I th`ink we're live now. Good morning. I'm Melissa Fairbanks. I am the Analog Semiconductor and IT Supply Chain Analyst here at Raymond James. We are thrilled to welcome back SiTime to our conference again this year. This morning we've got Beth Howe, the CFO. I believe this is your second year attending,
Yes.
so we got you early on. Before we start, I just want to read some kind of, you know, like a safe harbor statement. So the discussions that we're about to have include forward-looking statements that involve risks, uncertainties, and assumptions, which are further described in SiTime's SEC filings, including SiTime's most recent Form 10-Q, and that actual results could differ materially and adversely from those anticipated or implied. SiTime assumes no obligation and does not intend to update any forward-looking statements. For more information, please visit SiTime's Investor Relations website at investor.sitime.com.
So to kick things off, I think for those that might not be as familiar with SiTime, Beth Howe, can you start with a short overview of the company just to give us an intro?
Sure. Hey, good morning. It's good to see you again.
Good morning.
SiTime is really silicon timing, and we define the category of precision timing. Our thesis is really that all the big trends in the marketplace, whether it be AI, autonomous driving, IoT, faster compute, lower latency, all need precision timing. And not only precision timing and the increased performance characteristics, but also to be resilient under challenging conditions of temperature, vibration, shock, small size, low power. And so that's really what we're all about, is delivering this precision timing to these increasing performance requirements from these different applications. And we're also very diversified. Our business is across automotive, aerospace, AI, data center, communications, as well as consumer and IoT. So a very varied application set that really drives our business. We've shipped over three billion units over roughly the last 15 years, and looking forward to the next couple.
And very unique business model in this industry. You are fabless. So not only a high-growth company, but also fabless. So you said the magic word, and we'll get into AI data center later in the conversation. But maybe just review what we've seen in demand trends over the past year or so. We did have an inventory overhang entering the year. We've seen sequential revenue growth since the March quarter bottom. So maybe talk about some of what you're seeing in your end markets.
Sure. Well, it sounds like you want to get to some of the industries a little bit later, but we'll get to this. So in terms of the markets, I think we've seen growth in all of our businesses this year. And as you said, we've seen sequential growth throughout the year and guided to that for our fourth quarter as well. Even though markets have been mixed, I mean, some have been on fire like AI, others have been a little bit more sluggish. But given our high-performance products, we've been able to grow across our end markets. We did see the inventory overhang that pretty much every company did through 2023, but really worked through that correction in the first- half of 2024 as our customers burned through that inventory.
But as we, like I said, even though the markets may be dynamic or mixed in terms of their performance, we have been growing in all of our markets.
Okay. We can talk about some of the unique opportunities that you've got in some of these kind of more recent catalysts in data center and automotive. But I would like to discuss some of what you're seeing in your more traditionally seasonal markets like communications, enterprise, and mobile, IoT. Have we returned to a more seasonal, more normal seasonal pattern like in mobile or IoT reflected in the strong September quarter growth? Or was this kind of a one-time inventory channel fill? Maybe discuss some of the dynamics that are going on with the normal seasonality.
Sure. So again, what is normal, I think, is a little bit, and I think that's changing a bit. As we grow and we grow across these multiple sectors, I think we do expect that some of these seasonal patterns may kind of dampen a little bit, because especially in consumer, you have that, which is the most exacerbated seasonal patterns. In addition, in terms of channel fill, I think one of the things that folks may not understand about our business is we sell to an end customer. While we sell through the channel, most of our business goes through distributors. We're actually using them to fulfill end-customer demand. So we don't stock and destock in a traditional way. We're really; we've built a bespoke product for an end customer, cloud service providers, automotive, whatever that is.
And so that's how we think about the industry is really fulfilling demand for those end customers. As we look at it, I think consumer will continue to have its seasonal patterns. Some people are surprised that many other techs also have somewhat of a seasonal pattern in the sense of more strength in Q2 and Q4 versus Q1 and Q3. So I think those will continue.
Okay. Okay. Great. So we have seen a pretty remarkable recovery in growth this year. I think if I'm doing the math right, the December quarter outlook implies full year growth rate somewhere in the high 30% range. Kind of an unusual story in semis in general. So it's good to see maybe return to more normalized conditions in some of those more seasonal markets. But in data center and cloud compute, I mean, this is something that has gotten a lot of people interested and a lot of people excited about the story. You have had some new opportunities there beyond just kind of the traditional x86 server opportunity. Maybe talk about how precision timing has become critical at all points in the architecture for some of these higher performance data center compute servers.
It really has been more and more important. You're seeing the importance of performance across the architecture, and maybe I'll bifurcate it into two, the kind of compute side and the networking side, so clearly in compute, we're seeing opportunities for precision timing, whether it's CPUs, GPUs, processing units, more general TPUs, et cetera. I think the more exciting almost is the network fabric, because you see the need for precision timing as you increase the data throughput in lower and lower latency, there's a higher and higher need for precision timing. You see that as you go from 400 Gb to 800 Gb to 1.6 Tb, whether it's NIC cards or SuperNICs, top-of-the-rack switches, all across that network fabric where you need that high performance, more throughput, lower latency. We're seeing opportunities across that landscape.
These are applications that really can't be addressed by some of the legacy quartz crystal timing solutions. You really need to have this type of silicon-based and the MEMS-based precision timing solution.
You certainly get better performance out of the silicon-based timing. I mean, quartz is a big market and there's going to continue to be a place. I think we are displacing it, but we're not going to displace it everywhere. It's been around a long time, so I don't want to imply that there will be applications where quartz maybe makes sense, lower performance, et cetera. But when you do need those higher bandwidth applications, small size or also lower power, those also are areas where we excel.
Okay. And I think some of that is extending into your auto and industrial, certainly in aerospace and commercial space markets. These have been relatively new growth drivers for you that you've invested in over the past few years. I like to call them kind of mission-critical applications, where if you have a failure at the timing solution point of contact, you potentially lose the ability to have self-driving vehicles or lose the ability to have communications within aerospace. Maybe talk about how this was an area that we really hadn't even talked about at the time of the IPO, which really wasn't that long ago, just several years ago. Maybe talk about how you've grown into that opportunity and where you see that market growing longer term.
Those really are interesting applications, and there's so much to talk about. Maybe I'll start with automotive and we can kind of go from there, so you're right. When we went public five years ago and we just passed our fifth anniversary, automotive really wasn't on our radar screen, and now it's become a meaningful part of the business. Automotive is an area that has a lot of challenges, and to your point, you've got to have performance and uptime. We've actually, to address that, actually just launched our FailSafe, which not only is, again, the high-performance timing, but also has a feature that allows it to kind of tell the system the health of the system that's going on.
So if it is looking like it may fail, it actually sends a signal so that the driver or the automobile can actually detect that and go and kind of exit safely as opposed to kind of having an abrupt failure. You can kind of think of it like a run-flat tire. Even if it goes flat, you've got about probably 50 miles so that you can get to a safe place with the automotive. And it's also an area that really highlights the importance of the precision under harsh conditions, right? Temperature, vibration, shock are all elements, EMI of a car that you've really got to be able to address in those applications. Another area that goes across those different applications of automotive, aerospace, even industrial is GNSS. This is an increasingly important area, right?
People talk about GPS spoofing and jamming and being able to have local time and what's called holdover so that the precision time is known for longer periods of time is increasingly important in these areas so that you're not reliant, for example, on a GPS signal to know the timing, which does everything from not only location, but also from safety perspectives in terms of understanding how the calculations that go into autonomous driving, and so being able to have long holdover is also increasingly important for things like GNSS.
Thanks. I'm glad you brought that up. That was an important one. So I think one of the, you know, all of these areas that are driving this outsized growth, what should we think about as a normalized revenue growth rate on these new opportunities? I know Rajesh Vashist would like to say one thing, and then as the CFO, maybe you're tempering that a little bit, but.
Well, you know, again, and rather than give guidance, I think for this year, as you said, we expect to be, and I think consensus is about $199 for the year, which gets us to roughly 38% growth for 2024. As we look out, I think as 2025, 2026, our target is to grow at roughly 30%, kind of +- is where we'd like to be. Again, some years we may be a little more, some years a little. And that's really made up of a variety of applications. I think we talked earlier about the diversification. You've mentioned it here, where we're in roughly 300 different applications across that spectrum from communications, enterprise data center, automotive, industrial, aerospace, et cetera. And about 10% of those are the really high growth applications. So you think about that broader base, maybe they grow 10%, 15%.
And then you've got these other applications, for example, in AI and other areas that are growing much, much faster than that. And so that's really that combination of applications and growth rates gets us to where we want to be.
All right. I think one of the more exciting opportunities as we get to this stronger top-line growth is the operating leverage in the model. Maybe discuss how you've directed your investment dollars. You did have a period during the supply chain crisis where you were able to invest more in R&D, and that's kind of what's delivering some of these new revenue growth drivers. But where the leverage is in the model, where you think longer-term OpEx needs to be, any areas where you feel you need to target increased investment in the near term, or do we just benefit from this huge leverage?
No, I think it's a great point. And so if I step back a little bit, we grew dramatically through that period, kind of 2020 to 2022, and invested and basically used that cash flow to invest in the business. And we're benefiting from that now as we launch a number of new products into the marketplace across our different applications. The majority of that investment was actually directed in our communications, enterprise data center application space for two reasons. One, we saw the growth coming in that area. And secondly, we are a platform-based company that then we build derivatives and configurable products on top of that. And those CED, as we call communications, enterprise data center products, then we can kind of leverage into those other areas, for example, aerospace, defense, automotive, industrial. And so the derivatives really apply much more broadly.
And so that's really what we used the cash flow and R&D for as we really grew very quickly through that period. And then we saw 2023 as kind of a temporary bump in the road, if you will. And so we trimmed, we tightened our belts, but we wanted to maintain those investments because R&D is really truly the lifeblood of our company. And we've seen that coming out the other side. And now we're in the process of bringing to market those various products. Some of them are just about to come, some are launching, some are ramping into mass production. And so we're really benefiting from that and continuing to invest in R&D. So I think about going forward, we want to continue to make those strategic investments where it makes sense, whether it's R&D, go-to-market, right? We see lots of opportunities.
But on the other hand, I think there's a lot of operating leverage in Q3, the most recent quarter we reported. We actually had 75% flow-through gross margin to the bottom line. So strong operating leverage in the quarter. Now, I'm not going to say we're going to do that every quarter, but I think we can over the next kind of year or two grow revenue much faster than the cost and therefore get the operating leverage in the model. I want to invest where it makes sense, where we've got good ROI, absolutely make those investments, because I think that's the right thing for the long- term. But overall, I think there is a good operating leverage in the model.
And I should have this number memorized by now, but I know the number of new products that you've launched each year continues to accelerate. And we've got a significant number of new products that have launched this year that will probably fund the growth or fuel the growth for the next several years. So maybe you kind of touched on it a little bit, but we did see the gross margin dip below 50% earlier this year. It has started to trend higher going forward. What is a normalized margin? We kind of saw peak margins during that peak supply chain crisis period. How much of that is going to be sales volume versus mix? As a fabless company, you do have your own dedicated capacity with your suppliers. But what should we think about gross margin going forward?
First, I think you just misspoke.
I may have.
We dipped below 58%, I think just below to 57%, but we've been above 50% pretty much since 2020. I think 2019, we're at 46%, but we've been well above that. And our long-term target is to get back above 60%. But we've been in the 58% range here. And I think really the gross margin is driven by three things. One is the sales volume, and we've seen the benefit of that. The second is product mix. And as we look at it, the areas of like AI, even industrial aerospace are accretive to our gross margin. I think as you look at over the long term as well as the medium term, that mix is really going to be a driver of gross margin. As AI is growing faster than the rest, that is a positive to our gross margins.
And then the third one, and I think you touched on this as well, is the new product introductions. And we have seen, and I've talked about on the call and other forums, the fact that as we've launched roughly a dozen new products or are launching into the marketplace, it takes a little while to get those process yields and cost curves kind of to the mature levels and get down to that area. And so we've really, I'll call it prioritized a bit, getting product in the hands of customers or revenue growth and really focused primarily on revenue growth. The cost curves will come. And so I think as you look over the next couple of quarters, as we move into 2025, we will have that focus on getting those product costs and yields more optimized.
That should be a benefit to gross margin over time as well. We really, again, want to prioritize revenue growth and getting products in the hands of customers.
I think maybe one of the other times that we spoke, you like to focus on gross profit dollar growth, not necessarily gross margin expansion.
At the end of the day, we want to get that. It's really about dollars, whether it's gross profit dollars or net income and driving cash flow.
Yep. Okay. I'm going to stop here and question.
Hi. I'm relatively new to this story, so maybe just could you just tell me how you size the TAM that you're in and what's the long-term growth outlook for it, and just how much of this market do you have and say your largest competitor have in terms of how this market's divided up and your outlook going forward for that?
As we look at the TAM for timing, the overall TAM is roughly $8-$10 billion inclusive. Our service market or SAM is in the kind of $3 billion range as we look at that. And about $1.3 billion of that is the communications, enterprise data center. You've got about $500 million in automotive and then the remainder in industrial, consumer, et cetera. So that's the size of the market. In terms of the competition, we're the only company that has a full portfolio of timing solutions, which are resonators, oscillators, and clocks. And we expanded that portfolio about a year ago when we acquired the assets, clocking assets from Aura Semiconductor , which expanded our TAM as well.
In terms of the competition, so the competition is really quartz vendors and there's, I don't know, 25 of those, NDK, Epson, Kyocera, a number of those, or the clocking vendors like Renesas, TI, Skyworks. So it's really a broad set of competitors, but they're really doing in general different technology. And so our thesis is really about using silicon-based solutions for timing to disrupt the, call it 70-year-old quartz market.
Is this a leapfrog of sort of the existing competition type situation allowing you to gain share?
It really is a leapfrog, and it's a leapfrog on two dimensions. One is performance, and this gets kind of geeky, but performance characteristics around performance under high temperature, jitter, there's holdover, there's different stability characteristics of the performance of the technology, but also differentiation in terms of, I'll call it supply chain and availability. But because we're based in silicon, we have very short lead times. We're programmable. And so, and as you saw through the shortages, we were able to deliver in weeks and months versus quarters, and maybe lead times got to be 20 + weeks or even longer for some of our competitors. So our ability not only to have the performance characteristics, but also customer service, delivery, programmability is a differentiator for us as well.
Thank you. That actually segued perfectly into my next questions.
Okay.
Since the IPO, even around the IPO, there was some discussion about you did expect new competitors to come into the market with a similar type of full timing solution, silicon-based timing solution. And it's been several years. We still haven't really seen a competitive challenger in the market that has the breadth of the portfolio that you do. Are you starting to see anyone coming close? Do you have an understanding of who's potentially working on this and what the competitive dynamics are going to look like going forward?
Well, I think we've continued to maintain a multi-year lead on the competition. There is a little bit of competition in the silicon-based or what we call MEMS timing. There's a small startup in Canada that I know that is working on a solution. Most of the competition, though, is more of the traditional players and quartz. I think the other thing with the traditional players, a lot of them like TI, Renesas, Skyworks, is the timing pieces of their business are relatively small portions, one, two, less than 5%. And so really it's not their core focus. Whereas for us, all we do is timing. And I think customers really appreciate that focus on timing and the breadth of the portfolio.
Whereas others might have a point solution for different places in the timing portfolio, we can offer them the full portfolio across oscillators and clocks, whether it's jitter cleaners or buffers or TCXOs or OCXOs. And so they really appreciate the fact that we're focused on it and that we have that breadth of portfolio.
I think the Aura Semiconductor acquisition really kind of helped you leapfrog because it completed. Maybe discuss how that helped you to save some upfront R&D dollars because they had existing products in their own portfolio and you were just able to integrate that. Did that result in either accelerating your new product development or complementing the existing R&D, maybe kind of saving some overall spend there?
Yeah. The Aura Semiconductor acquisition that we did, we completed that about a year ago here, really expanded our portfolio, and it's something we talked about since the IPO. Our primary focus at the IPO was in oscillators and that we always wanted to expand into clocks as well, and this acquisition really fulfilled that promise of having that broad portfolio, and we're really excited about the Aura Semiconductor transaction. It's gone really well in terms of the integration of the people and the technology, as well as from a customer perspective, as I was saying earlier, that customers really appreciate a single provider having that broad portfolio.
And even some of the products that maybe weren't as highly differentiated, some of the buffers. Actually we found customers really like to be able to buy not only the differentiated products, but also some of these more standard-based products because they can get it from a single provider. And so it really has, I'll call it, fulfilled the promise of the IPO, if you will, in terms of that broad portfolio and given us a calling card with customers as well because the clocking solutions, the clock tree is also something that's designed earlier in the design cycle versus the oscillators. And so having this broader portfolio of clocks has also given us a new engagement point with our customers earlier in the design process to really have that architectural conversation around the whole clocking tree. And so we've also found that to be very beneficial as well.
That's a really good point. That's a really good point. So following Aura Semiconductor, are there any technologies, products, platforms, or something that would be complementary to the existing portfolio or where you're looking at several years out?
Well, I think right now we're focused on this portfolio and continue to bring products there. Again, in terms of looking at opportunities, if there's opportunities to complement to grow our portfolio, we'll definitely look at them. But I think the portfolio that we have in terms of resonators, clocks, and oscillators really does give us that full portfolio. As you look more kind of out there, again, our resonators is a smaller part of our business at this point. I think that could be bigger. And then if we think bigger or think more broadly, you think about atomic clock precision and at some point having an atomic clock kind of being able to be handheld, if you will. Now, again, those are very expensive. That's a long way out.
But at some point in time, there may be the opportunity to really look at how do you have that level of capability kind of available.
Yeah. Excellent. So one of the other great aspects of this story that I love, you've got about $18 in cash per share exiting the September quarter. Maybe talk about what some of the priorities are for the cash use. I know the Aura Semiconductor acquisition, you do still have some obligations that you're holding back some of that cash, but talk about where you as the CFO like to try to direct this.
So our cash at the end of Q3, as you said, about $18 share, $435 million on the balance sheet and no debt. I think that's also important. That's net cash, not just gross cash there. So as I look at that, first and foremost, having a strong balance sheet for a company like SiTime is really important. We have so many customers that are tens and hundreds of billions of dollars of customers. And so they rely on us. And in many cases, we're sole sourced for them for the products that we deliver. And so having a really strong balance sheet is a really critical part of that conversation. You also see that in the inventory. I want to make sure we're carrying sufficient inventory for assurance of supply for those customers.
So, first and foremost, that's really important to us as a company, kind of the size and characteristics of SiTime. Beyond that, again, we'll look at opportunities to blow that cash. But again, at this point, fulfilling the Aura Semiconductor obligations got a little bit more to go on that, as well as then maintaining that strong balance sheet. Really important.
Okay. We did see CapEx increase. You are fabulous, but you do make internal investments in some of your capacity. CapEx increased as a percentage of revenue last year, mainly just due to the lower sales volume. But what should we think about as a normal run rate percentage of revenue, or is it going to be lumpy depending on where you see opportunities for investment?
So I would think CapEx more in dollars than a percentage of revenue. To your point, last year, the percentages were more a function of the revenue than the CapEx. But we do make investments from time to time. You saw some of that in Q3 as well, that we're investing in equipment for our systems. We also sometimes will look at the economics. We are fabless and so rely on OSATs for the building and manufacturing of our products. But depending on the specific products or the OSATs, we may do some sharing with them. For example, we might decide to invest in some of that CapEx for their line when we think the economics make sense for that or we'll co-invest along the way. So I would think about that. Really, it's an investment cycle versus a percentage.
Okay. Perfect. We just have a few minutes left. Any questions from the audience? Nope. Did you have any closing remarks that you'd like to kind of, I know Rajesh Vashist normally likes to put a big wrapper on it at the end, but I don't know if you'd like to take the opportunity to have some closing remarks or maybe something that we didn't ask about today that you've been asked about recently?
Thanks so much for having me. I'm excited to be here. I think.
Oh, sure. Sure.
So just in terms of the sales cycle to the customers you're selling into, vis-à-vis your visibility, is this the type of thing where you kind of have a long-term design win type of sales cycle and then your visibility is more related to the end market demand and that's shorter? Maybe just a little explanation like that because it sounds like you get pretty well embedded over a long period of time and then as would be typical, shorter-term visibility for your end markets and that might be a bit more macro-sensitive.
I think you characterized it well. So we have design wins with our customers and those design wins give us then the visibility in that socket. And the length of design win will depend a little bit on the end market. Consumer design wins can be one to two years. Aerospace and defense can be five plus years, very long cycle. Automotive in the middle, kind of there. So when we get a design win, then we've got that visibility and then it varies, as you said, based on that end market demand. And so as we look out over the next year or two, and that's also where we say sometimes it takes a while to ramp a new product because you've got to then get it into the hands of your customers and get it into the next design win.
Are you usually a single source vendor in the markets you're just touching on?
As we characterize it, we characterize roughly 80% as single source. And that's really in three categories. One is single source in the sense of we are the only product that can work and has the performance and the characteristics for that customer and they single source it for that reason. The second, and that's about 30%, about 30% is they single source with us. If they had to in a pinch, they could requalify another competitor and have that. About 20% is single sourced by choice, i.e. they could use a different vendor, it's compatible, but they choose to be single sourced with us. And the remainder is multi-sourced. And both are important. So having a single source clearly very good for recurring revenue. In some cases, we also want to get in the door.
Sometimes the way to get in the door is to be that second source in a socket and then build the relationships from there.
That's a really good point. We're at the end of time, but if you'd like to just, I really appreciate you coming here and talking with us again today.
Thank you for having us and really appreciate being here today.
Great. Thanks so much, Beth. Thanks everyone.