semiconductor analyst here at Barclays. Lucky enough to have SiTime CEO, Rajesh Vashist with us. Thank you so much for being here.
It's a pleasure.
I think you want to kick off with some disclosures, and then we can [inaudible]
Last time I did not do that on the presentation. I'm not going to put you through that again, but the usual disclaimer is on. That's it, right? All right, perfect.
All right. So again, great having you here today. It's been a busy couple of weeks for you.
Yes.
I want to start out with the Aura Clock business that closed last week. Can you talk about why you did that deal, from a strategic perspective, and why was this such a must have?
Right. SiTime is focused on timing, and timing has two or three components to it. One is the oscillators, which is most of our business, the resonators, which is business we're going to grow into. We know how to do that natively. But the clocking business is one that we've talked about since the IPO, and one we wanted to get into, significantly. It is because the oscillators we sell and the clocks that we now will sell, sit at the same motherboard, the same system that customers consume. I t fulfills our promise of being a single stop purveyor of all timing-based products. W e found one of the best companies with the best technology, found it at a fabulous price, so the rest of it is the hard work of doing it.
Yeah. So does this take you into another realm in terms of the competitive environment? When you're moving into more of a full solution, does the competition change and does your go-to-market need to as well?
Absolutely. A s you recall, we've been competing against quartz crystal vendors, people like Epson, Kyocera, TXC, et cetera, mostly Asian. T he quartz business is a 60, 70-year-old technology, so there's a general slowness of innovation in this business. We're now moving to clocking, which is a companion chip to oscillators, and in clocking, our competitors are Renesas, TI, Skyworks. P eople and competitors like that are clearly a little bit more semiconductors. They're more aggressive, but at the same time, they're not very focused on timing. Timing for most of them are 10 billion revenue company, and the timing piece of it is about a $100 to $200 million business. So this is not something that's investment for them.
So it gives us the edge in going to our customers in the networking telecommunication market, whether it's a Google or it's an Infinera or it's a Nokia, and giving them a full solution, fully integrated. It's a game changer for the customer.
When you look at the model and how that changes when you're moving into this full solution, both from a spend perspective and from a margin perspective, can you help just give me a feel how your business changes? Your go-to-market seems like it has to change because there's new competition, but it's, it's easier for you, right? You bring more to the table. But how does it impact your P&L?
At the top level is where the biggest contribution is. It's a little bit slower to get to revenue because it needs some momentum, needs a design win momentum, and we're not getting revenue. W e're going to get the momentum revenue as we go, probably call it 2025, 2026. Gross margin is actually favorable. Most of clocks sell around 70% gross margin, so it's a more favorable business. O n the OpEx side, because we're getting fully formed business or fully formed, excuse me, products from Aura, 50 products by the early part of 2025, it's gonna be relatively low OpEx. T hat sounds like a pretty good deal. Something that grows your top line, the gross margins are equal to or greater, and the OpEx impact is relatively light. So that's another reason why I like this business.
Very helpful. So from a broader perspective, zooming out a bit, you talked about the model, you talked about the strategy and the customers. You've done a very good job historically about talking about your TAM and SAM, and both of which have kind of grown. What does this do to both of those? Does it expand your market? And do you think that your served addressable market is a bigger increase than your TAM? Meaning you go into a lot more areas than you couldn't have previously.
Absolutely. T here are the numbers, then there's the actual customers we go with. T he best part of this thing is that we go to the same customers.
We go to the customers that we're already talking to, and they've been underserved by some of our competitors because this is not the area of investment for them. M any of them are relieved to find us owning this part of the business, particularly in the area of networking, telecommunications, data center, which is a big, important piece of business for us, but also in automotive, also in aerospace and that portion of the market. Now, as far as when we look at the impact on the business, I think all of that gives us a very strong impact on the business. It just takes a little bit longer. When we did the deal, we had to do a creative deal.
We couldn't do a full acquisition because Aura had revenue coming out of entity ban list China companies. W e basically took the rest of the world for this business itself.
Helpful. Do you feel like this is the last acquisition that you need to complete the portfolio? If you look at just the three moving pieces that are in your industry, you cover them all now, so-
In this year?
In this, yes, in this area.
In this year.
In this year. E xcuse me. So do you think-
Just kidding.
Do you think that there's additional assets that is a must-have? Or do you think that there is a direction that you see this going, where you need something that's inorganic?
T he fact is that you're absolutely right. We did fill in a gap with 50 products to come, in our, in our business that, was the clocking piece. The other part of it is that we didn't get any revenue with it, so it's a slower, slower ramp to revenue. I f by chance we could get a clocking business that was already had revenue, that would be a pretty good deal for us, right? In the area of MEMS technology, we don't see anybody. W e do see that timing is becoming more important. People continue to buy timing as IP. There are people who sell timing their software, there are people who sell timing modules, there are people who sell timing systems. Some of these systems, for example, sell for $100,000, $200,000.
I'm not saying that SiTime is going into those tomorrow or even day after tomorrow. What I am saying is, SiTime is focused on timing. SiTime is focused on delivering the world's innovating, innovative timing products. T hat's what keeps my eye looking for, companies to buy or not buy or partner with.
Yeah. Super helpful. So yeah, I think we've covered the M&A portion early. I want to move to a topic that I think I've started most conversations with this week, which is just [inaudible]
Yes.
You've clearly seen this exacerbated cycle. You guys experienced it as everyone else did, but talking about a first recovery, which is a good sign. How are things tracking since earnings, and what areas of your business are you seeing that are turning on, that are leading to you being so confident about a recovery from here?
Right. T he general message for anybody, ourselves included internally, is that when a downturn happens, there's no place to hide. We, as management teams of companies, look for places to hide, but there isn't any, so we embraced that early on. Last year, we said we saw the downturn in consumer, then we started talking with all the businesses, and we saw that there was no running from that. We embraced that, called down on network telecommunications, on data center, on automotive. The only one which hasn't turned that much is aerospace and defense, but even that is not going down, but that's flattening out.
On the other hand, we called the up, which happened for us in Q3. The consumer was, you know, as you would expect, the up, and it's leading to some increases in data centers. It's leading some increases next year in electric vehicles, particularly in China. W e see all of this coming around the fact that we are coming our customers and specifically their CMs are digesting all inventory. We think that that is done by the middle of the year, coming year, 2024, and that we start to get normalized in Q3, Q4, and we're looking for 2025 onwards for the next two years to be significantly growth years for SiTime.
Do you think there's a difference in inventory levels between the CMs and your end customers? Is there worse behavior in a certain end market than another? I know that you had called out the consumer downtick prior to some of your other markets, but I thought what was very, you know, beneficial to you is that the areas in which you said that you were originally going to win back from the IPO started to recover, too.
That's right.
C an you just talk to what strengths are you seeing in those core markets, and are those the big customers coming back, or is that across the board?
T he consumer business is coming back. We see our biggest customer come back. We see some pickup in some of the other consumer products on the light industrial. I n general, I would say sometimes that those companies are, because they deliver similar margins, they pay greater attention to operational issues such as inventory. Sometimes companies that are consuming $5,000 chips and $20,000 chips, $40,000 chips, they are less focused on a $5 and $10 chip, and we know what kind of companies those are. O ur job now has begun not only to pay attention to the OEM, but also to the CM. Our top 50 OEMs represent 150 CMs represent the top 50 OEM.
W e're trying to get as high a level of clarity on an ongoing basis and trying to build reporting structures, not in fancy, just salespeople pushing hard and trying to figure out where CMs are, and we think that that's going to be the value.
One unique feature of this cycle, other than the rapid build of inventory at the end, was just the pricing. And I think companies, both because of wafer cost increases, as well as the ability to pass the cost of customers without deteriorating, saw a massive increase in both their gross margins as well as their ability to handle. Where is pricing today for you guys, versus where you started at the beginning of the cycle? And do you think that these levels of pricing are sustainable?
A s of Q3, we said that we have seen a steady growth in ASPs, and even in the results on Q3, we talked about ASPs that are holding or not declining or increasing. So one of the features of SiTime is that we're accessing a $10 billion market, and we have, as of now, a $2 to $2.5 billion SAM. A s of now, analysts such as you have it at $200 million per year. At that level, it allows us to have a huge playing field for us to pick and choose the differentiated business that needs our differentiated product, and the buy side of the aspect of that is that high-priced product. W e have started to lead with premium pricing since the pandemic.
We've found that not only do customers buy us because of quality or for performance reasons, and there's about 15 different reasons to buy us on performance reasons, but they also buy us on supply chain, and there's about six reasons which have to do with programmability, capacity, quality, reliability, and focus on that. W e're finding that 80% of our business is single source. You can imagine that gives us significant price stability. We're not gouging the price. We think our price premiums are fair price premiums for what we get.
I think that's something that's been unique, is the resilience of your margin profile. You guys are talking about a recovery back to that 60s range by the end of the year. How do you get confident in that? Is it just because you think or are there any other factors that help build that?
Well, some of it comes mechanically. As we get back on revenue, we get to pull back a few percentage points, about five percentage points with the, with the overlap, right? If you go back to $60 million a quarter, all of that expands. If you get back 100 points just through that. But then we do have the pricing. The other thing is the pricing mix has shifted, so we are shipping more and more higher performance points, which are likely to be higher priced. But we've also taken this opportunity to focus on cost. And even though we still have high cost wafers, particularly in the MEMS area, we are paying attention more to yield. We're paying attention more to spending minor CapEx for better enhancement of our product.
Then at the same time, you also are introducing new products, which I would imagine it's very easy to lift the price on these products as well. So can you talk about at least the ASP of a couple of new products that you've announced, and then can you give us the timeline for when you're gonna announce your first integrated product? I think you gave some timeline on that when you did the deal.
Right. So first of all, today we introduced the Endura Epoch product line. So Endura is our mil aerospace, especially hardened product lines, and in that, we introduced our highest price oscillator for the commercial market. In the commercial market, the Epoch product is for $30. So it wouldn't be hard to imagine that if military aerospace defense, that would be five to 10x higher price. So that's an example right there of the higher price. Another product that we introduced sometime back was the Cascade product line, which was our first actually Aura product. In a sense, we didn't get it openly, but we OEMed Aura clock, we integrated it with our oscillator, and we won about 10 design wins.
Sorry, excuse me, 100 design wins, and we hope to have solid revenue around the $10 million range for that in the coming year. So that, in fact, answers the second question about integration. Integration is not that hard to do. It's not that easy, but it's not that hard. W e are already selling this integrated product. W e have the one plus one equals three between the clocks and the oscillators already demonstrated.
I want to pivot to the competitive environment. Y ou've talked about the new competition that you're getting into with the acquisition, but historically, the quartz competitors had capacity issues during this period of time. S omeone comes to you and says, "Hey, I think that most of those customers would actually go back to cheaper options, at least in the low-end markets." How would you respond to that? Two, in your higher end markets that you've always outlined this core to your business, the differentiation from a technology perspective, why do you keep them using your product versus new potential?
The quartz issue of quartz supply becoming easier already happened in the H1 of 2022. We've now been for more than a year in the, in the regime of having easy supply of quartz crystal. Now, the side point on quartz crystal is that there's no such thing as easy supply. They still have 22-week lead times. In some cases, they have 35-week lead times, and the so-called ordinary lead times, which is 14 to 16 weeks. It's never a great supply situation in the world of quartz. But the point is that we've already experienced that, and in my opinion, we've hardly lost any revenue as a consequence of that.
Primarily because those who came to us at that time all appreciate the quality, reliability, the supply chain aspects of it, even if they didn't get the performance aspects of it. Now, pushing to the people who want the performance aspects, it's our job, it's our engineering job, it's our marketing job to deliver products of such exceptional value that, quote, "Customers just got to have it." So that's what we lead with, and if it takes us six years to develop, like it took us to develop the Epic product. So the six years to develop it, to get it to a level where on nine metrics, with quartz crystal, we beat them on eight. You know, that's the kind of product we want to win. I like to join battle where the battle is already won before the battle is joined, right?
We do that by defining the products early for winning.
Y ou talked about that $200 million mark, and you also talked about being selective in where you went with your product because you have that ability to leave the price. When you have that capability, you oftentimes can steer some of your revenue goes into markets you want to go, because you know, you have that luxury. When you look at your end markets into 2024, could you just walk through, you know, rank order or just give some comments on each, about where you see the most growth in getting to that $200 million range?
I think that data centers continue to do well for us. I think that even in the radio access network, there are certain markets, India, Turkey, the Balkans. T hese are small markets with small revenue, but they look like positive places for us. We see industrial, and in industrial there's robotics, there are farm equipment, that we see doing well for us. Consumer continues to do well, and automotive in particular, automotive out of China continues to do well, starts to do well, to be more accurate. Finally, the aerospace business, with Viasat or Kuiper Satellite, with SpaceX, where, you know, we think that that's good to be had.
So we see in the H2 of the year, in particular, we see a nice steady growth of pretty much every piece of business. And that's why we think we get to exit Q4 at quote, unquote, “normalized run rate.” Maybe not at the levels of 2022, but still very healthy, setting us up for 2025, because my eyes are on the 2025 prize. We want to do well in 2024, but not to be expected in 2025.
Is it just the rolling on of new products from the acquisitions you just did that add to the 2025?
The fact that we would have had two years with our own need to develop products, products like Epic, products like Elite RF, products like Elite X. They, they will have greater time to get the design win, and by the time they'll be in full flight.
Got it. I just want to dive a little more into the consumer end market, just because that market moved the most, and that market would also be subject to some of the customers that were most at risk to switch from quartz and then switch back, right? What do you think the right run rate for that business, and then with the largest customer that you have in that business, you guys have gone through a variety of different, content changes there. What are your expectations there?
Well, we think that our largest customer continues in the consumer products, where we add value in the watch, in some of the audible products, in the computing products. I think we reach a base level of, you know, keeping us around the 25% mark. I think that's a good number for us for next year. Consumer gets us to the 30% to 35%, depending on how things go.
In terms of the enterprise, Mil-Aero, et cetera, could you point out one particular area that you think as an end market is gonna offer the most growth for you? Because you've talked about content at those customers being very strong and the desire to have a higher end of technology. I f you could pick one area into the 25 time period, which you're the most excited about, you're bringing these new products to the table, what end market are you marrying them with? Is that across the board, or can you pick the first couple?
I would pick the data center, because whether it is in the rack or on top of the rack, whether it's optical, whether it's switching, whether it is in computing, whether it's acceleration, I think SiTime, particularly with the Aura products, is in an exceptional position for getting revenue out of those markets. So I think it'll do well. I think it will do really well for us.
Generally, with timing, you scale with the market as a whole, just because you have so much exposure to every business, right? When you look at a theme that we're talking about with all our companies deal with, is AI. When you assess what could that do for your company, do you have some specifics in what you think that could add to growth? Or do you just think broadly, A, more servers use, more successful?
Yeah. So more. That is, in fact, the right answer for now. More servers, more GPUs, more performance-based products, SiTime. W hat we haven't quite understood fully is what does AGI do? What is the level of synchronization? What does latency do in that business? And it's such early days that we've given ourselves about a couple of quarters to figure it out, but I just have a hunch that that's a big, big growth market for us. We just have to identify in what way and what time with which customers.
One thing that is more of a CFO question, but I'll give it to you because you're well-versed. Inventory levels on your own balance sheet ticked up during this period of time, and I think two of the questions on the call, you said most of that was with wafer. You wanted to make sure that you had wafer.
Yeah.
Could you talk about one, where is the target for inventory for your business? So where should you take perspective or even dollars, where should that go? And then other than that wafer stock, are you holding any other product there that is the problem, or is it really just [iaudible]
Yeah, it's almost all wafer stock, and in that, I would say 70% of that is particularly MEMS wafers, which have an eight to nine months l ead time a nd given that we are 80%, single source with some very critical customers, we think it's prudent to have inventory. Now, the other factor, because even based on that, I think it's a little bit higher for what I would like, but there are two other reasons for having done this. One is, we have a very reliable supply chain, and it pays to keep it running smoothly. It pays to keep Bosch and TSMC and all our back end guys running smoothly, keep our testers up. T o some extent, we overbuild just to keep the supply chain lubricated, as it were.
The second reason is more me, which is that I think that if we didn't understand what speed the market would go down at, I think we're not going to understand what speed the market will go up at.
Therefore, that's an opportunity for us because we'll have the product, we'll be able to satisfy demand, and I see that as a prudent use of resource. Having said that, I think where we are is pretty much of the absolute dollars, which should probably be about $10 million or less on a running basis.
We have a couple of minutes left. I wanted to open it to the crowd, if there was anything here, and I have a couple more. Something that's also coming up in terms of companies that saw pricing increase. So you clearly saw the ability for TSMC and others to raise pricing. That led to stock.
Yeah.
If you see, and I think there's been a resilience from the families to keep pricing high despite falling utilizations.
Yeah.
I think that many think the next step is some leniency on pricing. If you see your founder being more lenient on pricing, feel like you're in a position where you need to give cost break.
You know, when the cost of goods went up, and the cost specifically of TSMC wafers went up or Bosch wafers went up, we didn't pass it all on. We passed some of it on, and some of it [inaudible]. Therefore, when it goes away, it's unlikely that we're going to go back and offer anything substantial in price reduction as a consequence of that. The other part is that, as you rightly point out, some of these guys are sticking, these guys, the wafer foundries, are sticking with the prices, you know, because they, why not? They like that, and they can sort of get away with it, and, you know, foundry business is now the new oil business, right? So I, I think we're here to stay with this price increases. I don't see them.
I have one more on the acquisition. When you look at the earn-outs, can you give us a little color on what those are based on? Is it just standard revenue hurdles, or is it product hurdles? Any color there?
Yeah, so we have two kinds of, we call only one of them an earn-out, but the other one, $148 million plus $2 million SLA, $150 million, is also, in a way, not quite an earn-out, but it's based on performance. So it's based on delivery of products as they go. So that $150 million, for example, we paid $36 million because we got 20 products right away at the close on Friday. As the year progresses, we're gonna pay something at max, something like $70 to $75 million, if we get all the products that we're supposed to get in the remaining Q1 of 2025. So that's also an incentive kind of earn-out. But the earn-out earn-out is based on revenue that SiTime produces with those products.
Based on the goodness of those products, based on the goodness of the performance of the products, we think that'll be accepted. If there's accepted cost, we give them revenue, we give them the earn-out, and it's over five years, and it has a cap of $120. Most of it, because it's based on revenue, probably due to the last two years, 2028 and 2029. So probably almost nothing gets paid in 2024, and some smaller amounts get paid in 2025 and into 2030.
Very helpful. Sounds like exciting times ahead.
They are.
Thank you, Rajesh, for being with us.
Thank you.