Monolithic Power Systems, Inc. (MPWR) Q2 2025 Earnings Report, Transcript and Summary

Welcome, everyone, to the MPS Second 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 released a written commentary on the results of our operations. Those 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 Q2 2025 earnings release, our Q2 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.

Good afternoon, and welcome to our Q2 2025 earnings call. In Q2, MPS achieved record quarterly revenue of $664.6 million, 4.2% higher than the first quarter of 2025 and 31.0% higher than Q2 2024. This performance reflected the ongoing strength of our diversified market strategy, consistent execution, continued innovation and strong customer focus. Let me call out a few highlights from the second quarter. We continue to see diversified revenue growth across all of our end markets. We began initial shipments of our power solutions to support our customers' new ASIC-based AI products. Storage and compute revenue grew sequentially off a strong Q1 as we continue to see demand for both memory and notebook power solutions. MPS continues to focus on innovation and solving our customers' most challenging problems. We continue to invest in new technology, expand into new markets and to diversify our end market application and global supply chain. This will allow us to capture future growth opportunities, maintain the supply chain stability and swiftly adapt to market changes as they occur. Our proven long-term growth strategy remains intact as we continue our transformation from being a chip-only semiconductor supplier to a full- service silicon-based solutions provider. I'll now open the webinar up to questions.

[Operator Instructions] Our first question is from Tore Svanberg from Stifel.

Congratulations on another record quarter. Michael or Bernie, I was hoping you could talk a little bit more about the September quarter, what the setup is there? You're guiding for 8% sequential growth at the midpoint. I was just hoping you could give us some puts and takes of the six end markets in the September quarter.

Sure. Happy to, Tore. When we look at Q3, we've got enterprise data growing between 20 and 30 percent sequentially. We also see a seasonal uplift in consumer. And then with the exception of storage and compute, all of our other lines of business are up high single digits. In storage and compute, we just have a little bit of caution, primarily because you're coming off of 2 very strong quarters in Q1 and Q2.

And just as my follow-up, you mentioned the ASIC program is now starting to ramp. I was hoping you could add a bit more color there. Are we talking about multiple customers? Are these primarily vertical power architectures? And I guess, a really important question is back at the Analyst Day, you gave us that $4 billion SAM for your enterprise data market. And there's a lot that's happened since then. So I was just wondering if that number is starting to move quite a bit upward.

You're right. Since the Analyst Day, things are changing fast. And everything is good. And after a couple of years, these enterprise data segments and clearly, you establish a winner or losers, okay? And that MPS has appeared to be a winner. We do engage multiple customers. If it's not all large customers, all large customers or potential customers that come in. We have a lot of design wins and design activities. While that threatens our ramping up in the near terms, also a lot of small emerging players. And we see the peripheral but not only in the data centers, but all kind of applications. And that's what we're very excited in the long term. The short-term ones, and I said in near terms, it doesn't mean then, okay, next 6 months, next 12 months, okay, I keep saying them, okay, our forecast revenues have been -- is always a plus or minus 6 to 12 months. So you said earlier the $4 billion, okay? That's what we said and that's what we see it, and we're going to get there.

Our next question is from Chris Caso of Wolfe Research.

I guess the first question is with regard to enterprise data. Previously, you had provided some guidance on that to be flat, plus or minus 20%. You talked about guidance for the September quarter. Any more visibility with regard to the full year guidance for that? Any more narrowing of that range? And whatever kind of color you can provide on your expectations there?

Sure, Chris. The market, as Michael just said, remains dynamic. We have fairly short lead times. So even for Q4, we don't believe we have our arms around all the business, it's likely to occur. So that's just said, the nature of the dynamic of this fast-moving market. So at Q2, we identified the range as being flat to down potentially 20%. And while we're not guiding on Q4, in addition to the growing Q3 sequentially by 20% to 30%, I can say that Q4 will be up sequentially.

Well, whatever we said in the beginning of the year, we feel comfortable.

That's helpful. Just in general, you've obviously listened to the calls from some of your peers that there's some degree of macro uncertainty out there. Some of the customers -- some of your competitors rather have expressed some caution and some concern about some tailwinds in certain areas. I wonder if you could comment on that with respect to your business. And in general, as compared to 90 days ago, is there anything that's changed in your view of the overall markets or your expectations with respect to the year?

Well, I'll give you an arrogant answer. Sorry, but I don't listen to any other costs. I'll tell you that, maybe I've known from you, okay? And we focus ourselves as always. Market condition is market conditions, okay? We provide the components to the multiple segments. That's where we focus on it. And we focus on the internal execution and execution with our customers demanding in futures. And that's what we always do. And whatever happens, happens as long as we're much better than everybody else.

If I could add to that, in Q2, we used the phrase that we were cautiously optimistic about the outlook for the balance of the year. And I think that still describes how we feel from the standpoint that we have seen a broad-based continued strong demand profile in all of our end markets. However, the ordering pattern because there is a little bit different risk pattern to remain with short lead times, so as a result, we're not necessarily building backlog that we have visibility out beyond 2 quarters. That's a little bit different from most recoveries that we've experienced. But again, I want to stress that we feel very good about our overall positioning for the remainder of the year.

Our next question is from Quinn Bolton of Needham.

Congratulations on another good quarter, Michael, Bernie, Tony. I guess, Michael or Bernie, just wanted to ask, as you start to ship into some of these ASIC platforms, can you give us a sense, the ASIC platforms, do they tend to be sourced by multiple PMIC suppliers? Or do you tend to see sole source sockets for a given generation of an ASIC, and then the ASIC vendor may source to first generation with yourself and, say, a second generation with a competitor. That's a question I've gotten a fair amount. And so just wondering if you could give us some sense on whether those ASIC programs tend to be single sourced or multiple sourced.

Yes. We see a variety of single sourced and double sourced, multiple sourced, high-cost source and low-cost source, and all source. And we deliver what our customer -- we develop system vertical power, again, and which is more module like solutions and even the chip side, and we do whatever our customers demand. Maybe they came and say that, okay, 2 ways. I disappointed that you didn't ask any more specific technical questions.

I'll say those for Analyst Day or maybe they call back.

Okay. Let me add real quickly there, is that each of the end customer has their own reasons for how they're selecting their suppliers. Some of them is supply chain resilience, others want innovation. Again, like every opportunity we have, we provide strong customer focus and consistent execution. So that's what makes us feel that we're very well positioned across all of these opportunities.

When I tell you the high cost and low cost, multiple customer, one source, two source, multiple source, they all choose, nothing but they choose.

Okay. Got it. And then the second question, I'll move to the automotive end market. Kind of wondering if you guys could give us your outlook for the second half of the year. What are the biggest drivers of growth? I think you have a couple of platforms with Western OEMs set to ramp where you have some pretty good content. Wondering to the extent that those ramp, does that drive growth half over half in the automotive business? Or are you looking for sort of more of a flattish half over half in that segment, in the second half?

So automotive, and we've been very consistent on how we've described the rollout for calendar '25 that we enjoyed a nice step up sequentially from Q4 to Q1. We anticipated that, that would be flattening a little bit in the middle part of the year and then picking up end of the Q3, Q4 as these new content opportunities come online. So while there is some back and forth on the SAAR and units and in particular, with individual companies, we're less affected by that than the timing of these new content ramps.

And Quinn, I think I just wanted to add on that, right? I know we're hyper focused just on the year. But I think if we step back and look long term into 2026, the opportunities around 48-volts, some of the zonal architectures, I think they continue to be opportunity for us going forward, and this will be a growth area for us over the long term.

There's too many things I can't remember.

Our next question is from Ross Seymore of Deutsche Bank.

Congrats on the quarter and guide. Just want to dive first into the enterprise data side. You mentioned in your preamble or the press release that both the AI side and the server side were strong. Can you talk a little bit about any differences between those 2 growth rates, composition, kind of the breakdown of ED between those in both 2Q and 3Q?

Again, something that we've talked about as it relates to enterprise data is that the lines between traditional CPU and AI are getting a little blurry. So it's very hard to make clear statements of relative growth or importance. Having said that, I think that the overall profile both for the near-term or midterm and the long term remains very positive.

Great. I guess as my follow-up. There's been a decent number of concerns about pull-ins and tariff-related activity. Obviously, you haven't mentioned anything on that. But outside of the enterprise data segment, when we think about the cyclical recovery that's happening, are you seeing any evidence of that kind of tariff influenced behavior? And/or do you think the cycle itself is really what's driving demand?

We believe the cycle is driving demand. We really don't have enough information to support a change in our customers' ordering pattern that would be related to tariffs.

We don't want to potentially know that, yes. These are our controls. And again, I mean, whatever happens, happens. By the way, our inventory is low, Ross.

I didn't want to go there, but it was nice to see.

Our next question is from Rick Schafer of Oppenheimer.

Congrats on another nice quarter. You guys make it look easy. I wanted to ask a quick follow-up on the $4 billion enterprise data, SAM number. I was curious if that considers the eventual conversion of server CPU to 48-volt. Does that factor that in? Or is that incremental to that number? And second part of that question, I'm curious how much does HVDC increase that SAM or that TAM? And when do you expect direct current rack power to start really taking off? I think you started sampling last quarter, if I'm correct.

Okay. These are the 800-volt systems and 400-volt systems. And yes, we've started sampling. These are not in the factors. And both what you mentioned, the 48-volt servers and also the 48-volt systems and 800-volt system, we're this far in the future, so maybe far in the future is maybe a couple of years, a year, 18 to 24 months kind of things or maybe even longer, okay? We don't want to call the market, okay? We are the only solution providers. And this one, we believe, ultimately, all data center will convert into this type of 48-volt and 800-volt systems. And that's what we're targeting, but that's what we emerge and focus our development, not only in the last couple of years. We said many years ago, back even in 2016, we foresee 48 volts, that will be the solution. And we became one of the key supplier in that. And last couple of years, 3 years ago, we started working on the 800-volt systems. And also not only that, also the BMS, the battery management systems, and these are absolutely fit for that type of applications. And not only for vehicles, energy storage and the data centers, this is all about the energy utilizations.

Rick, the only thing I'd add, I know you asked very specific enterprise data question, but remember, we kind of think about the overall data center opportunity. And whether that's optical module growth, whether that's going to be memory, all those things, I think, play in as opportunities for us. So I'm just trying to get a step back a bit from only focusing on the enterprise data segment.

I appreciate that, Tony. And that actually leads me to my second question, which is I know it's not your largest segment, but communication seems to be firing on all cylinders. I mean, satellite, WiFi, 5G and transceiver power that you just mentioned. I didn't know if you guys can elaborate at all or talk at all about order trends, order velocities there, outlook for that segment? Like just basically any color you'd be able to share there.

Sure. So if you look at about a year ago, we saw a large step-up from Q2 '24 to Q2 of '23. And a lot of that was in the core networking telecom business. And that's sort of plateau -- that element has plateaued. But at the same time, we saw growth in the optical modules within the data center that's been growing very nicely. So right now, I think that we're positioned very well but I don't necessarily have a strong signal of additional investment in the network category.

Our next question is from Josh Buchalter of Cowen.

Congrats on the record results as well. Might shock you, but I'm also going to ask about enterprise data. As we get into the back half of this year, any metrics or guidelines you can give us on how much this new AI ASIC is contributing to the back half of the year, how it compares to your lead GPU customer? I mean is this opportunity comparable in size to what you've been able to generate on the GPU side?

Well, all these questions that are being asked are being similar questions, so we'll answer okay? We're looking in the future, even near-term future looking good. But that's only about, what, 25% of MPS business. And the bigger revenue growth is the rest of the company. And I hope we should have more questions on the rest of the business.

I will take the subtle hint there and ask you about auto.

Auto and H1 that add together maybe only 40%. How about something else?

All right. I mean I got the storage and compute there. We can take this anywhere.

All right. I'm just joking. Ask us whatever you want to ask.

Let's take with storage and compute then. I mean you mentioned some caution there into the back half of the year. Is that sort of an inventory dynamic? And you had gained a bunch of share, I believe, on DDR5 to start the year. Is there still more room to run from that on the share and content side on the DDR side within storage and compute?

Josh, it's an excellent question because we had a very significant step-up of our position competitively as well as from revenue in both storage and notebook. So again, the reason that I used the term cautious is because both end markets tend to be a different demand profile from like automotive, for example. And what I mean by that is historically, notebooks have always been like consumer and been expansion in Q3. We had such an atypical seasonality with a buildup in Q1 and Q2 that it just pays to be a little bit cautious there. Likewise, on memory, I have nothing to indicate that there's a slowdown or a change in the market positioning. But again, it's just that in the past, they've had historic boom and bust cycles. So that's the only reason that I'm offering. Now having said that, we were pleasantly surprised in Q2 that the results for that particular group came in better than expected.

Yes. In auto, we grew significantly this year, right?

Yes. Our full year results are going to be well above what we've been doing...

40%?

Yes. Probably, for the full year, we'll be between 40% to 50% growth for the year.

And that's the reason Bernie probably asked to be even more cautious, okay? Not going to be 100% next quarter, okay? That's what we mean relatively what cautious means, okay? Cautious is not expected another 50% or higher.

Our next question is from Gary Mobley of Loop Capital.

Bernie, I appreciate the fact that you don't have a lot of visibility out into the fourth quarter, but I want to ask about the seasonality of the fourth quarter. Typically, Q4 might be down, what, mid-single-digit percent sequentially? How do you see it shaping up this year?

I don't have a seasonality anymore.

I think Michael said it all there. Again, if you look at that historic trend, and I don't know the last time we actually fulfilled being down, it's in a fairly narrow range. So I think flattish is probably the easiest way to describe the outlook.

Helpful. All right. So it sounds like you've got plenty of capacity, plenty of inventory. Can you remind us what sort of annual revenue you could support with your internal and external capacity? And can you confirm whether or not the book-to-bill ratio is in fact trending above parity?

It's 2 separate questions, but I'll try to address pretty quickly.

I'll answer first, the second part. Our inventory is low.

Our current capacity, and we've talked about this in the past, is to be able to support $4 billion of revenue with diversification of 50% of that outside of China. So what we're trying to do is be able to support all of our customers' requirements in whatever supply chain profile they're looking for. When you look at the book-to-bill ratio, and I commented on this earlier, that we're having sort of an atypical ordering pattern when you consider that we do believe we're in the middle of a cyclical recovery that's very broad-based. And what I mean by that is the ordering patterns are much more short term. We're not building book-to-bill ratios of like 1.4, 1.5, where we'd have backlog continuing out into Q1 and Q2 of next year. It's really a more near-term focus. So with those short lead times, that's the only reason I have a little bit of concern about Q4, and I don't want to send a negative signal, it's just that that's the nature of the demand profile.

That's it. I don't want to say negative single either for the low inventories. We are expanding our supply chain. And we can meet it in Q4 in our customer demand. And for next year, we start to -- even now, we continuously qualify the newer supply, and whatever it takes to meet the customer demand. That's what we always do.

The only thing I'd add, I don't know if it was part of your question was, in addition to the overall capacity, the geographical balance of it. And what we said is we would, by the end of the year, have half of that capacity outside of China, half of it inside. And to Michael's point, we just want to be able to and believe we can meet customer demand no matter how they want to route their product.

Our next question is from Kelsey Chia of Citi Research.

Michael, Bernie, congrats on the strong results. So I have a question on customer concentration. So it's great to hear that you guys are shipping to the ASIC platforms. So does it mean that MPS sort of back to the historical kind of diversified growth where there's no one customer that's more than 5% of your sales by the end of the year? Or is the ASIC ramp sort of lumpy as well that can sort of tilt that kind of customer concentration?

I think that when we had the high customer concentration, particularly in enterprise data, that was an aberration from our normal model of being broadly diversified in terms of customers, end markets and geography. So I think now that the portfolio of market entrants is starting to build up, and we're going to have exposure to all of those opportunities. You'll see us go back to a more normal profile of customers not contributing more than mid-high single digits.

Got it. My second question is on the growth rate. So it seems that the analog industry has sort of been going through a downturn in the last 2 years. And potentially for 2026, we could see pretty strong growth due to the cyclical recovery. And you guys have a 10% to 15% outperformance target versus peers. So that would imply sort of like a close to a 20% growth rate perhaps for next year. Is that a right assumption? And if you can provide some color as to which end markets would be driving majority of that growth based on the content or design wins?

Sure. I think that your rule of thumb as far as our traditional outperformance and also what the broader market looks like for '26 are both accurate. So I think within plus/minus a couple of percentage points, I can support those numbers. Again, as far as the particular end market drivers for next year, we believe it's going to be broad-based, although with all of the enterprise data opportunities ramping next year. That will probably be a key contributor.

Our next question is from William Stein of Truist.

First, I wanted to clarify about the short lead times and ordering patterns. Is it fair to say that the only thing that's really going to cause that to stabilize and lengthen is your extending the lead times that you quote to customers, which likewise is sort of difficult as long as revenue is fairly meaningfully below your capacity level? Is that a fair way to think about it?

Yes. I know it's a correct characterization that is meaningfully below our capacities. I don't know if that's an accurate statement or not. But overall, it's a fast changing market and customers are updating their models, okay? We have just to keep it up.

Think we're being responsive to real demand. One thing we haven't touched on is that our channel inventories in each of the geos, major geos for us are down in the quarter. So they're also very lean. So right now, we believe we're meeting real customer demand.

Got it. By the way, Michael, what I meant was comparing the revenue guidance -- revenue results and guidance relative to a $4 billion level of capacity, there's a gap there, right? So that's all I want. It wasn't a critical one.

Yes, $4 billion, okay, and it's not only for -- I mean it's for enterprise, okay? We're building capacity towards it, it's the case. That's the process.

Got it. And the other thing I wanted to ask about was to comment on the product development and revenue trajectory in 3 areas that you've highlighted in the past as sort of unique growth opportunities. One is modules, the other is converters, D to A and A to D converters. I think you hired a team a couple of years ago. We haven't heard that much about it. And the other is eMotion, which I know is ramped, but I wonder how meaningful that's become relative to your overall sales.

Thank you very much. First thing, the e-commerce is a kind of flop, right? I answered your question, that part of your question a few quarters ago, okay? But the good news is, the module business is really growing other than in the enterprise data, and industrial side, even consumer side. And we offer those solutions that our customer doesn't want to get into the detailed design. And we provide a solution for them. And these are revenues that next year is about 10% to 15% of our total revenues, okay, other than the enterprise data power modules. And actually this is very much related to when we provide assisting solutions. And so we're transforming companies, as Bernie said earlier, to be system providers, a solution providers, and that's what our customer wants. If they want to chips also, we provide chips. And at the same time, they help MPS revenue growth. We're not depending on only selling chips. A few years ago, I'm talking about I'm tired of selling chips only, okay. And that's where our revenue grows, okay? And the other things that you're talking about the data converter. The data converter is kind of slow moving. And we are releasing a standard product for that. And for only some $1 billion, $2 billion revenues, maybe contributes very little and that doesn't -- will not move the needle. But as a product, in the product categories, that provides a total solution. That's a part of our picture. That really benefits the top line growth in terms of solutions. eMotion, finally, we get the needle moving. And it's been for a while, we get over about $100-some million in the past few years. And they're not too bad. It is slower than the MPS total growth. But now robotics. And we see it, AI-driven robotics will be -- okay, we see a lot of opportunities, and a lot of potential to grow in the next couple of years. And we provide total AI power, and not only AI power, we provide all the actuators, actuator solutions and motion controls and as well as a barrier, as BMS solutions. These all combined, all together, the motion will grow a lot faster than in the past few years.

If there are any follow-up questions, please raise your hand. There are no further questions. I'd like to just say a few closing comments. I'd like to thank you for all joining us this conference call. I look forward to talking to you again during the third quarter 2025 conference call, which will likely be held in late October. Thank you. Have a nice day.