Monolithic Power System Inc (MPWR) — Q2 2024 Earnings Call Transcript

Welcome everyone to the MPS Second Quarter 2024 Earnings Webinar. My name is Genevieve Cunningham, 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, VP of Finance. Earlier today, along with our earnings announcement, MPS released a written commentary on the results of our operations. Both of these 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 risk and uncertainty. 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 earnings release, 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'd like to turn the call over to Bernie Blegen.

Thanks, Gen. Good afternoon, and a big welcome to all of you. Let me open by saying MPS reported yet another record quarter with Q2, 2024 revenue of $507.4 million, exceeding the high end of our guidance. Our strong revenue growth was attributed to three factors: increased demand for AI-powered solutions, improving order trends in several of our end markets, and initial revenue ramps associated with design wins secured in past years. Separately, we continue to expand and diversify our global supply chain to ensure supply stability and capture future growth. As we have emphasized for many years, our results reflect the continued success of our proven long-term growth strategy and our transformation from being only a chip supplier to a full solutions provider. I will now open the webinar up for questions. Gen, could you please ask the first caller?

Our first question is from Matt Ramsay of Cowen. Matt, your line is now open.

Thank you very much, everybody. Good afternoon. Michael, Bernie, congratulations on, I'll just say, it's nice to have a calm part in the storm relative to what all of us have been dealing with over the last two weeks, so congrats. I wanted to ask a couple of questions that I imagine the enterprise data segment is going to be a huge focus on this call. So maybe I'll just start with the first question. Bernie, if you could maybe give us some directional color by end market in your guidance for September, I'd appreciate it. And secondly, Michael, if you could talk a little bit about what you're seeing in some of the non-enterprise data segments, like things like consumer, comms, auto. The model's been driven by this big growth in enterprise data, but from what I can tell, your company's had a lot of design win success in some of these other areas where the macro has really been challenging. And I'm wondering if you're starting to see any green shoots there yet, for recovery in some of those other segments? Thanks, guys.

Sure, I'll keep my comments fairly short as far as the Q3 outlook by end market. Essentially, the bookings trend that we carried into Q3 has improved quite a bit over the last few quarters and has been fairly broad as far as end market participation. If you look specifically at enterprise data, you'll see that we took a fairly large step up in each of the last two quarters. And then in the third quarter outlook, we see continued growth in communications, storage, and computing with incremental improvements and also in our other groups.

Yes, your second question, as Bernie said, we do see some of our other business that would not relate to AI. And now it is difficult to separate AI from non-AI. Other businesses we designed in, which we shipped products to about half a year ago and a year before we see the market starting to wake up. But that's whatever it is. And we do see a lot of new designs and new requirements which are much higher power, especially for communications. And, as a matter of fact, across the board. These designs are much higher power than previous versions and these are all related to the new designs associated with AI. These design wins will likely trigger revenue gains in two or three years.

Thank you very much, guys. I appreciate it. I asked two questions, so I'll get back in the queue. Thanks.

Our next question is from Dustin Fowler of Oppenheimer. Dustin, your line is now open.

Hi guys, I just have a couple of quick questions. So for power isolation, I believe you ramp later this year in auto followed by data center next year. I think both markets could be Greenfield opportunities in the hundreds of millions each. Could you just give us a sense of the ramp in both markets, the competitive landscape and maybe any market sizing if you can?

Dustin, I apologize. This is Bernie. I didn't hear which two markets.

Oh, sorry. That was auto and data center for power isolation?

Okay. Yes. We do start to ship in the gate and the product for the higher powers and for data centers. And this is only the beginning. And I would say that, that's very true. Auto is becoming known, and as ADAS technology becomes more integrated into cars, we believe that in two or three years, all cars will feature that technology. And we have design wins with all these car makers, especially in the EV sector.

Okay. And I guess as we think about kind of auto this year, I think the street has auto models flat. I guess how realistic is that given the 70% content gains you have with your largest auto customer before power isolation, plus the wins with Chinese OEMs? I guess maybe for Bernie, are you modeling auto flat this year? And how should we kind of think about share gains?

Well, we cannot pick the models for you, and I can't give you a model. My accuracy is plus or minus 12 months. And going forward, we are very confident.

Yes, I think the near term outlook for automotive remains a little bit fuzzy. For example, in Q2, we were expecting a nice uplift, which did not occur. It’s correct that positioning within the next two or three quarters is hard to predict for auto more broadly and EV specifically, but our long-term positioning is only getting better with additional design wins.

All right, thanks guys.

Our next question is from William Stein of Truist. William, your line is now open.

Great, thanks for taking my question. Michael, I'm hoping you can elaborate on that comment you made about solution selling. I think that might have been the term used. I think in the past, we've been conditioned to hear about modules versus semiconductors. Can you take a minute to explain what sorts of products and what end markets, and what growth you see coming from that activity?

I like that question. Yes, okay. As we said, we started that journey a long time ago. At the beginning, maybe I didn't know, we didn't know what we were talking about, but that's the direction. That includes a part of e-commerce, and we sell plug-and-play solutions. The effort is that the journey will be; we are known to be a semiconductor company. Yes, we do. When we sell a silicon chip, we are, in fact, selling a solution. We know all the technical details in terms of how we implement the solutions. So our goal is to monetize all our knowledge. MPS, in the end, doesn't make anything, but we make money from our knowledge. We're capitalizing on all our capabilities. So today, selling solutions is probably 20% to 25% of total MPS revenue. Every piece of silicon used to sell somewhere between $0.50 and $1. Some of them are even lower for the mass market, and we vary little now, especially in the consumer business. Now we're turning to solution-based company from silicon-only options.

Will, if I could just add two comments quickly. The first is, this is specific to your question, that it's across all of our markets. It primarily focuses on those that have longer design cycles, where we can add expertise in design solutions. The second point is, while we are driving a higher percentage of our business to the modules, or to the more complete solutions, we have the ability, like no other company, to deliver the type of solution our customer wants, whether it's a die, a chip, or some package in between.

Yes, many traditional semiconductor companies find it hard to distinguish whether they're solution providers or selling modules, solutions or semiconductor only. This is especially true for AI power solutions. There are very few companies using a silicon-based module solution to power up AI solutions or data centers.

If I can add a follow-up, please. Can you talk about what you're doing in vertical power delivery? I know many companies are claiming to have it, or be ramping it, and I know you guys are one of those companies. Maybe you can talk about what your efforts are yielding in that area? Thank you.

We are in the leading position. Okay, I said it – there may be one or two companies that have capabilities similar to MPS. The market is huge and growing. Our leading position is not something that happened overnight; we've developed that over the years. Not many semiconductor companies have this type of capability.

Our next question is from Quinn Bolton of Needham. Quinn, your line is now open.

Hi guys, let me offer my congratulations. Maybe just first, Bernie, a quick follow-up on Matt Ramsay's question on the drivers of sequential growth in the third quarter. I think you mentioned communications and storage and computing, but I didn't hear you specifically say enterprise data. I just wanted to confirm enterprise data is still one of the big growth drivers in the third quarter. And then I'll get to the two questions?

On enterprise data, first off, yes, it is expected to grow. However, we have grown significantly each of the four prior quarters to this, so the rate of growth is slowing down. I offer that because there are many opportunities that have yet to ramp with a number of different companies that we're currently working on designs with. But I was calling out in particular that both storage and computing and communications will be good drivers in Q3.

Quinn, you know about VR 13.5, and we start to see significant revenues from it. That was just the beginning. VR 14 hasn't really ramped up yet, and we have a lot of designs increasing it. VR 13.5 is like an MPS test case. During the shortage, many revenues and volumes shifted to MPS because our competitors couldn't deliver, but we did. The benefit is for VR 14, where we will be significant players in the market segment.

Particularly in the second half of this year, it looks like that segment is starting to take off.

Great, that was going to be served. I guess my first question was just could you talk within the enterprise data about some of the opportunities, whether it’s on the CPU power side or just broadening out your AI portfolio? And then I’ll have another question on vertical power if I could squeeze in that second question?

Go ahead.

So the question on vertical power, I mean, just some different competitors may mean different things when they're talking about vertical power. I just wanted to kind of get your guys' view on the market. When you guys talk about vertical power, is that true vertical power where the voltage regulator is sitting underneath the processor substrate and supplying current vertically into the processor? Or do you guys consider sort of stacked packages, where you've got one or more voltage regulators, or phases in a module with the power inductors and capacitors, and other things that might be stacked vertically but still delivering current laterally into the processor? What's your definition of vertical? Because I think there's some confusion in the market of what may be vertical and what might look vertical but actually still supplies current laterally?

Yes, you're right. Some customers use top surface lateral power supplies similar to server CPU power supply. Others are turning to the backside, still using lateral power, and the more advanced one is directly under the CPU. To answer the question, MPS provides all three solutions: chips, all modules; we have everything there.

Quinn, if I could also add to this, the necessity for different deliveries of power becomes increasingly important as you get into higher current. In addition to either lateral or vertical power, we also have liquid cooling solutions that we believe we have a position on.

Yes, Quinn, to answer your question, the most efficient and difficult to achieve is to make a module directly under the CPU. That's the highest efficiency power delivery, but it's also riddled with technical issues. Many companies take a conservative route, while some just use top surface lateral power delivery for those very high powers well over 1000 watts. The power modules should be directly under the CPUs for the highest efficiency.

Got it. Thank you, Michael. Thank you, Bernie.

Our next question is from Chris Caso of Wolfe. Chris, your line is now open.

Chris, can you hear us?

Sorry, I was on mute. You can hear me now?

Yes.

All right, sorry, thank you. The first question is on seasonality, and how that may be changing for you. In the past, the December quarter typically was a seasonally down quarter, but your business mix has changed quite a bit since those days, and the revenue is now coming from segments that are not as seasonally driven. With that, how should we look at December seasonality going forward, or is the old model no longer appropriate as we look forward? And if there's anything you want to comment on with respect to December?

Yes, we're seeking your advice. What is our seasonality? With our new business and inventory oscillation from our customers, we don't know. You could see a huge shortage followed by an even larger oversupply, and now I assume our customers have consumed all their excessive inventory. The market is waking up. Is there another potential shortage? I don't know. It's difficult to forecast and the lead time is still very short.

Okay. So we'll do our best for everything good thing?

Yes, yes. You can tell us.

All right, that's fair enough. As my follow-up question, I wanted to follow-up with vertical power as well. What we know, particularly within the enterprise segment AI, we've got some higher wattage processors coming. Can you talk about what you expect for content on those, and what in particular you expect for vertical power next year? Because I know there's some debate around whether the ecosystem is ready for vertical power, and does that make a difference with regard to your content as you go into the next generation of processors, given that the wattage is certainly going higher?

Yes, I can tell you this. Our customers do not appreciate discussing how many dollars per unit, or GPU unit, okay? They don’t want us to talk about that. And second, I don’t want to explore that.

Okay. Understand, but I mean, just in terms of the opportunity, maybe you could speak in more general terms about the opportunity that’s available to you?

Oh, sure, yes. And now you look at our revenues, right? For a couple of companies, last year was like what, 1% or 2%, or maybe last year’s, okay, couple of percentage of our total revenue? Well, today, it’s like we’re running at 20%. So that’s a full $500 million, okay? We felt this was just the beginning. There are a lot of other companies where our opportunities and design wins haven’t ramped up yet with companies like Google and AWS and Meta. Those are coming later. So I see huge opportunities, and I do see that this is not just MPS, but the market is way too big. We don’t want to turn into an AI company only.

Yes, I think the only thing I’d probably add on top of that besides the direct AI opportunity is the AI trickle-down effect. Whether that can be memory, it can be optical, it can be pulled through networking. I think there are a lot of opportunities for MPS, beyond even the direct AI business?

That’s a good comment, yes.

Our next question is from Tore Svanberg of Stifel. Tore, your line is now open.

Yes, thank you, Michael, Bernie. And congrats again on another stellar quarter. I wanted to follow-up on enterprise data, but I'm going to move away from the processor power management, and ask you more about opportunities elsewhere in the data center, especially power supply. I know historically you haven't talked a whole lot about GAN or silicon carbide, but I'm pretty sure you have some activities there. So can you elaborate on what you're doing on the power supply side of data centers?

Yes, I like that question. Okay, I like that question. But I can't tell you the revenue yet, as it's a small amount. Since 2016, we started to develop our own silicon carbide. We're not selling it as an expense; it will be a part of our solutions. I said very early on, we are not intending to sell just a power device only. Now, we designed those three-kilowatt, six-kilowatt, and twelve-kilowatt power supplies. You guess what? It’s for data centers. We have very little revenue from it, but the beauty of MPS is that we use a huge amount of power supply in our own test plots, making us our own customers for testing equipment. We design and invent how to test these modules particularly, as well as burn testing. MPS supplies these power modules to our vendors, who incorporate them into their systems, becoming our test equipment. This is how we're starting to generate revenue. However, for 4G and for data centers, these are still in the initial stages, as we engaged them a couple of years ago. Revenues are small, but we're looking for big revenues from that segment, as MPS provides the highest efficiency and smallest size.

Yes, that’s great perspective. As my follow-up, I recall, because we've gone through a few cycles together, I remember usually you use downturns to gain a lot of share. I know everyone's excited about data centers today, but what are some areas where you're starting to see more meaningful share gains from a design perspective, now that we're sort of navigating through the bottom of the cycle?

Sure, so I think that we've been very clear about the positioning that we have in automotive, and how that's continuing to grow across a multitude of different platforms with various Tier 1 suppliers and OEMs. The most recent area for improvement regarding Greenfield opportunities is related to communications. We are seeing this in wireless and 5G, and I believe we’ve seen some early indications of continuing investment in that segment. That’s probably the area that will be most visible, but I don't want to forget industrial, which, although it presently constitutes a relatively modest part of our business, does not show a lot of momentum towards growth but holds a lot of Greenfield opportunities that will materialize in the next four quarters.

Yes, go ahead.

Yes, I was just going to ask, and the RV market, is that going to be categorized under automotive or industrial?

Oh yes, no, no. Yes, okay. That's one of my favorite ones. Okay, yes, I think it's going to be positioned in the industrial market. Because all these products share the same signatures as we do for data centers and for our own test equipment. These products share a lot of the same characteristics, especially for robotics that are moving, not stationary. These robots need that type of power supply, charging, and battery management. They share very much the same characteristics as RVs.

Yes, those could fall under industrial, consumer, or even humanoid markets too. So I don't think I want to just categorize those...

I think it's probably on the industrial side.

Sounds good. Thank you and congrats again.

Our next question is from Quinn Bolton of Needham. Quinn, your line is now open.

Hi guys, just wanted to add a quick follow-up question. There have been a few semiconductor companies this earnings season that have mentioned having licenses to Huawei being revoked in the quarter. It raised some questions from investors. I'm just wondering if you guys could address this; do you have a license to ship to Huawei? If you did, has it been revoked? Is there any real exposure at MPS to that customer? It's obviously not impacting your near-term outlook, but I figured I would just ask, as we've received a number of questions on potential Huawei exposure across the semi coverage? Thank you.

Quinn, I can make this pretty simple. We don't have any licensing or contractual arrangements with Huawei at all, so they can't be canceled. Any business we have with them has always been done on a purchase order basis, and we have no indication of a change in that relationship.

So Bernie, I guess there was more of an export license question. Do you need an export license to ship to them? I think that’s what’s raising concerns?

Yes, our product is not, not like in the 20 nanometer or below that. As far as I know, we are not subject to any export limitations.

Yes, you're right.

As far as I know, we're not subjected to export limits, but I will consult our legal counsel to answer your question precisely.

Yes, okay. Okay. Thank you.

As there are no further questions, I would now like to turn the webinar back over to Bernie.

Great. I'd like to thank you all for joining us for this conference call, and look forward to talking to you again during our third quarter conference call, which will likely be in late October. Thank you, have a nice day.