All right, let's get started. Good morning, and welcome to J.P. Morgan's 52nd Annual Technology, Media, and Communication Conference. My name is Peter Peng, Small Mid-Cap, semiconductor analyst here at the firm. I'm pleased to have Michael Hurlston, President and CEO, and Munjal Shah, head of IR at Synaptics here. I've asked Michael to start us off here with an overview of Synaptics, a summary of the March quarter and June quarter outlook, and then we can kick off the Q&A.
Sure, Peng. Thanks for having us, right. Appreciate the opportunity to come in and chat. Synaptics is a, a broad-based semiconductor provider. We service almost every different market that there is, but we break our business down into three distinct lumps to help people understand it a bit better. One is mobile, and in our mobile business, primarily, it's touch controllers that get sold into high-end Android phones.
The second business that we talk about is enterprise and automotive, and that, as the name would imply, we have semiconductor solutions that go into enterprise class goods, whether that be PCs, enterprise telephony, docking stations, so we've got a multitude of different businesses in there. And then, of course, we have exposure to automotive. Our automotive business is primarily today the solution that drives the displays in cars.
The third business that we talk about is Core IoT. Core IoT is made up of a processing piece that we spent a lot of time on this particular call talking about. We've branded it Astra. Looks like Google is now out talking about Astra, so we'll have a little dispute with them on that. Our wireless business, which is predominantly Wi-Fi. We've got a bunch of different technologies in our wireless basket, but really the most important to us is Wi-Fi. Peng, you asked about the last quarter. I think it was in line with expectation. What we saw was our Core IoT, this basket of processors and wireless, doing quite well, and our enterprise and automotive, you know, sort of bubbling along at what we would consider a bottom level.
Into the guide, we showed, again, more growth in Core IoT, with enterprise coming up a bit, enterprise and automotive. I'd say in general, we've been a little bit surprised by the lack of recovery, I would say, in our enterprise and automotive business. We would have expected, had you talked to Munjal and I probably three, four quarters ago, we would have expected that to pick up a bit more strongly. It hasn't, and then, you know, our Core IoT business is performing largely as we expected. It's grown very rapidly, albeit off a small base, but it's, it's grown nicely and, you know, we'd expect that business to continue to grow.
Got it. Okay, let's kind of just turn to the near-term stuff because there's a lot of emphasis on the recovery profile. So if we look at the consensus estimates, the Street is forecasting sequential growth through the second half of the year and into 2025. Just given your visibility, orders, and new product ramps, how are you thinking about the recovery profile?
Yeah, I mean, I think the Street has taken down the numbers over the last couple of cycles, and I think that's fair. I think that we definitely see recovery. It's been slower than we would have expected, as I said, primarily driven in that enterprise and automotive business. The enterprise and automotive business, we don't see anything sort of structurally wrong, but the demand profile has been weaker and the recovery less strong, and that's predominantly driven by spending in the enterprise.
I T spending in the enterprise, and IT spending in the enterprise, the, I think is growing, albeit relatively modestly, but for the most part, it's going towards data center, server spend, and not toward the things we, we play in, the gadgets and, and PCs, the, the sort of hardware, small hardware that sits in the, the enterprise. So, you know, I, I, I'm still... Munjal and I have talked about this. We feel like there is, the, the recovery is, is there. We're, we're putting that into our numbers. It's just been the slope of that recovery that's probably not quite what we would have expected had you talked to us two, three, four quarters ago.
Got it. And, you know, after several quarters of undershipping consumption trends, I think the team noted that inventory levels are, you know, sort of normalizing across multiple product areas. It's always hard to gauge your, you know, inventory levels at your end customers. So, what's, I mean, the best way is probably to monitor is booking trend. Are you seeing positive booking trends out of your kind of direct customers?
Yeah, no, I think bookings have definitely improved. We had, you know, again, four, six quarters ago, we had really very, very weak bookings. Bookings have definitely improved. We feel good about the order rate that's coming in. You know, do we feel like that's consistent with what we would call the real demand? No, it's still not there. The bookings are not consistent with what we would call pre-pandemic demand, and pre-pandemic demand for us is quite a bit higher than we are today.
But I don't believe we have an inventory problem anymore. I think there's pockets here and there, but for the most part, inventory has been cleared out, and now we've got an order flow that's much higher than it was four, six quarters ago, but not where we would expect it, where we at sort of our fundamental levels of demand.
Okay. Now, on the gross margin, you have a target of 57% longer term. Just help us understand how to get back to those levels, and what are the drivers for getting back to those levels?
Yeah, the gross margin for us is sort of bottomed out at 52%. I mean, we were at 62%, 63%, you know, maybe two years back. It fundamentally fell off because of mix. Our enterprise and automotive business is by far our highest gross margin segment of the business. That has been where we've seen significant demand problems, and as such, the gross margin has been greatly impinged. With that said, it's come up about 50 basis points, 50 basis points a quarter over the last couple of quarters. So we've seen quite a bit of improvement. I think, you know, our guide is to 53.5%, if I remember correctly. So we've come off the bottom from a gross margin perspective as the mix has improved.
But in order for us to get back to 57%, we would really need to see our enterprise and automotive business get back to a level that would drive the gross margin up. It is by far our highest gross margin. It's well above that 57% type of target, and we expect it to get there, but I think it's gonna take a little longer than we would have forecast.
Okay. On your design win pipeline has been pretty strong in 2023 amidst a pretty severe semiconductor downturn. I think it's pretty reflective of your differentiated technology and product portfolio. Do you continue to see strong design activity into 2024, and how are you seeing, you know, the different end markets behave from a design win perspective?
Yeah, I think we're doing well in three areas, Peng. One is our wireless products. I mean, we are coming off a small base. We don't have very significant market share, and we feel like we're winning at an outsized rate in the wireless business. So we've positioned it well. Wi-Fi, there's a lot of different competitors, but it doesn't take a lot of movement in the share profile for us to have a significant impact on the company, and we think we are generally taking share in that business. Longer term, we're excited about our processors, and where we have differentiation there is on the AI component. I mean, what's AI? Nobody knows, but everybody wants to future-proof these devices that are at the edge.
Our play, as you know, is not having AI in the data center, the NVIDIA, Intel, AMD's of the world that have done so well in that business. It's more now running models at the edge of the network and having the ability to do inferencing on chip rather than depending on the data center itself. There are just a ton of emerging use cases that are coming up with that type of product, and our customers are kind of innovating alongside us. We feel like that's gonna be something that is a very, very differentiated part of the portfolio, one that won't hit for some years. It won't hit probably for two years, but we, given our design trajectory there, feel really, really good about that being a meaningful contributor.
And then the third area, and I know you've got some questions about that later, is our TDDI, right? Our automotive business. Our automotive business has some puts and takes, but on the put side is the drive to have more and larger displays and have an integrated touch and display technology that goes into those larger displays. So we are winning at an outsized rate there. We feel really, really good about our pipeline in automotive, and it's just a matter of seeing those designs through and getting them into production.
Okay. Switching to your business, the Core IoT. So within your Core IoT segment, the wireless business represents, you know, the largest portion of the revenue mix and is one of the fastest-growing opportunities. You have a solid position in the high-performance segment of the IoT market with your Wi-Fi, Bluetooth combo chip. You know, you have the Wi-Fi 6, the 6E upgrade cycle, and then you have some new products like your Wi-Fi 7 and then your broad market chipset later this year. So as you emerge, you know, from this downturn, how well-positioned is this business?
Yeah, this—I would say this is our best-positioned business. We're—we've invested in this, we've. As you know, the history of that business was taking a design database from Broadcom and repositioning and retargeting what Broadcom had done very well, which is execute in the mobile phone space, and move it to the IoT. And we did that, I think, very, very well. And now on top of that, we are starting to bring to market our own designs. So for the first, you know, year, year and a half, two years, we were simply taking Broadcom devices and then reselling them into a new market. Over the course of the last year, we've started introducing our own products, and those are much more geared toward the IoT market.
You have to take cost out of those chips for it to really make sense. And so we've got a much better cost point, as you correctly said, in this high-performance area. And high performance typically means distributing video. So you're moving video over a wireless link, and that, ideas for that market segments are security cameras. That's a big segment for us. The other segment is a good one, is like set-top boxes, OTT streamers. We've done super well in that segment. But really, for us to expand the wireless business, we need to go into the broad market... and broad market are appliances, home automation, industrial applications, and for there, you definitely need a much lower cost point, which we think we can bring.
Our engineering team has been working on this now for the better part of two years, and we think we can redesign, re-architect the Broadcom database and bring its cost point down to a level that makes sense, and we can get gross margins that are consistent with this 57% gross margin target you mentioned a second ago. So we feel really good. I mean, I think we've done a good job from an engineering perspective. Those chips, the broad market chips, the first ones we'll sample at the end of this calendar year, and we would expect those to start contributing meaningfully at the very end of our fiscal 2025. As you know, we're on kind of mid-year, but more meaningfully in 2026. And then the other initiative that we have going is Wi-Fi 7.
So Wi-Fi 7 will give an ASP uplift, will create a new design cycle around the high-performance area, the security cameras, set-top boxes, things like that. We think we're first to market with a Wi-Fi 7 device for the Internet of Things, and that too will be sampling to customers at the end of the calendar year, and we'd expect that to contribute again in late fiscal 2025, early fiscal 2026.
Got it. And kind of leads to my next question is, you know, you're taking costs down as you leverage for core technology and high performance to go into the broad markets. It's also a more fragmented customer base, you know, more distribution. So maybe help with some of the initiatives that you're trying to build out in your broad market channel business.
Yeah, that's, that's definitely, I think, our challenge. We have the right products both in the processor domain and the wireless domain. What we don't have is a channel. We've not done as well as some of the broad-based microcontroller guys, and even some of the wireless competitors that we have, have done a much better job building out that sales channel. We are. Historically, we've been structured to go after very large customers, Amazon, Google, things that, companies that service the Internet of Things but are much, much larger. That's the way our sales force is set up. So we have kind of a multi-tiered plan to get into the channel.
The first is on our wireless products, module makers, people that take the semiconductor solution, package that up with power amplifiers, other RF circuitry, and even some passives, and put it into a package. It's like a system in package type of approach. Interestingly enough, they also have a pretty big channel, and they have customer support, software support engineers that can go and service that channel. So we've actually, on the wireless side, spent a lot of time over the last couple of years developing module partners. We've had one that we've talked about that represents a big, big chunk of our revenue, and they've done a super job for us, selling into a diversified customer base.
But we're in the process of bringing it up a couple more, and those will help us with the scaling problem as we build behind it the necessary infrastructure for the channel. On the processor side, it's a bit different. Again, we're going to market first with ecosystem partners and helping them, and they help us sort of leverage the technology and go broader. We are going to need to move more quickly there to build up a channel. So intersecting this sort of two-year time horizon that we've talked about with our AI processors, we need to build up the structure of a larger sales channel, and we've started working on that now. We're already talking to distribution partners. More importantly, we're taking care of the things that we can control.
A unified SDK that's very easy for customers to use, a lot of the collateral, a lot of the self-help that they go to the web to go get answers, so they don't have to come into our support team directly. We feel like we're building that out in parallel, and when we land and get really, really ready to go broadly with our processor, our AI processor products, we should be in good shape from a channel perspective. But for sure, you called it out right, Peng, that is something that's a work in progress for us. We figure we've sort of solved this in reverse, where we have a very, very good handle on the products, but we need some work on the go-to-market piece.
Okay. Before we kind of move on to some of your AI initiatives, take a pause and see if there's any questions from the audience. Just raise your hand and we'll pass the mic over.
Yeah. Thanks, Mike. Thanks, Munjal, for joining us today. As we cover a lot of semiconductor companies, much like yourselves, many of them are coming from the bottom of the cycle and, you know, the question for a lot of our companies is: there's still some inventory correction, inventory burn with some of your customers, some of your customers have stabilized, some of your customers are starting to drive you guys to ship to consumption.
So I know it's always difficult when we're at this part of the cycle, but I'm just wondering if you, if you have a sense of-- you guided for, I think, $245 million in revenues for the June quarter. I'm wondering, as you talk to customers, look at distribution, sell-through and so on, where do you actually think that the normalized level of demand sits relative to that sort of $245 million guide in the June quarter?
A good question, Harlan. Look, I think we were sort of the first company to enter this downturn. Just by virtue of the exposure to end markets that we have, we were sort of first in. We were sort of the first also to call the bottom, right? So we called the bottom a couple quarters ago, and while we haven't seen the uptick, we haven't seen any further downtake. Where we see things, it's been a complex issue to explain in that we think the inventory is now largely washed out. We see pockets, but to first and second order inventory is not our issue. Our challenge is that the demand, the actual. So we think we are now shipping, at this moment, what is the demand?
We think, again, to sort of first order, we're shipping right at where the demand is, and to Peter's question, bookings are consistent with that sort of view. But what we think is, and we definitely hear that from our customers, this is not what they would expect in the real world when everything sort of shakes out. They expect demand to come back in, to a much higher level than we are today. When that is, is a TBD. We would expect that if you asked me, I would say, "Hey, back half of the year, we would start seeing a lift," but we would've felt probably two quarters ago that we would start seeing the lift now.
So we would expect that there is going to be a lift, not because of inventory, but just because the demand itself is depressed mostly, again, in our business. Now this is more specific to Synaptics, is that we have this enterprise business that when we talk to the IT guys that are controlling the budget, they are allocating their dollars toward AI, toward data center build-outs, and away from PCs, hardware. They're sweating those assets for longer than they did previously. But at least as far as I see, there's nothing systemic in it. We would expect demand to come back, and when it does, our numbers will come up.
Okay. Maybe just talk about some of the AI initiatives. So, you know, AI is beginning to proliferate at the edge, right? It's where the inference is done on the device to minimize latency. The team has recently launched the Astra IoT Compute platform to intercept this growing demand trend. So maybe just give this opportunity for you to talk about, you know, how this platform works, what kind of technology are you embedding it in the platform, and the opportunities that you see.
No, we're super excited. I mean, I think, you know, we've already been asked a set of questions this morning about it. We launched it in April officially, but we've been out talking to customers about the platform even before that, and we already have design wins. Those design wins will take a while for it to start hitting the numbers, just by the fact that this is the first decision that customers are typically gonna make in a new product, is what's the processor that's gonna drive their next generation goodie? But we already have wins, and we're pretty excited about those. They're meaningful wins in the area.
And what we have is we have a range that goes from a microprocessor, think about a full-throated processor that can run an Android, that can run, you know, a real operating system can be run under the hood. 2- and 4-core, capable of TeraOPS of inference, very, very, very strong offering. We think we're the first to have a true AI offering that's at the edge. People are talking about it. People are, you know, making a lot of noise. We have it. Our customers have it. They're already working on designs that they're embedding in their next-generation devices. What's interesting about it is, again, an underappreciated problem is the compiler.
You need to take this code that's trained, that's developed in the data center, that's the model, and you need to be able to now compile that onto the chip itself and run it locally, and that, there's compression in that, there's memory optimization that needs to take place in the compiler. We have that compiler technology, too. That was an acquisition that we did probably two years ago, that turns out it's a really key piece of the puzzle. You know, what, what we can bring in this area, again, is use cases. So we're innovating alongside our customers. Our customers might have very simple, you know, people identification types of use cases, or they may have more complex use cases where they're trying to actually do face ID or something like that. We can do that.
We can do all of that on-chip at the edge of the network. And then we also have an MCU, so a microcontroller type product, that, again, is extremely innovative because it's designed for low-power use cases. I would go in a battery-powered camera, I would go into electronic shelf label, I would go into anything with a very small battery that's gonna need some measure of AI. And what we've done here is we have the ability to kind of ratchet up the power as different events occur. So let's assume that it's operating in very steady state. We're consuming, you know, microwatts of power. It's a very, very, very low power consumption device that is running and able to detect events.
And then when it does detect an event, when something happens, okay, is that, let's talk about a security camera example, is that a squirrel? Is that a cat? Is that a person? It now steps up to the next level of power consumption. It's an innovative technology. We call it gears. Now, it moves into from first gear, it moves into a second gear, and there you're going to consume a little bit more power. You're going to be doing some more inferencing. You're trying to make a determination, as I say, "Okay, what is that object?
Oh, it's a person." Okay, now it moves to a second gear, a new gear, and now it's trying to say, "Okay, well, is this a person that we recognize and understand should be on the property, or is this, you know, somebody that, that shouldn't be there?" So at each level, at each gear, we're consuming more and more power, but the whole idea of the chip is to be able to operate it, because 98% of the time, you're in this very, very low power consumptive state, and you want to be able to operate off a battery. And we are able to do that by having a, an extremely low power at that level, but then ratchet it up as we're making more and more decisions and doing more and more inferencing.
So the microcontroller, we think, again, is a very differentiated AI product, and we now have products that span from an MCU all the way to an MPU, a four-core MPU, that again, is able to do teraops of compute, and one software stack, one, that extends from top to bottom. So our customers, many of whom are going to use an MCU, and in some cases, use an MPU, a microprocessor type of unit, they can get from us one SDK, one software load, and have that extend across the entire portfolio. So it's the compiler, it's the actual ability to do AI, which we have today, and it's the extensibility of the software, Peng, that really gives us, I think, a differentiation and a level that, you know, nobody else has at this moment.
You mentioned some of the design wins that you initially see. Can you just kind of provide some color on what end markets are you or end applications are you seeing these strong demand pull from?
Yeah, I mean, I would say three interesting ones. One is kind of the home appliance, home automation market. We're, we seem to be doing relatively well there. The second is security. So we talked a little bit about the security camera use case, and it might be a processor that goes into a central control unit that is doing a lot of the compute there, and the central pulling information from cameras and things of that nature.
Then I'd say the third is video conferencing, interestingly enough, and the video conferencing use case seems to be able to enhance video quality, sound quality, depending on the link. We're able to do that actively. I mean, it seems very, very simple, but, you know, you have a lot of differences in sound and video quality over these Zoom links, and we're able to manage that, and our customers find that pretty darn compelling.
And during our last earnings calls, you talked about a $20 billion SAM opportunity for this Astra platform. I mean, can you just kind of flesh out some of the revenue potentials that you can see for Synaptics and the go-to-market strategy here?
Yeah, I mean, look, Synaptics has been characterized generally by having a very dominant position, whether you think about our docking station business or our audio headset business, or even our PC touchpad business, we have, you know, a very, very dominant share, and frankly, small markets. And so we are using that dominance in those small markets to generate the cash and allow us to invest in a couple of really big markets that are very fragmented, and those are wireless, right? Everybody understands the wireless semiconductor market. It's pretty competitive, and this processor opportunity. So if we're able, as you correctly said, our calculation is a $20 billion SAM of things that we can go after, and that excludes data center, automotive, big, big markets that we don't intend to pursue.
If we can move the needle even slightly and become, you know, 1% of that, 10% of that, whatever, that's a big move for Synaptics. And so we feel like we're on track to be a meaningful player. It's not we're never going to get to 50% share or 80% or 90% share like we have in some of our other businesses, but we most certainly can become a 10% player in this large market, and we think we have already the design traction. It's probably not, you know, a $1 billion-$2 billion design traction, but we already have the early signposts that we have the runway to build into that kind of number over time.
Okay, just kind of switching to your enterprise and automotive. Another area that you guys have, you know, AI exposure is your PC segment, right? The inventory digestion is now behind us, and there's a lot of optimism around the new AI PC launches for the second half of this year. You know, our tech team is forecasting a 3%-4% AI penetration rate this year, and then 10% in 2025. Maybe just help us frame the opportunity set here for you guys.
Yeah, look, I mean, we kinda go as the PC units go, and we talked about this on our call. We've not seen orders that would indicate a big bounce in the second half. We hear the exact same thing that you're reading out, Peng, which is that there's a lot of optimism around growth in the second half in the PC sector, driven by the AI PC. We haven't necessarily seen orders consistent with that. But if it happens, look, it's good for us.
I mean, we have exposure in the PC. We have an AI chip, actually, that goes in the PC that a lot of people talk about, and that's the user presence function. That opportunity alone for us is something that we think we can build on. It's a very small piece of our business today, but we think it's something that can grow for us over the next couple of years. But, you know, as if units increase in that PC market, it's good for us. We just haven't seen orders that would be consistent with that.
And then your enterprise segment, it's, you know, things look like it's, you know, bottoming out. You're implicitly also guiding that there's gonna be maybe modest sequential growth in the June quarter. You know, from my perspective, and you sort of alluded to it, it seems like, you know, companies are kind of reprioritizing spending to AI, you know. So maybe just talk about how do you see the long-term, you know, structural health of this end market?
Yeah, I don't think that there's anything that's changed in our view as to structural health. We think that there's a demand level. We're under shipping the demand level. We certainly are, you know, I think, again, inventory issues are behind us, but we feel like the demand, the natural demand of PC, the PC units are down. We know that docking stations are down considerably, right, relative to where we'd expect them. Audio headsets, which we have exposure to, down. Enterprise telephony down, all below what we would consider sort of the normal demand rate. And to your previous question, until we get some return in that, our gross margin is gonna be impinged, because those are our best gross margin businesses.
When we talk to the IT managers, they know that they've sweat these PCs a lot longer than they need to. They've sweat the docking station assets, headset assets for a lot longer than they would in a normal cycle. So we'd expect it to come back. We'd expect demand to increase. You know, we'll see what next year brings. IT budgets are typically set at the end of the calendar year, as you know, and it goes into the next calendar year, 2025. There's nothing that would lead us to believe that there's not going to be some step up in IT spending, but you know, it's probably too early to call that because the IT budgets haven't necessarily been set.
Yeah. Okay, one more final question. Back, I mean, this kind of ties back to it. Back at your Analyst Day, you talked about a 10%-15% revenue CAGR. I guess, what's the confidence level in, you know, getting back to those growth rates?
Yeah, I mean, we talked about, you know, high growth in Core IoT, right? That would be, you know, growing at a pretty good clip above the 10%-15%, and we've seen that. The Core IoT, I think, has done actually quite well over the past couple of quarters, delivering outsized growth. So that's happening. I think the issue that's impinged our overall growth is the choppiness in enterprise and automotive, and we've spent the balance of the discussion talking about that. You know, we don't think that there's anything structural in that business. That business should be growing at sort of 3%-5%.
You know, this guide is in line with 3%-5%, but I wouldn't be surprised if it continues to chop along for the next couple of quarters until we see an inflection in enterprise IT budgets, and that's when I think you're gonna see that business return to sort of a 3%-5% CAGR with our Core IoT growing, you know, something, something much higher than that.
Okay, with that, time's up, and thanks.
Thank you, Peng.
Thanks for being here.
Thank you.