Cirrus Logic, Inc. (CRUS) Earnings Call Transcript & Summary

Great. Thanks, everybody, for joining us today. I'm Rick Schafer, Oppenheimer's semiconductor analyst, and I'm joined by Cirrus CEO, John Forsyth. I think most people probably know John joined over a decade ago, I believe, through the Wolfson acquisition, as I recall. Been the CEO since 2021. And again, time fly. So great to see you, John, and really appreciate you able to make today.

Yes, yes. You too, Rick. And you're pretty much exactly on the money. The Cirrus acquisition of Wolfson closed in August of 2014. So we're talking a decade -- yes, this month.

Yes. Decades. So maybe I'll just kick things off. I could kind of a kind of high level, I guess. I mean, Cirrus has been in every iPhone, I think since the beginning. Your largest customer, I think you count at 90% of revenue and -- if you assume handsets of mature market, which I know admittedly is a big assumption. But if you assume that, that's sort of a mature market. I mean, can you walk us through sort of your primary strategy for growth as center on generation content gains there at your top customer because, obviously, you guys have been a practice and, very adept at that over time. And maybe as part of your answer, I don't know if -- maybe you want to give, but maybe if you could walk us through a little bit for tease a little bit about so where the bigger opportunities are going forward for further content gains.

Yes, absolutely. Yes. Thank you. So to be clear, we aim to grow both within our large and outside of our largest customer. So as we do the former, that obviously means we're going to keep growing outside of the customer -- outside of our largest customer, if we want to kind alone that concentration. And yes, to your point, in the quarter we just reported, I think I'd like to discuss where it's 88% revenue that obviously fluctuates over the course of the year. But we're getting into a kind of big build period now as well. So yes, it's a chunky number that's right up there. But that collaboration and that relationship with that largest customer has been enormously successful over a period that's more than -- it's approaching 2 decades. It's more than 1.5 decade at this point where we began with one audio socket, and we've obviously branched out in many ways from there. So, as I said, we aim to grow both within that customer and outside of that customer and our strategy for growth is really based around 3 principles. Number one is maintaining our leadership in the smartphone audio space in those flagship devices. That's kind of the golden goose. It still represents the big single component of our revenues. And we have significant investment that's kind of coming to fruition this year in those areas with some new content we're expecting -- new content that's going to be in devices that we're expecting to see launch in the back half of the year. The second of those principles is to branch out within -- staying within the smartphone and locus, but branch out in content terms beyond audio into areas that I've referred to as high-performance mixed signal, which encompasses a number of different areas, camera, haptics and power being the main ones that we've talked about. And then the third element of our growth strategy is around taking those technologies and IP that we've developed both in the audio space and the high-performance mixed signal space and leveraging those to address other markets. And that's something that we are increasingly doing in the PC space, for example, and I'm sure we'll get on to that where we've been seeing some success that we're excited about. So yes, in our largest customer, the chief kind of growth opportunities for us and the way we think about our growth there is both iterating on content and sockets that we have today in ways that enhance the features and performance. And looking at new areas where we believe we can be a differentiated innovative supplier where they would potentially be, I think, what you call a general content gain like a whole new socket for us. And we've shown our ability to do that. As I said, in the HPMS space, having added over the past few years. We added the camera and then power. We've obviously been shipping haptics for a little longer, but I think we've established now a lot of the credibility that we can solve hard mix signal problems in areas beyond the audio space and be in a really good innovation partner for that. So we have multiple kind of irons in the fire from that perspective. And maybe just to kind of wrap on this point. As I mentioned, we're particularly excited that this year, we have new content coming in the audio domain, and that's both the audio architecture from our perspective, from a silicon perspective, really consists of a comp things. It's a codec and boosted amplifiers. Boosted amplifies driving the loud speakers you hear from the phone and the codec during a lot of processing of both the audio output and input is, what's coming from the microphones and so on. And we have new generation versions of both of those components, ramping right now for a device launch that we anticipate in the coming months. And that represents really -- a really significant evolution of the devices that we've been seeing for a while now. We do expect those to -- typically that it will ship for quite a while. So the current generation today will have been shipping for 6 years when they get updated. So from that point of view, we also get the benefit of a lot of leverage from the R&D investment that we make there, and we can apply those resources on to other tricky mixed-signal problems where maybe we see content expansion opportunity with our customer.

Okay. Thanks for all that detail, I really appreciate it. A question I've had and certainly we've gotten from the as well. But just I'm curious, you guys have posted some pretty remarkable upside the last few quarters. And I don't know if you could just give a little bit of color on your forecasting process or -- just trying to understand better sort of how some of those surprises came about? Again, it's several quarters in a row and again, kudos for the significant upside, but just trying to sort of understand the process there a little bit better?

Yes. Thank you. Okay. I mean kudos, but also -- and I appreciate the nice comments, but to be clear, it's not what we're aiming for, like if we're going to surprise people, I definitely rather we were surprising on the upside, obviously, but -- our goal when we guide is to give the -- our best estimate, the most accurate estimate we can give of where we're going to land for the quarter. it's not to give -- have a kind of head fit where we give people a range and then plan to surprise them on the up, that wouldn't help us long run, you know what happens, which is people would bake in that assumption. So I don't believe that helps investors or at all. So what I'm saying, we've just been -- we've just been really bad at accurately forecasting. So I do want to at least explain a little bit about the environment that has contributed to that. So -- and I guess the other point I should make there is there hasn't been a change in our forecasting philosophy. We don't -- as I said, our goal is accuracy. And the kind of factors that we take into account when we're doing our forecasting is obviously the kind of order book from the customer. We have, as you would expect, our own extensive networks in the supply chain, we talk daily with the contract manufacturers. We have our own information on what their inventory position looks like. And we also look a lot at historical patterns and the historical trends within any given quarter. And so typically, that process mostly directs you to the right place. Over the past kind of 2 to 3 quarters. And I guess maybe as we just reported it, the June quarter is a really good example to talk about. We normally see in a typical year, much more movement in the demand picture from the customer than we have this year from our customers than we did this year. And it can be movement in both directions, but the typical pattern at this point when you're in the June quarter is that you've got past the kind of holiday period, then you get past Chinese New Year and so on. And you normally see some tapering of demand, some modulation of demand, some rescheduling and moving around. And when I say typically, like we look back many years at the equivalent quarter and how things shook out and how the backlog kind of got rescheduled and so on. And there really isn't any previous year, which looks like this year, where we didn't see that tapering of demand at all. We didn't see that modulation, and that's been a feature actually of this cycle, which has been really sustained, really healthy sustained demand all the way along since the launch really to atypical degree. I -- and then -- so that's what we've seen, and that's been to us coming in higher than we anticipated in the June quarter and previously. I would say that if I were then speculating about contributing factors to that. I do think one of the stories in this cycle has been very proactive use of incentives on the retail -- in the retail environment. I think there are lots of science that have been -- those have been leveraged towards very effective demand creation. And because those have been driving the demand, there's been obviously some modulation of those incentives in order to keep demand and supply really well balanced relative to previous years. So I think a lot of people have done an amazing job to deliver that, but that's how it's kind of translated into the picture that we see.

All right. Got it. Thanks for all of that. I pivoted, if I could, to PC, just because it's kind of an exciting relatively new opportunity for you guys. And PC power converter, obviously, greenfield opportunity for Cirrus. I was hoping you could speak just for a second about where the PC Cirrus Power is winning? And maybe you can talk about who you're winning against primarily you're beating in that space. And what's the strategy there within PC because it's kind of a new channel for you -- correct me if I'm wrong, but it seems like a relatively new channel from you that you've got to sort of green shoot grow. And so -- is it the same kind of idea that if you go back in time, 10, 15 years, you did a handset where you sort of led with audio in this case, do you lead with sort of power converter, get your foot in the door, land and expand. You know what I mean, and then you start branching out where you're selling codecs and you're selling all your apps or whatever, the rest of the lineup into PC over time. So again, as you grow that new channel?

Yes, it's an interesting story. So number one, yes, it's mostly greenfield for us. We're starting from, yes, more or less 0 is the base. So that's always exciting to have that opportunity where we don't feel especially exposed to the -- what's happening in the PC market as a whole. I mean, to some degree, yes, but it's all upside to us, tree be told. We're actually going in parallel with different components. So I'll break out the components that we have in the PC space. We have power conversion to the point you are making, and we have audio, which is made up of these 2 classes of components again codec and boosted amplifiers. And then we have haptic driver as well. And we actually got our first initial kind of footholds into some of our PC vendor partners in the audio space. To begin this is pre the Lunelake reference design. And really, it was during that period of the silicon semiconductor crunch where we -- some people were left hanging. They didn't have audio parts. And we had some off-the-shelf parts, which were not ideal. They're not perfect, but we could supply them. And at that point, that was -- supply was Kings. So that certainly opens some doors to us, which might otherwise have been -- we might have been knocking on them for quite a long time otherwise. But we also knew that there was something important having in the audio space, which was a transition of the audio architecture and the primary interface used to connect audio components, which is going to transition away from a legacy interface called HDA. And so we saw -- any time you see that kind of architectural discontinuity or a shift we saw an opportunity to enter the market with some fit purpose products that supported the new interface, new architecture, new driver architecture on the Windows side. And we developed a made for PC-specific codec and made for PC-specific boosted amplifiers. So we've seen -- and we saw a really, really good kind of positive responsiveness to that within our customer base. In parallel to that, on the power side, we developed a PC-specific power converter. And this is what you are alluding to. This is leveraging IP that came from the Lion acquisition we made a few years ago, and really provides more efficient DC to DC power conversion within the laptop space. So less waste heat effectively relative to conventional book boost power converter from somebody like TI, for example, supply a lot of those products. It turns out that the key for getting wins on the power side has very much been the Lunalik reference design. So we've partnered with Intel and Microsoft on that. And that's been driving really good uptake, both on the audio side and with the power converter. And I think as a general rule, it does seem like OEMs, no DMs aren't crazy about on picking the power architecture of a reference design. Of course, they swap some components in and out. But we have something that's relatively unique there compared to the competition, and we're seeing really good design momentum around that. So we've -- yes, so the -- that spread of products that we've developed and delivered. They're really winning in parallel at this point, and we're kind of excited about that the collective opportunity that, that represents for us.

And speaking of that, I mean, I know I think I've heard you say a few times that you hope in the next few years, PC could rival sort of Android in terms of contribution to your top line? And I don't know if you want to expand on that at all, just only because as you said, relatively greenfield market for Cirrus, it's at least a 250 million unit TAM. I mean it's a massive market potential, right, to be sort of growing into and especially you start thinking about this lineup of potential SKUs that you can be selling into a PC. If I don't know if you could just touch on why you don't think that it could be bigger than Android for you guys given the relative size of Android in your model right now versus ILS? And then I was just curious if you could just kind of go out there on either talk about the PC TAM or talk about what your actual versus potential content per PC could be something like that?

Yes. Great. Actually, great questions. Let me talk about comparison with Android first and then talk about how we see the market opportunity for us. On the comparison with Android, a great point. Yes, I think it could exceed our Android business. I actually think on the Android business, I mean, I've mentioned that elsewhere, like we -- we love the business we have there. We love the customers. We don't put a lot of R&D into that for multiple reasons. But one of them is a lot of that business is wrapped up in China-based smartphone OEM. They're great partners and customers of ours, but the fact is there's a geopolitical cloud over that, which nobody can ignore. They need -- they never know what export control change they're going to wake up and read about. They need domestic suppliers for any silicon they're buying from somebody from the U.S., I believe. And that's just going to limit the opportunity and probably mean there's a lot of cultivation of local -- lower-cost competitors in the long run. So for us, that's something we regard opportunitilly not bullish about it long run, and we definitely don't build our strategy around it. So I think the PC business, yes, we should certainly get level with Android and then there's really no reason why it can't exceed that, it should for us, especially given that range of content that I talked about. So when we look at old market opportunity, yes, so 250 million units. We are not going to be relevant in a reasonable chunk of that. I think the way that we think about it, very broad strokes is similar to the smartphone world, which is the higher tier kind of premium tier, we should be all over that. And then when you look at the mid-tier, there's going to be a good chunk of that where we're highly relevant as well. And then once you get more into stuff that is really built to a budget and where they're making a lot of compromises on user experience or physical sales or whatever, then we're going to be less relevant. But we believe there's a lot of value to capture in that kind of upper half of the market. To the tune of -- it's $1 billion plus SAM in 4 years from now. That's the estimate that we have, and we're certainly looking at ways that we can continue to expand that where we're relevant within the entire TAM. And that's going to be distributed across those different components that I talked about. We see a lot of factors that are driving demand. So in audio, in the U.S., again, you asked we are winning guests. I mean the typical audio supplier for PC World is Realtec. What we saw coming out of COVID was a desire almost universally across PC OEMs for a better audio experience as part of a better audiovisual experience because so many people are having so many more conversations through the screen. That demand and that -- the importance of that remains the case today. It's really not a marginal benefit as well when you compare an A, B. In fact, we did this many times with OEMs of wiring out or replacing on a design that they have replacing the existing components with our amplifiers and just blow them away every time. So we can provide a hugely better experience and help our customers do that, and that's great. On haptics, PC World has been very slow to catch up. Where I think the non-PC laptop manufacturer of no chip haptics, Trackpad about 10 years ago for the first time. And that -- so it's been slow to catch up. I think Windows was actually slow to deliver what was required for a really good haptic Trackpad experience, but the 0 doubt that the user experience is considerably better with haptics. You've got better force sensing, better feedback. It works perfectly across the entirety of the Trackpad. When you have a mechanical Trackpad, it's a lever basically. So there's hinges somewhere. And so if you press it in the corners, you don't get that good click experience. Have to Trackpad solve all of that, and you can make them as big as you want, so with that mechanical trade-off. They also have a lower -- a smaller typically can make them thinner when you look at the stack up. So again, as you wanted to make these devices thinner like it to a better experience. It's going to drive demand for haptics drivers. And then on the trackpad side, as I said, we've got a need for efficient power conversion to the extent that you're doing local or device side AI tasks, on these devices, that's only going to increase the need for getting everything you can, every jewel out of the battery that you can without wasting it in power conversion. So we've seen really, really positive response to what we're providing there. So yes, when you take all that together, we see a $1 billion plus opportunity. And obviously, we want to capture as much of that as we can.

Thanks for the detail. And I should have asked this part of my question. Is there any -- are you relatively agnostic whether it's an ARM-based PC or whether it's x86? Does that move the needle materially for you guys? Or -- I'm just curious that's a tens of your potential content...

Yes, that's a great question. That's a great question. From a potential point of view, we're totally agnostic. I think -- that's one of a number of factors that's kind of making the PC world interesting again as well. I think just that -- we'll see a lot of gains in power efficiency, a lot of competitive pressure, do that. It's great. Now we're kind of -- further along with x86 stuff by virtue of working closely with Intel and Microsoft around Lunar Lake. And that still obviously is the bulk of what our customers think about when they think about laptops. But we definitely see a lot going on in other architectures, on-based architectures we're very engaged on those fronts as well. And it probably doesn't need to be pointed out. But as you get to an architecture that's kind of ARM-based and boosted amplifier and on, you're talking about something that looks even more like smartphone from an architectural perspective. So that gives us an additional and familiarity.

A bit again, that's right, to AI, kind of stealing a lot of the Sunshine next year, in particular, for the group. I know there's a lot of talk about moving compute more to the edge, i.e., PC handset, whatever. And your last customer has been a leader in AI assistance forever now deploying their own GenAI. So I'm just curious from your perspective, do you think AI is going to drive a meaningful either TAM increase or refresh cycle or anything that's going to help Cirrus Logic, and -- I don't know if you could talk about sort of your "AI Play" like how do you guys benefit?

Yes. Thanks for asking. I guess a couple of caveats upfront. One, you know our style. We are not high merchants. So we don't -- we have attached AI to everything we're doing and kind of I thought about [indiscernible] So there's that. And then secondly, I think it's important for the investor community to know that when we guide, we're not baking in some assumptions about a super cycle or whatever. It's going back to the earlier conversation about guidance. But I think it intersects with us really in 3 ways and kind of like in increasing levels of specificity. So yes, I'm a big believer in on-device AI. I think we just -- it's just no way of addressing the privacy and latency concerns other way. So it's going to be a feature of mobile devices. So yes, of course, the extended drives an increased upgrade cycle, we'll benefit from that. Somewhat more specifically to the degree it creates increased pressure on the battery and power resources within the device and good space. I believe we stand to benefit from that. Most of the investment that we do is about comparatively advanced genomes for the kinds of devices that we deliver. That means we typically deliver products, which are smaller and more power efficient to our customers. So if you need more board space, freed up so that you can have more room for battery, more room for memory, because there is 0 doubt that you need a lot of storage and RAM to run an effective GenAI model, then -- we can help solve the problems associated with that. We also, of course, in the power conversion and control space. So this is somewhat more sophisticated than what we're doing on the PC side so far. We are -- so we've been shipping that product in -- with our largest customer for a little while now, and it sits between the battery and the rest of the device. And this is especially relevant in situations where you're making -- we have big spikes in demand for power from the battery, even more so when those occur when the battery is in a diminished state of charge. And those are -- without our technology, those are the kind of situations where either you risk damaging the battery or you get some kind of user experience degradation. So you start dropping frames and video, the system starts slowing down or even worse browning out to lose functionality. So we can provide solutions that address that. We have that IP today. That makes us very, very relevant in a space where people are more concerned about power efficiency and handling kind of peak demand. So -- yes. The 3 ways that I think we stand to benefit. Number one, the accelerated grade cycle to the extent that, that's something we see. Secondly, just all the demands on board space and power efficiency. And then thirdly, I think really the interface for a lot of AIs to the extent that there is more done through the camera, photographically, and through audio, through voice and so on. I believe we can enhance that experience because those areas, the interfaces with the real world, what you hear from the device, what the device hears from you and the environment and then the camera itself. These are areas where we are providing critical innovative silicon to our customers today. So we believe there are kind of more interesting problems that will come up with the horizon that we can help address.

And just since you mentioned camera in there. I mean there's a lot of talk about your largest customer potentially switching vendors for their camera. Are you -- does that matter to see this logic? I mean, are you going to support Samsung, Sony, does it matter to you guys?

We're pretty agnostic on that front. I guess our camera controller has drive -- they do a bunch of things, but they stabilization, focus, lens moving and so on and the main things. A lot of those are not connected with the sensor at all. So if we're moving optical elements for stabilization or auto focus, or some kind of exotic or elaborate lens assembly then that's completely decoupled from the sensor. We do also provide functionality, which is -- which enables the sensor shift stabilization. But that, again, should be pretty agnostic about the sensor unless the weight of the sensor changes radically or something which I did. .

Just dolomitic you were on cameras. Another question that I had, I keep asking these random questions. But your largest customers a lot of stuff in the press about the potentially begin to in-source their own modem, their own baseband next year. And I didn't know for Cirrus Logic, does that in any way, shape or maybe a dumb question, right, but is it impacting us a positive, negative, any impact at all, particularly Cirrus Logic, if it happens?

I don't think it's a dumb question, Rick. I have been asked it quite a lot, I will say. So which means I'm very glad you asked it because there's probably on a bunch of other people's minds as well. I can't think of any way that -- I think it's completely I can't think of any way that makes a difference to us. Certainly, the modem is just a completely separate subsystem. And as far as we're concerned, I don't think it has a material impact on anything we're doing in Power haptics or camera.

Okay. I just wanted to check that box -- It's for going to this. And then something I haven't heard you mention again, I did hear you clearly saying you're not high merchants, and I think -- I don't think anybody would accuse you that. But some of the bigger names in the supply chain for your biggest customers, the folks that actually make a lot of their devices, have told us repeatedly firsthand, that haptics for the side of the phone, replacing the mechanical power, mechanical volume buttons that's not a dead project and that, that could come back. I haven't heard a time line or anything on that. And I don't know how much you're willing -- I know you guys have said as far as you're concerned, it does happen. And again, maybe it goes back to your MO as not being machines, but what can -- what can -- if anything, what can you say about the potential there? It seems it's live and sometime in the future.

Yes. Well, yes, I mean, not high merchants, but I also don't like -- I mean if we thought there was something coming for us, I would try to read from that. I'm not looking to kind of jump out of a cake with some surprise for investors like I don't think that serves anybody super well. Yes. And just in case if anybody from my largest customer listening, we've never commented on what functionality was involved in the piece of silicon that we developed and then which didn't ship last year, which we talked about. And everything around that, that we've been associated with, that remains completely 0 in our model. So we make no assumptions. And I really don't -- I mean I hear the same rumors that you do. But the -- that could mean any number of things and there's plenty of ways of addressing some -- at least some of what's in those rooms without using any silicon from us. So yes, I got nothing -- no exciting news or teasers around that, and there's nothing that we -- nothing baked into our assumptions or expectations for the future around that one even.

Okay. Fair enough. You mentioned a couple of times some of the new products you've got coming out. And obviously, the new 22-nanometer codec. I mean, I'm curious there -- I know there's obviously some start-up costs and associated with that, and you're probably improving yields and everything still in the normal cycle of things, new products. But in the past, we've seen when you guys like me 55-nanometer like you said, as I've lost track 6 years ago, whenever that was. You did benefit on the gross margin line. It definitely showed out to you guys kind of ticked up into the mid-50s. And I'm just trying to understand the puts and takes this time on this transition. And could we see a similar benefit for a while in the model? Or -- and I don't know how much you could share on that, but I'd be curious your thoughts.

Yes. I mean there's a lot to unpack there. First of all, yes, like everything is more expensive on '22 and with the new product -- it's our first '22 product. So we don't -- we're not yet driving volume on that node. But really, everything is more expensive. The entire design process, simulation, wafers and so on. So you can think of that from a couple of ways. One is that, of course, we want to get paid for that. It's -- there's a lot of benefits that come from '22. We believe we're delivering a lot of value to our customer there. But I also think the reality of the situation for most companies like us, maybe even more because of the customer concentration, gross margin improvement is going to come mostly from working very hard on the supply chain side. So we've got a lot of work to do to get down that curve. We guided gross margin a bit more positively this quarter for the September quarter, because we had -- we took a lot of the initial kind of bootstrap costs in the June quarter that we just reported. And as we go forward, yes, we're going to continue to work really, really hard on the supply chain side. We believe there's opportunity to, yes, improve yield, improve wafer price as we drive volume on that node. And hopefully, we see some benefit, but there's no reason in our view to update the long-term model, the kind of [ 49 to 51 ] model at this point. I think to the extent that we can move that needle over the very long term, I think it's going to be driven primarily by business outside of our largest customer.

Perfect. And I know we've got about a minute left, so I'll sneak one in just I'm going to use every second if I get the. But I think you do boosted an I believe, remains 55-nanometer, I'm sure you'll correct me. If I'm right, I'm just curious what was in there as you were moving '22, like quite a lag, I guess, on that transition for boosted amp? And by the way, second part of that, it's just -- you may not be able to answer this, but what's the general rule them gen-to-gen with ASPs, when you make a lead like this and maybe doesn't apply or maybe it does apply to boosted amps even though you stayed on 55-nanometer. Is there a rule of thumb on ASP improvement?

It's -- actually, that's been all over the map in the past. So I wouldn't talk in terms of rule of thumb. There's been some ASP. The benefit -- we don't need to -- or we didn't believe we needed to jump from 55 to 22, for example, with boosted amp because the calculus around that depends on the ratio between analog and digital circuits. Analog doesn't shrink very effectively. In fact, often grow is to maintain the same performance as you go out -- But that's not to -- I mean, just to be clear, the 55-nanometer boosted amplifiers that we're bringing to market are it's a radical new architecture that's really significantly more efficient, and it also incorporates a lot of stuff that used to sit around the amplifier on the board. So you end up with major board space savings, major increased deficiency and actually potentially lower system costs for the customer. So to the extent that we can get an ASP accretion there, it's driven by the fact that we're taking other costs off the table and improving the system view from a customer's perspective. And we're very, very excited about the architecture that we're -- that's instantiated here as is the customer.

Great. That's very clear. So anyway, John, I may have taken a minute or so over, but I really appreciate it as on catching up. Great to see you, and thanks so much for your time in your in our discussion.

Yes, thank you so much. Yes, great talking with you as always, and thanks for the invitation and the opportunity, and thanks to everybody who's joined this session today.