, we have possessed the best technology and best customer service. And we're solving problems, we demonstrated that we can have qualities and shipments, okay, all these categories, we are the best company. And in terms of which AI company, we don't really care. We engage with the large companies and the small companies.

And we want to demonstrate that this is the best technology. When revenue comes, it comes. And given times, as we said it, given time, every -- all the true color will show.

Our next question is from Ross Seymore of Deutsche Bank.

I guess for my first question, the automotive side of things. You talked about getting an incremental design win in the ADAS side of things. Can -- there's a lot of choppiness in that end market across different geos and at different points in time, cyclically, geopolitically, all those sorts of things. But from a secular growth perspective, can you just talk about ADAS as a percentage of your revenues now versus user interface or USB and those sorts of things?

And how ADAS penetrating more of that market changes the growth rate, the content, the diversity of it. Just kind of want to capture that ADAS theme and what it really means to you.

Sure. Why don't I take a start at this. So I think that we've demonstrated a history of establishing a strong presence in various markets with differentiated technology. You might recall going back always with USB ports for automotive.

And now it's ADAS. And what this does is it has a cascading effect where we're able to get adoptions and the design wins, and we call out when these begin to ramp. And that's actually been more important to us then the -- necessarily the -- what the SAAR is for a particular end market. And in this one, in particular, we have started with a lot of ADAS opportunities, particularly in the EVs because they're faster to come to market.

But what that's given us the opportunity has been is to showcase all of our other technologies and now we're starting to see those ramp, whether it's in body electronics or different applications. What we're more excited about is as we look ahead and the transformation of the end market for automotive as it moves into 48-volt and zonal electronics.

Yes, we don't know the breakdown, okay, to answer your question exactly. I mean maybe Bernie has some ideas, okay, but I think it's less than half. And I think it's well less than half.

It's considered less than half.

Yes, yes, yes. Okay. And I'm just looking at the number of cars and also our revenues, and it cannot be more than half. But in the ADAS side, more and more cars and including combustion engine cars, they are adopting ADAS. We will see a significant growth in the next few years. That's where we anticipated, okay.

I guess as my follow-up, pivoting to another thing you talked about in your press release of moving from a silicon -- chip-based supplier to more of a solution provider, how do you think about the gross margin implications of that over time? You guys were a few points higher than you are now a couple of years ago. We've talked about what does it take to get you back to kind of the upper 50s from the mid-50s-ish where you are today. Does that system approach help?

Or is that actually a headwind in the gross margin as you go forward as much as it might even be operating margin accretive?

I don't so. I don't think there will be a headwind and a lot of the large systems that we're building, okay, we're kind of learning how to do it. And it is creating a lot of issues once the volume goes up. And so we are learning.

And I think as things get a lot better since last year, I guess, okay? I mean -- and it will improve quite a bit. We actually making our own test equipment as lot of people know about it, okay. We eat our own dog food and creating all fully automated test systems.

Those type of products have never existed before. And I think ultimately, we'll improve the yield and improve the gross margins.

Our next question is from Joe Quatrochi of Wells Fargo.

Maybe first, I just wanted to ask, embedded within the 4Q guide, is there any help you can kind of provide in just thinking about the end markets? I Think there's some seasonality to maybe like a consumer in the fourth quarter. And then I think enterprise data you guys were previously thinking that would be up somewhere like high single digit sequentially in the December quarter. Is that still the case?

Okay, to anticipating a market, that's a very difficult call. And you guys are betting on which stocks, okay? I mean, it's difficult, okay. But we have our way of operating our business.

We do the best. We do the best develop a technology, engage our customers closely and probably the same way that you pick -- how you pick the stocks, okay? And so you engage customers closely. Whatever happens happens.

And you can't predict what the market is. And we don't do that actually.

And if I could just add to that, that since we last talked about the second half of this year, nothing has fundamentally changed in our positioning.

Okay. That's helpful. Maybe just also following up, but I think in the press release or your prepared remarks, you had a comment around the first design win for a full BMS solution for robotics platform. Can you talk about what drove that and how you think about the revenue opportunity ramping there and more wins in the future?

We probably get too excited to talk about those. And the fact is that we got excited is because we see robotics happening. They kind of -- we kind of predicted it the BMS is going to happen, okay. And we got in and we're designing and it's our customers engage with us, okay, there's a ground-up system that we developed.

And we see more and more this type of system will happen. So now this kind of system will be become a reality. So that's the reason we put out that.

Yes. And I'll just add on very specifically on that one. Clearly, we called it out because of the fact it was the first opportunity that we have the design win on. In terms of a revenue ramp, that's really starting in 2026 and it's not in and of itself necessarily a needle move around the model. But I do think it starts the wave of these full solution design wins that we might have going forward.

Yes, that's kind of -- it's the same time when you're making the point that is, okay, we do start to do robotics stuff for actuators and the IC for actuators and the BMS charging, wireless charging. And these are back a few years ago. We even don't know that the robotics were taking off, okay? I've been talking about robotics since 2017 or '18. And -- but the AI assist robot more and more believes, that will really start taking off.

And so that's kind of the projects that we think -- we pick the winners, okay? And we're glad to see it, that's why we're probably too excited to talk about this.

Our next question is from Quinn Bolton of Needham.

Congratulations, Michael, Bernie and Tony. I just wanted to start with a big picture question, just looking through this earnings season. Intel has talked about sort of shortages of server CPUs, we've seen hyperscalers significantly increasing CapEx. NVIDIA talked about $0.5 trillion of demand in '25, '26. I guess my question is, about a year ago, I think you guys were seeing very, very short lead times in that business and dealing with some level of pricing pressure.

I'm wondering as you look into second half and more importantly into next year, have you started to see any change in customer lead times? Are they giving you better forecasts across the enterprise data segment? And is the pricing on voltage regulators and vertical power, has that changed at all over the last quarter or so?

Quinn, thanks for the question. Let me start that this remains a very dynamic market. We're responding to a variety of requests when it comes to the orders and the expectations from our customers. So in some ways, we're getting improved predictability because we're adding -- we're layering, as I said, more customers into the mix.

But as far as the market itself, and particularly with all the blockbuster announcements that have been coming out recently, you can see how quickly things are changing. And what our position is, is that we can't control our customers necessarily, but we can position the company to be as responsive as we can.

And is this sort of dynamic market? I mean if folks are scrambling sort of to get capacity. Has that had any lifting effect on pricing? And then I'll ask my second question.

Any market segment started from the -- and ramp rapidly at the beginning always cause these imbalance, supply imbalance, okay? So in the AI side, clearly is like that. I mean once it goes on, things will smooth out.

Yes. And then being specific to your question, I don't think we've seen any recent or sustainable trends in pricing one way or the other.

Okay. Perfect. And then a second question for you Bernie. Gross margins have been sort of on a -- sort of ticking down over the last year or 2. Can you give us any sense, do you think they sort of stay in this mid 55%, 55.5% range.

Is there some point next year that you start to see gross margins starting to move higher, either driven by mix or new products? Or should we be thinking about margins being fairly flat over the next year or 2?

Sure. As I've commented on prior calls, we've seen about 3 or 4 quarters in a row, where I'd say that we have seen a strong uptick in demand, and that continues even today. What makes this cycle different than ones that we've experienced in the past is that the orders are more short term in nature. We're not seeing a large buildup in backlog in future quarters.

And so without that visibility, it limits our capacity to be able to manage the mix of business that we want to be able to have expansion in gross margins. So for the foreseeable future, until the demand profile changes to elongate the buildup of backlog, I believe that we're going to be in sort of the steady range, plus or minus 20, 30 basis points in the mid-55%.

We have a lot of products, okay, a few thousand products, okay? And a few 40,000, 50,000 customers in to move -- and it's very stable margins. And with the transitions to more solution provided companies, okay, as I said earlier, and all this has to be automated, okay? And as time goes on the margin will improve and -- but not quickly.

And that math is very big. And so we know it's operated on the low end of our margin profiles, but the longer term will improve and we stick with a gross margin range.

Our next question is from Tore Svanberg of Stifel.

Yes. Michael, Bernie, Tony, congrats on another record quarter. By the way, some of that dog food you're referring to must be pretty proprietary stuff. But my first question is on the Enterprise Data segment.

So that's about an $800 million business right now. And my understanding is you're on a journey here, right, and you're still selling predominantly chips. You are obviously moving into module subsystems, eventually systems. So I'm not looking for any numbers per se, but could you just sort of let us know where we are in that journey.

I mean building an $800 million business with chips and where could we eventually go here with, obviously, more and more subsystem type solutions for enterprise data.

If I understood your question correctly, okay, that I can answer that way. We're anticipating doing millions of millions, multiple millions units per month type of a shipment. And all of these integrated, highly integrated modules never existed before. So how do we test this thing, how we achieve the single -- low single-digit PPM failures, okay?

That is we never encounter that kind of issues before, okay. And so we started using our own robotic systems to make that happen. So now we achieved very high volume, 100% automated, including reliability test. And that's our features.

And the goal is building multiple million units a month for that, 10 million a month for that. That's the near term goal.

Yes, that's very helpful. And did you also have a response to my question on the enterprise data. Again, where are we in this journey towards delivering more system-level solutions, especially talking about rack level power and so on and so forth?

This is at the very beginning, maybe I think it's module power still less than, way less than 1/3, okay. And less than 1/3 of our revenue and it grew in the last half year. And the -- we expected it to grow. Some got delayed and now started happening in the next years could be much more.

Maybe just to add on to the back of that, Tore, I think as you start talking about some of these solutions for 800 volts as we discussed, those are like '27, '28 revenue ramps. So I think it supports what Michael was saying that we're still at the front end of this opportunity in the data center for us.

No, I think in the investment communities, okay, it's -- if you have 800 volts technology data center transformer is like a flip a switch. The light is turned on, it turns on. It's not like that, okay. The light to turn on, it takes a couple of years, okay, more than a couple of years to make -- to see the revenues, okay. These take 3, 4 years.

Our next question is from Rick Schafer of Oppenheimer.

And I'll add my congratulations to you guys. Just maybe if I could start with an auto question. Kind of a follow-up, but I know we talked about BMS robotics. But I know BMS becomes a bigger contributor for your auto segment next year.

And you mentioned a couple of things, Michael, in an earlier question, but I'm just curious if you could give some guide rails or provide some guide rails about potential content trends for MPS as you start ramping some of those BMS opportunities? I mean again, not asking for dollar content per car, but does it double or triple those kinds of numbers? Like what does it do to your potential content per vehicle moving into that BMS space in a more meaningful way? And as part of your answer, I'm curious, where's some of the lowest-hanging fruit is for you guys?

I mean, is it 48-volt, is it power isolation, that kind of thing?

It's actually all of them, okay. BMS revenues for auto and for EVs is a bit far away, okay? And -- but our customer, it's a very -- this is a very concentrated market. There's a few players and a few, well it's not a few car makers and gradually, all of them, they have -- they want BMS.

And so our customers are glad to see MPS is to develop that series, that type of a product, too. And in terms of other low-hanging fruits, okay, we see 48-volts as a trend. And we develop those products back a few years ago, like 5, 6 years ago, and we provide all these integrated solutions rather than these discrete ones, okay? And the size can be 6, 7x smaller.

And the other one is the 800 volts for EV. And now all goes up from 400 volts to 800 volts, okay? And in the China market, a lot of cars already have 800 volts. So our silicon carbide solutions came in not for traction inverters and for control systems.

These products will be shined. And so that's -- I pull out my hair at this moment.

And Rick, maybe just to shape you a little bit on timing there to make sure. I think what we said is the layering of opportunities in auto really sort of out of the end of this year and next year starts with design wins that we have, bringing new content to market per vehicle. You start to see zonal designs hit market next year then ramping through into '27. And then the BMS and traction inverter solutions they're really kind of more like '27 and beyond, just to make sure you understand sort of how those revenue opportunities are layering in.

And if I could ask my follow-up just it's on HVDC. And I appreciate the timing and commentary you provided a second ago, Tony. But I'm also curious, I mean, I'm just trying to figure out the right way to think about that emerging market. I mean, like can you give a sense of how HVDA compares to sort of how you've described at your Analyst Day earlier this year.

Maybe how you describe the 48-volt accelerator power opportunities or in any terms you want, just to try to give a sense of what that market represents to you guys or what you think it could.

I think the 48-volt system is clear, there's a reason why the 48 volts, is going back to those telecom times, okay? And telecom systems, 48 volts plus/minus 48 volts? And plus/minus 45 volts. And first, it started with -- in the server side, we're talking about for years this thing.

And it has to be the solutions because once your current goes up, everybody remembers the car using a 6 volt batteries and then it became 12, is ultimately moving up to 48 volts. I mean these are all for control systems, 48 volts and the data center is really happening 48 volts. I see -- that's why I predict all the building -- the building automations are going to be on 48 volts. And the building will be a DC power solutions.

And the opportunity is great. And as we put out to engage our customers with the building automation systems, and we proved the point, actually, it's so welcome for that type of a product. And so that's my view at this time.

And rick, I think part of your question was also on the 800-volt high voltage DC for data center as well, right?

Yes.

And I think on that one, we've been pretty careful about trying to go size the opportunity because, one, it's very far out. Two, we don't know how it will ramp in the market. I think what we have said is since we don't play in that part of the market today, the business we get is sort of all accretive to our overall SAM going forward. But I think we want to be careful about size in the market yet, given how far out it is and not knowing how it will layer into the data centers going forward.

Our next question is from Gary Mobley of Loop Capital.

Let me extend my congratulations on the continued strong growth and continued execution. I appreciate the fact that you still only have about 3 to 4 months of visibility given the capacity that you can support and the quick turns business you can support. But can you confirm whether bookings continue to improve sequentially and what are the seasonal considerations as we look out into the first quarter?

It's again very difficult for us, okay, to predict that, okay, what's the booking, what's the -- where are the bookings. We build our inventories. We try to build, okay, we look at the inventory now, and we try to build up way below our models. And whatever comes we anticipate it, again, we can swiftly to adopt.

Just to add to that is that we really don't have a lot of visibility into the first half of next year. We can definitely point to the normal drivers as far as both enterprise data and automotive are very well positioned for new revenue ramps, but getting both the timing as well as getting that to balance out, we don't have a strong view on Q1 yet.

Okay. Appreciate that. And if I'm not mistaken, your distribution inventory as of midyear was at the low end of your 5- to 8-week target range and it decreased in the June quarter. What was the trend sequentially for the September quarter? And when might you take that distribution inventory back up to maybe the mid- to upper part of that normal range?

Yes. Currently, the Q3 channel inventory was unchanged in terms of days from where it was in the prior quarter. So we take from that, that we're satisfying real demand at this point, which again is a reflection of the quick turns business that we're working with.

Our next question is from Chris Caso of Wolfe Research.

Yes. The first question is on enterprise data. And what are sort of the puts and takes as you look into next year? And of course, this year, there was some changes in market share in that, which affected that business. But I guess I'm going to assume that things are cleaner as you go from this year into next year? And I mean, one, do you expect to grow that business as you go into next year?

Well, it's cleaner. You said that this year, the next year, this year, we are doing pretty good this year. And see, we -- and as I said earlier, the module business is growing. So all this area MPS technology shines.

And the power -- the higher power, the better it is. And because we provide the highest density, power density products that fits this market perfectly. And in the next couple of years, you will see it. MPS is a major player in this market segment.

And also the market is big. We -- it's not -- okay, we don't want a place, so we don't have to be MPS only, okay? We want to have multiple competitors. It's good for the industry.

Yes, I could see enterprise data growing in the range of 30% to 40% in 2026 for us. Much of that, though, would be back in the second half of the year. So while we've seen a number of new players that have been layered in, I think the material ramps are more weighted to the second half of '26.

That's very helpful. If I could follow on to that since you provided a little bit of color on that, Bernie. When you look at that 30% to 40% growth, is that -- because I know that some of the vertical power of designs, for example, you have more content. What's the driver of that?

Is it fairly broad-based? Is it skewed towards some of the ASIC solutions more towards vertical power, whatever kind of color you can give behind that 30% to 40% expectation?

As a CEO, I don't know how to make a 30% to 40% cost. I don't know. And the opportunity is there, okay? If we didn't deliver 30% to 40%, the stocks I see from $900 to $400, what kind of f***** is that? And so I don't want to make them very hard, just waiting for the numbers, let the numbers show it.

Our next question is from Kelsey Chia of Citi Research.

So my question is on the competitive landscape. And I was hoping if you could share more, especially with regards to material side of things like gallium nitride, silicon carbide, I think your biggest enterprise data customer has been signing a lot of partnerships with all these semiconductor companies and I was just wondering, MPS positioning in those. And if you actually see those materials as being important in the next generation of the power modules and chips.

We do our own silicon carbide and we're building the modules. And also, we are seeing -- we are using again -- but that's a very, very early stage we are evaluating it. And also, don't forget about silicon, silicon power MOSFETs have evolved, we engaged a lot of new developments. And a lot of the data showed it can be very cost effective and also can compete with the silicon carbide.

That's very new -- that's a very, very recent development.

Yes. Okay. Got it. And I would just like to have a sense of how do you guys feel today versus a quarter ago.

Especially, you guys have come a long way since the start of the year when you're dealing with all these market share changes, visibility on the ASIC customers and things like that. And given the slew of announcements from all these big mega partnerships, how do you see that relative to opportunities? And also given the maturity of the supply chain, I believe, like things are probably -- the supply chain partners are getting to a good cadence. So how do you guys feel with regard to those recent announcements relative to your opportunity set?

I don't measure quarter by quarters, I measure by multiple years. So I can't tell you that.

I guess the simplest way is we're very broadly indexed across not just the merchant vendors or large ASICs, but medium and small time opportunities. And all of these need to find their way into the marketplace. And it's right now, we're still very, very early in the process. So as Michael said, it's very hard to sort out in any particular time period.

But I think that we're as well indexed amongst all the opportunities as anybody in this market.

Our last question is from Jack Egan of Charter Equity Research.

I have one on enterprise data and then one on modules more broadly. So the shift to modules in vertical power delivery with the custom ASIC ramp should be a pretty big tailwind for MPS. I'm kind of wondering about what the main drivers have been at least so far for those customers that are switching from lateral to module to vertical power. So I'm not really sure if you have this level of granularity, but among the major benefits like higher power density, higher efficiency, smaller footprint on the top side of the board, et cetera, is there any one characteristic that's kind of being cited by your customers as the main reason that they are moving to those modules or vertical power delivery?

Well, we don't see from a chip to module. Whoever stays with the module stays with module, starts with the module. Whoever stays with the chip, stays with the chip, okay? And so MPS provides both, okay, in both chip solutions and module solutions at this time. And so I don't know if it answered that question for you.

Got it. Okay. And then just kind of on the modules more broadly. I think I believe, if I understood it correctly, last quarter, you mentioned is that modules outside enterprise data could be like 10% to 15% of your total revenues.

And so I was curious how much of your revenue base or I guess, addressable market outside enterprise data would be eligible for switching to modules. I mean, even if you're looking several years into the future, how high could that mix of modules outside enterprise data go?

That's a good question, okay. And we want to -- we build those modules and again, very similar to enterprise modules, okay? And since 2017 industrial markets adoption is kind of slow, actually faster than telecom, okay. And these are 2 market segments that we focus on.

And then to our surprise the auto industry also want to use it because it's easy to implement. And so -- they don't want the semi equipment. And that's a large segment, we didn't realize that, okay? I mean now we see all these revenues are happening there.

So I think that in the next couple of years, it will be growing faster than 3 or 4 years ago. And so we're picking up a business -- the rate of increase is picking up.

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

I'd like to thank you for all joining us in this conference call. I look forward to talking to you again during our fourth quarter 2025 conference call, which will likely be held in early February. Thank you, and have a nice day.