Silicon Laboratories Inc. (SLAB) Earnings Call Transcript & Summary

Welcome to day 1 of Citi Global TMT Conference. My name is Atif Malik. I cover U.S. semiconductors, semiconductor equipment and networking equipment stocks here at Citi. It's my pleasure to welcome Matt Johnson, President and CEO, Silicon Labs as well as Mitch Haws, sorry, Thomas Haws, I'm confusing, Mitch is at Skyworks and now he is at AMD. The format of our discussion is a fireside chat. I will go with my questions first and then open it up to your questions. If you have a question, please raise your hand, and we'll get the mic to you. Welcome, Matt.

Thank you.

Matt, I want to rewind what has happened this year. Your revenues troughed last year in December quarter, and you've seen some sequential recovery in first half going into September quarter. Are you seeing clear signs of demand recovery and path to getting back to the $200 million plus sales per quarter? And what are the kind of the goalposts you're looking for the signs of recovery?

Sure. Quick answer is yes. All the data points coming in are encouraging and going in the right direction, whether you're looking at the bookings trends, customer expectations, what we're actually billing, pulling requests, it is a consistent improvement in the right direction. Not sufficient to say, hey, this thing is over, and it's a recovery, a full recovery, but they're moving in the right direction. I think it's the rate that's been so difficult to call. That being said, back in Q4, when we were at our trough, we said we see a path to sequential growth moving forward, that has not changed. We basically have -- you can break it into 3 big drivers of that growth. The first is the inventory -- excess inventory destocking at our end customers. That's been working down in terms of the count of customers carrying too much and the average days of excess. So that's encouraging to see, that going in the right direction. That's been the primary driver of our revenue growth. As revenue gets closer to true consumption, that's helped us in the first half. We've shared even in Q3, our guidance is below consumption. So that gives you a sense of its improving, not done, but going in the right direction. Next big driver is design win ramps. We have been winning design wins at a strong pace over the last few years. In fact, we've had design win growth throughout this entire cycle. And we are now starting to see those ramp, which is very encouraging. Not just 1 customer, multiple customers, multiple geos. So it doesn't step function when that happens, they feather in, but that's going to be the next growth driver for us as the destocking becomes less of a factor and design win ramps will be more of a factor in us driving sequential growth. The third leg is the end market, much more difficult to call, much less clear. Our customers are getting a lot of conflicting signals, and we're not banking on that to drive growth. It will happen at some point. It will come back, and that will be awesome. But we're not banking on that as we look to the next few quarters.

Matt, just that topic, are you seeing this recovery typical of past cyclical kind of corrections? Or are there things that are different about this setup?

I think the asynchronicity of buy-in market has been interesting, but I think the biggest factor is the inventory. We'll do case studies as an industry on this on the other side of the cycle. But I don't think we've ever seen this level of inventory buildup and depletion required in any cycle in recent history. So I think that's definitely been different and kind of amplified the peaks and troughs, at least for us.

Okay. And I thought it was interesting you mentioned that the next kind of goalpost is the design win activity and ramp. And can you just update us in terms of your design win funnel. You guys had a very strong momentum on Series 2 and even Series 3 going into the upturn and -- so what is the typical kind of a time frame for the design win to translate into revenues?

Sure. On average, once we identify an opportunity, it could take a year or 2 to convert that to a design win that we know will ramp at some point. And then it could take another 1 year or 2 to ramp. And then it could take 4 to 5 years once it ramps to reach its peak, and then it works its way back down. So every customer market, everything is different, but on average, that's what we've seen. The only caveat to all of that is during this cycle, we've seen a push out in terms of the win to ramp, which has been a function of many things in terms of not having all the right pieces from a supply perspective, delaying to burn down inventory, all these different things have pushed out design win ramps. But as I said, we're starting to see those now, which is very encouraging.

Okay. Then just on the product side, can you talk about early customer feedback on Series 3? How does Series 3 differentiate from Series 2, as they're both in the field? And also touch on the AI aspect of these products, how is that innovation panning out?

Sure. So for anyone who's not familiar, our current generation is maybe poorly named Series 2, which is actually our fourth generation of wireless technology. That is kind of our platform across multiple products, and it's our workhorse in terms of opportunities, design wins and revenue growth for years to come. The reason it's successful, it supports all the key ecosystems, protocols with industry-leading security, industry-leading battery life, and it's what's driving those design wins I've talked about in that future growth. So very happy with Series 2. And for those of you who are unaware Series 2 even includes integrated AI/ML inference engines, so you can do battery-powered inference at the edge in our products, and that's production released can used by our customers today. So we were doing this stuff before the market even had a big need for it because that's our strategy that our products lead with the capability and then the market moves into that. Same with security, same with ML, et cetera. Series 3, which is our fifth generation, which we've just started sampling will be a significant step function on the elements that we compete on. Wireless performance, battery life, security, compute. We've shared 40 to 100x more compute than prior generation. So do not think of it as a replacement for the existing Series 2. Think of it as an addition in terms of performance, because some of our customers need 1 level of performance, some will need the new level of performance. And people get confused, because they see it, like, how could that be what an IoT application needs, it will move into it over time. That's what happens in our space. If you create that capability, they will fill and take advantage of it over time. That's exactly what's happened with Series 2. That's what will happen with Series 3. But do not think of it as a replacement. Think of it as a higher performance, broadening our capability as a company, and Series 2 will drive revenue for growth for another 6 or 7 years. As Series 3 starts to ramp during that time frame as well. Anything you'd add, Thomas, I might have...

Yes. Can you provide some examples on the use cases of Series 3 where it's getting more pull or end markets?

So again, it's early days. We've shared publicly that Series 2, maybe 12-plus products in ICs and then software, hundreds and hundreds. Series 3 will be 30-plus products. And we're just sampling our first one. So this is the way it works. It's early days. But let me try to pick -- let's take an example, light bulb right? A light bulb today with Series 2 could have wireless connectivity. Obviously, you can turn things on, turn them off, you can control. You could have really strong security. You could have very strong capability from a differentiation if you wanted to have a lot of control and features. Series 3, you will be able to do new things from a compute perspective that weren't possible in that application because you'll have a lot more processing. You'll be able to do new things from an AI/ML perspective, because there'll be a lot more inference power at the edge of that application that's available. You'll have next level of security because during this time frame, you're going to need post-quantum capability for products. So all of that capability will be there as well. And that's why it's not a replacement for because using that example, there'll be tons of light bulbs that will need 1 set of capability, but there'll be a bunch of new light bulbs and applications who want that next generation capability with the compute flexibility, the next level of security, new machine learning applications. So both can coexist at the same time, and it's important that people understand that because there's a tendency to think of it as very digital that, okay, once the new 1 comes out the old 1 goes away, it's not what happens at all. We're still selling a ton of Series 1.

And then recently, you disclosed a significant win in continuous glucose monitoring or CGM, can you talk about the overall size of this market and other design wins you have referring to this space?

Sure. So if anyone is not familiar with CGM's, continuous glucose monitors would be the patch that you wear. This is a market that we were not shipping anything into a couple of years ago. We shared back in Q4 on a public call that Dexcom was going to be ramping us as a supplier. We've also shared since then, we've added more than a dozen other CGM wins and customers since then. And we're starting to see those ramp. So what's interesting about that space is it has a lot of volume potential, one, because the market adoption is relatively low. If you look at the number of people that would be eligible. And it also drives a lot of volume because these are disposable. They're only used for a couple of weeks and then you need a new device applied. So that combination of early days in the market and disposable creates a lot of volume. And it's interesting to us in the healthcare space in general, that's why we added that focus 4 or 5 years ago because we thought that there'd be a lot of need and applications there, and that's what we're seeing play out, which is really exciting. So think of it as a space where no revenue a year ago, and starting to ramp revenue this year. It will be a growth engine in the next few years moving forward and definitely has the potential to be a 10% plus contributor to revenue over time.

Any kind of update on electronic shelf labeling, smart metering end markets that -- what's the health of those markets?

Overall, solid and good is a quick answer. What we've shared for anyone not familiar, smart metering is a strong market for us. We have a leadership position in gas, water and electric. I think on the last earnings call, we shared that globally. We are participating in every major smart metering rollout period. That's not true, no period, with the exception of China domestic. China domestic, we don't have access to that market, but China export we do, which is exciting. Now that market is not that dynamic. It's pretty stable, steady growth. But that strong position and the need for those types of applications, that's an easy strong 5- to 7-year growth engine for us, which is really exciting to see. ESL, electronic shelf labels, for anyone not familiar, these are the digital labels that would go replace a static label of years past. The benefit is dynamic pricing. You don't need someone to actually update the price. It can be done electronically, digitally, remotely. That market, we also have a leadership position in. We shared on our last earnings call that we've shipped over 300 million units over the last few years, lifetime to date, and we see that ramping. The best proof point to that is most retail environments you walk into, you don't see these devices broadly deployed. But most major retailers have shared publicly that they plan to, and that's really exciting to see. And the reason is the maturity of the technology is now there, the reliability of robustness and the returns are really good for the retailers. So that combination is driving rapid adoption, and we're very well positioned to index and capture that opportunity.

Great. On Bluetooth, it has been a success story for Silicon Labs. Can you give us a sense of what the contribution to your current revenue is today from Bluetooth?

So yes, easiest way to think about it for anyone not familiar. When we talk about Bluetooth revenue, we're talking about something that is only Bluetooth. We ship a lot more Bluetooth integrated into other products that could be integrated into a subgig part, a Wi-Fi part, a 15.4 part. But we don't count that as Bluetooth. We count that as Wi-Fi, 15.4 et cetera. For Bluetooth, easy way to think about it, our largest positions by wireless technology would be 15.4, which is Zigbee and Thread and then sub gigahertz, and they're about the same in size. The fastest growing for us is Bluetooth and it will approach those other 2 areas in the next 1 year or 2. And that's based on share gains that we've been able to secure over the last few years. And by the way, a majority of those share gains have not presented yet, because they're won in design, but they haven't ramped yet. So that's to come in the next few years. So very happy with our progress in Bluetooth, very happy with what we've seen there, and we expect that to continue to grow for years to come. And then the third -- sorry, fourth one is Wi-Fi, earlier days, but the products that we released are getting a very good reception, have set some records in terms of opportunity for anything we've ever done and we are now securing designs and expect that to ramp next year. So we'll see Wi-Fi starting to contribute more to our revenue in 2025.

Great. Now just going back to the cycle, are you seeing any differences in destocking or restocking in the 2 segments, Industrial & Commercial versus Home & Life?

You want to take that one ?

Yes. Atif, I think what we've seen is that our Home & Life business kind of entered the inventory cycle, maybe a couple of quarters before the Industrial & Commercial. And so what we've seen is it's starting to come out on the other side a bit earlier. That said, in the June quarter, our Home & Life business was up 39% sequentially. And our Industrial & Commercial business was up 35% sequentially. So relatively similar sequential growth. That said, we did say on our Q2 earnings call that we would expect the Home & Life business to grow a little bit quicker sequentially, but still seeing positive signs in Industrial & Commercial as well.

Okay. And the question we get a lot on pricing. And correct me if I'm wrong, I think earlier this year, you talked about some pricing pressure for one of the larger analog company in the U.S. But how have the pricing trends kind of progressed through the year with the rate?

No big changes. Standard for what you'd see in a downturn, people trying to use price to win business, doesn't work that easily and quickly. It's a long-term game. We are very comfortable with our position. We haven't changed our gross margin model or outlook, haven't lost any share. And we see a path to navigating this. It's pretty typical to pass downturns in that regard that people try to use price, but you got to have all the features, performance, capability platform. Otherwise, a customer takes the price and they have to derate their performance and their application as well. So we feel good about that. But I think in general, the environment is what you'd expect for a downturn.

And what are you seeing in China in terms of your recovery and where you see your kind of normal exposure to China over the long period? And are you seeing any substitution on IoT parts in China from domestic China suppliers?

Yes. Atif, I think if you went back a couple of years, our China as a percentage of revenue for domestic consumption in China was maybe 20%, 25%. And where we're at now is low teens as a percentage of revenue in terms of China domestic consumption. We haven't really seen very positive signs coming out of China. And just in terms of the economic recovery. We're not banking on that to kind of underwrite our growth into 2025. But we are opportunistic there. We continue to see design win traction there. We will win there when we're offering a superior technology, but there obviously is a preference for China domestic in terms of Chinese consumption.

Let me pause here to see if any there's questions in the audience.

The earlier comments you made about the 10% of revenues. I mean we're talking about CGMs. Was that CGM specific or health care as a whole?

That was CGM specific. Health care as a whole will be larger than that and a pretty decent growth engine for us moving forward because CGM is great and awesome, but there's a lot of other applications in the life and health care space as well that we like, and we see good progress.

Yes. And if you all want to ask, I can just repeat it, if you don't need the mic. So no problem.

Just curious on the ESL, I'm just curious about like what's the current penetration rate right now? Like how much is the adoption right now as a retailer sequential growth...

Like globally?

Yes.

It's a good question. I would guess single-digit percent for sure. Is it mid or low single digit? I don't know. It's tough to call because there's a lot of debate on where that total market is in terms of potential. But I'd say definitely a single-digit domain. So we have easy 5 to 7 years of growth in front of us in that space as retailers adopt and start to deploy that technology. And again, the easiest proof point is most retail environments you walk around, it's pretty light in terms of deployment, but the returns are really good, and most retailers publicly have shared, they want to go for it. So I think we're early days.

How much do you think it's going to grow?

What's that?

How fast do you think it's going to grow?

Well, tough to call again, but what we've shared is we've shipped over 300 million units into that space over the last few years. And I think that was -- I'm not trying to overset expectations here, but I think that was early days in the market. And now we're seeing more retailers commit, more retailers prototype and start to deploy. That will drive those numbers up. And we're very well positioned in terms of market coverage, technology and our position there. So it should accelerate sometime in the next few years, just difficult to call exactly when. But again, we're early days, and there's a lot of potential there. So I do expect an acceleration.

Matt, just looking at the model, even at -- going back to $200 million a quarter, the company is still not -- I'd assume the profitability level that you'd think is acceptable...

No.

And you're not going to shrink your expenses to get there. So what do you think is the revenue level of the company is sized for right now? And how long will it take us to get there?

Yes, sure. So for everyone that's -- what's our plan around OpEx, profitability? How do we see that playing out. So simply said, we're definitely sized in OpEx to be a larger revenue company than we are today. We know we are a larger company in terms of consumption and in terms of the design wins that we've secured. So our plan is to, one, hold our OpEx flat until we get to profitability as fast as possible, which we're getting close to that. Revenue is going up, gross margins will go up. That helps. Don't increase OpEx. And then after we get back to profitability, the goal is to get back to model profitability as fast as possible. And for that, we'll be holding OpEx flat, again with the exception of absorbing merit and bonus, not paying bonus while not profitable. Once we start getting back there, we'll start paying bonus. But aside from that, you can expect the OpEx to be very measured. So easy way to think of it is during the time when revenue is declining, we reduced our OpEx every quarter during that time. Now that revenue is growing again, we're going to hold it flat with the exception of merit and bonus until we get back to profitability, then get back to model as quick as we can. And the variables that we know destocking, we talked about that, design win ramps, that's going to drive our sequential growth from here. And at some point, the end market will show some more strength, but we can't bank on that. So we're operating on that design win growth until then. So hopefully, that answers the question, but we're not going to let OpEx go more than it needs to until we get there. And we are sized larger. So that allows us to capture the opportunity and not hurt our future growth potential while we're waiting for this to happen.

Can you give us an idea of what's the size [indiscernible].

Well, the easiest way is to look at the model that's publicly out there, right, that we share publicly at this revenue level, this is the profitability we should be at. So I think at like $1 billion, it's a 15%, if I'm not mistaken, 15% op margin. So that's what we're trying to get back to as fast as possible. And like I said, we won't let the OpEx go until we get there. But to be clear, we don't guide beyond that, and it's tough to call because there's uncertainty on the end market. So we don't want to set an expectation that is based on an assumption that's tough to deliver on it. What we can bank on is we know we won on design wins. We see those ramping. That's going to drive our growth. If the end market comes back, it will happen even faster. So that's the best answer I can give you without giving forward guidance. Other questions?

Matt, on this, the model, your gross margin is expected to be mid-50s range in the second half, which we believe is your long-term target model. So how do you see the upside to that range? Is it mostly volume-driven, top-line driven or there are areas within the mix that can perhaps with Series 2, Series 3 being higher gross margin products?

Sure. So the question is kind of do we expect the gross margin outlook to change.

Do you expect to kind of raise your long-term gross margin?

Yes. The quick answer is no. And let me try to explain. During the peak of this cycle, our gross margins were strong 60s and there was a lot of pressure to increase our model. Like this is the new normal. And we shared we think this is temporal. We don't think this is the new normal, and we stuck to our gross margin model high 50s, then in the trough of the cycle, gross margins went down. Mix was not favorable underutilization. We have to absorb a lot of costs. And again, we did not change our gross margin model. We said the model is the same. This is temporary. And as you're seeing now, it's been improving every quarter. I think we're guiding midpoint 55 for Q3, and we expect that to continue to go up as we progress and the mix improve. So the point is, throughout this cycle, peak or trough, we've been consistent and nothing's changed there that we see our ability to drive a premium versus our competition to drive high 50s in this space is unchanged. And if I'm honest about it, I think if there was ever a chance to reset it, it would have been during the cycle, and we didn't do that because we know what we're winning. We know our position, and we feel like we can continue to drive that.

And on the cost side, I remember during the opportunity you guys did a pretty good job in kind of transferring the higher wafer cost for your customers and your foundry partner, TSMC and they're talking about raising pricing for next year as well, maybe not on your geometries, but help us understand the cost part of the gross margins, like is that fully kind of baked into your outlook for second half?

It is. Yes. I think simply said that I think we were able to not get stuck in the middle during the peak or the trough on this. And I think the right expectation long term is there's not a meaningful change in those cost structures. If you look at what our industry is doing, moving manufacturing to higher-cost geos does not lower the cost, it increases them. All the capital needs to be paid for. That doesn't lower the cost. I think running at lower utilization increases the cost doesn't lower them. So I don't think there's -- all these things, materials, utilities, they're all pushing things up, not down. And I think we're in a position to not get stuck in the middle either way. And we've been able to do that so far. And again, if we were going to change it, and we were worried about that, I think this would have been a good opportunity, but we haven't needed to do that. Sticking to our commitment in the model.

And one last one on M&A and divestitures. You guys sold your I&A business to Skyworks a few years back. And are there other opportunities in your portfolio for kind of tuning in terms of improving profitability or -- and I'm sure you've assessed this during this downturn already. But are there other areas where you can improve either the margin profile or decide not to be in those markets?

Yes. The quick answer is no. I mean, we're always looking, always should be managing our portfolio. One of the best things is there's so much opportunity and you can't do everything. So that forces prioritization and decision-making on what's going to drive the best results. But from a divestiture perspective, we invested everything we wanted to when we did the divestiture, and I think that worked out well for both companies. Nothing in that regard moving forward. It would be the other side. Where can we acquire in a way that would accelerate our market share and our model that we're going after. We're not looking for small tuck-ins or technologies. We have all the requisite technologies we need. We've been acquiring companies at a rapid pace of 1 a year, and then we got to a point where we're like, okay, we have what we need. Now it would be how do we go faster and bigger in the markets that we're in? How do we step function and what type of acquisitions would allow that to happen. So that would be of more interest for us than divestiture. But always managing, pruning that portfolio. That's part of the job.

And one last one on AI on the edge. I get asked this question a lot. I feel like we all have been waiting for AI-on-the-edge to happen. And when you look at either smartphones or garage door openers and IoT or other apps, are there like killer applications or markets where you see a very strong case for AI in the industrial side or product side. And we'd love to get your thoughts on where you think AI-on-the-edge will be the first to take off?

Sure. What we believe is we believe AI at the edge absolutely is happening and will happen. We've even put it into our products and our Series 2 products were developed some time ago. We really operate in, for a lack of a better term, the embedded edge and embedded applications. That's more difficult for AI, because you need to have a lot of domain expertise. That's what embedded is. You need to know the application you're doing something kind of specific for that application. So what we're seeing is customers wanting to take advantage of machine learning and inference at the edge and it's happening, but they have to harness and master their data. They need to find ways to do what they do even better because of that, and there's a lag. So I do see signs that's happening, and it will happen, but it will take some time. We're betting on it. We're putting it in our products. And just like we explained earlier, our goal is to always be ahead of the need and the market moves into it, not trying to react to the market in that regard. And then to your last piece, killer apps, not yet, but we do see some green shoots and encouraging signs that there are those killer apps out there for some applications. But remember, we serve thousands of applications and customers. So it's going to be pretty broad. I don't expect just one, but there will be some growth engines in there in the next few years. I really believe and see the signs of that.

Great. On that optimistic note, we can wrap up our fireside chat. Matt and Thomas, thank you for coming to the Citi Conference.

Of course. Thank you. Thank you all. Appreciate it very much.

Thanks all.