Hello again, this is Melissa Fairbanks. I am the IT supply chain and analog semi analyst here at Raymond James. We are thrilled to have the team from SiTime here today. We've got the CEO, Rajesh, and we are pleased to announce Beth Howe, the brand new CFO is here as well. So I know they've got a pretty full day of meetings. Hopefully, a lot of you will have a chance to meet her. I guess to get started, if you just wanted to do a quick intro-
Yeah.
Otherwise, we can jump right into the questions.
Sure thing. Thank you so much, Melissa, for having us. Thanks for joining in. And this is the time one gives a safe harbor statement that, you know, there's forward-looking, et cetera, et cetera. You guys know the drill. So, just a reminder on all of that. Other than that, fire away.
All right. All right, sounds good. I think maybe it would be helpful to just start what you've been seeing recently in the demand trends over the past year or so. We did see some pretty strong corrections, you know, last year, finally found a bottom, for, for the revenue. It seems as though you're on a very strong upward trajectory now. Maybe talk about what caused the correction, as supply normalized, there was an inventory overhang that maybe, caught some of us off guard, and then what you're seeing in the near term as well.
Great. Well, as you know, SiTime went public about four years ago, around November. This—I was going to say this month, but we're already in the last month. So in November 2019, and it's been quite a ride.
Mm.
Within three months of us going public, there was the pandemic, which completely distorted supply chain. The demand went through the roof, the supply chain was weak, so SiTime made good use of that. We grew tremendously, well north of 50%, and most of the demand was fantastic. It came from new customers, it came from people who didn't know that we existed, it came from people who were desperate for product, and because of our unique supply chain versus the incumbent technology, we were able to satisfy demand. It's also true that customers, towards the end, as we know now, overordered, they either miscalculated or they deliberately overordered because they weren't sure when and how they would need the product.
Long story short, along with the overordering, sometime in middle of last year, the demand started to taper down. The lower demand started with the consumer side of the market because we have a very broad, diversified business. Started with consumer, then went to industrial, then went to networking, telecommunications, automotive-
Mm.
So nothing was untouched. So as a consequence, Q2 this year, 2023, was our lowest quarter. And Q3, we guided up, which we did well with that. We grew by 26 percentage points over Q2, and we've guided to around anywhere from 12-15 points up in Q4. So that is the opposite of what happened, which is the inventory is slowly being consumed. The demand is moderately up, particularly in consumer and light industrials. So it's almost the exact opposite of what we saw, and as a consequence, we expect that next year, as the inventory goes through in the first half of the year, things will get better.
Mm.
Then the second half, they'll accelerate because not only will the inventory be consumed, but the demand will have picked up. And then just looking out into 2025, I think that 2025 will be a very, very solid year for us.
Great. Yeah, I think, consensus estimates looking at 2024 and 2025 are some for some pretty, pretty strong growth and-
Yes
... and profitability. So, good to see, it seems safe to assume that the inventory overhang is largely corrected. Is that a safe assumption to make, or are there still some areas where you've got-
Yeah-
-pop?
... I think we have, when we take a look at our top 50 customers, we see their inventory declining, but we don't, the excess inventory now-
Mm-hmm
... but we don't see it fully consumed till anywhere between Q1, Q2.
Okay.
Let's say by the first half of the year.
Yep. Gross margin dipped to just over 58% over the past two quarters. I know this is a little, a little bit more detail than we normally go into, but now they're expected to trend modestly higher going forward. How should we be thinking about what a normalized margin is? You know, we are seeing higher ASP products that you're shipping, and that's probably some of the fastest growth areas.
Right.
How is that going to impact the margin profile going forward?
Right. So, first of all, I have to say that, you know, even though I'm gonna say I'm impressed by my own company, but I was impressed with the fact that while our revenue declined substantially, went from $285 last year to consensus has it around $145-$146 this year, our gross margins only dropped from 65 points down to 58-59. So just, you know, 5-7 points.
Remarkable, for sure.
So, and the reason I'm happy with that is because it points to the pricing power-
Mm-hmm.
-the uniqueness, the value that SiTime brings, which I think is unique in this. So it rarely happens, that scale drops so much-
Mm-hmm
...and margins relatively less. So in general, 65 points of margin, 30 points of net profit, that is our model.
Mm-hmm.
We were at that for several quarters in the last 3-4 years. We expect to get back into that certainly by 2025, and perhaps even by maybe the third-fourth quarter of next year.
Mm-hmm.
Certainly on the gross margin side. I like 65 points of margin because it is an exceptionally high margin, but it's not one, which is so high that we lose our degrees of freedom. I've seen other companies, mostly analog semiconductor companies, at 75 points of margin, and then not being able to do anything with that. So 65 is a good place to be, and I think that is our long-term model.
Okay. Now might be a good time to address the Aura acquisition. I was happy to see it closed on Friday. I think the announcement came through.
That's right.
Maybe talk about just review the acquisition, what Aura brings in terms of your ability to offer more fully integrated, more complete timing solutions, and how that is going to drive some of that longer term growth and margin.
Right. So Aura Semiconductor is a company that where mostly the engineers are all out of Bangalore, India. And when we went public four years ago, we said that there's three parts of the timing market, and SiTime is a company that's focused on all aspects of the timing market. So we started with oscillators, and even now, about 90%-95% of our revenue comes from oscillators. That is, out of the $10 billion market, that's about a $4 billion piece. The resonators, which is another $3-$4 billion piece, is also one where we have the know-how, the technology, the ability to enter that market. It's just that we choose to do it at a later date.
The part that is been always the most one that we have had our eyes on is the clocking piece, which is a relatively small piece, about $1 billion-$1.5 billion. The clocking part is very important to us because it sits right next to the oscillator technology and sits right next to the customer use case of oscillators. Now, we can do clocks ourselves. In fact, we are doing clocks ourselves, but what Aura Semiconductor deal does is it gives us 50 products over the next three years, 20 products right away, another 20 in the first half of the year. It gives us 40 products between now and June, and immediately connects us deeper with our customers in the networking, telecommunications, data center part of the market.
Those customers have the oscillator product sitting, and right next to it on the board is the clocking part. I'm thrilled because it's just the right product at the right time, and the single best, you know, deal we could have done, at the price we did, which I think is a very, very fair price.
It seems as though this really accelerates your roadmap into clocking-
Right
... and probably brings along some incremental revenue. You discussed it briefly on the call, on the earnings call, and I think there was a clarification about what the revenue contribution over time would be, that you know, it's going to take some time to ramp.
Right. So there's two values to the deal. One is, of course, the revenue, and as you point out, in the market that we're talking about, the data center market, the networking market, the core, the radio access market, design wins take a good 18 months or so. So the design wins are going to take us some time, and then the deployments take some time. The good news is, though, that the ASPs are high, the gross margins are high, so I think this is going to pan out for the long term. And we've said that in the next few years, we think we can get to $100 million at 70 points of margin.
Wow!
... so slightly higher than our average in this business. But the other value, which is a little bit more subtle and we haven't been able to put a number to it, is that we get drag along into our oscillator products and our resonator products. Another value which we haven't been able to put a number to, is that clocking is very architectural in the customer's systems. So whether it's a Google, whether it's a Meta, whether it's a Nokia, it gives us the abilities to have an architectural to meet a conversation at a higher level sooner than it did with some of our oscillator products. So I believe it provides quite a bit of value. I just, I'm not able to put a dollar value to it, but it's pretty significant. I think that does that.
In any case, we've become the only company with this acquisition to have resonators, oscillators, and clocks.
Mm-hmm.
Uniquely, so we don't see anybody else in the world that can do that and has done that. So we think we'll continue this very focused approach on timing in a $10 billion market, which gets to be a $20 billion market in about a decade.
So I think that actually brings up a very interesting point. You know, since the IPO, at the time of the IPO, you know, you really weren't seeing any substantial competition. You know, you had a several generation lead for some of the MEMS-based timing solutions. In the several years since the IPO, we really haven't seen anyone substantial emerge yet. What are you seeing in terms of the competitive dynamics? You did mention that Aura helps to kind of extend this leadership position, but is there anyone else yet on the horizon?
So first of all, I am pretty paranoid about watching out for competition and checking out competition. I never take it for granted that we are in a great place and rest on those laurels. Having said that, I'm also a big believer in identifying what is the best game to play. Winning teams define their strategy and then execute to that strategy, and keep on checking on whether that strategy is, in fact, the winning strategy. So far, the strategy of SiTime is to be focused on MEMS, focused on precision timing, focused on timing, focused on all aspects of timing, focused on the high-value part of timing, focused on premium price timing. That has been paying off. The fact that there has not been any competition is just because MEMS-based timing, precision timing, is hard to do.
Silicon Labs tried it, Maxim tried it, IDT tried it, and five startups tried it, and they all bailed after having spent a collective $400 million, $300 million by around 2013, 2014. So I do know that large company Murata, large company Kyocera, and one startup in near Montreal, Toronto, are trying-
Mm-hmm
... to come into this business. I still think it's hard. MEMS technology is just hard to do. SiTime is a leader in MEMS-based timing, but if you look at leaders in other MEMS technologies, MEMS in display, that's Texas Instruments, they've been leaders in it for the last 30 years. MEMS in RF filters, Avago, before that, Broadcom, before that Avago, before that. HP has been a leader in 30, 40 years in MEMS-based filters, and there's nobody really competing with them. I think the same thing applies to timing. SiTime has been a leader in timing for the last 10, 12, 15 years, and I believe we'll continue to be a leader going forward. But even if there is competition, as I said, it's a $10 billion market.
Even when we are $1 billion in revenue, and the market is $15 billion, we'll still be a very small portion of the market. So our goal is to be a player like Apple in the, in the smartphone, 15%, 10%, 20% market share, and 80% of the value, 80% of the profits coming to the company. I like that model.
I think that, that kind of leads us into some of, some of the end markets. You did mention, you know, precision computing. We've talked about auto in the past. In the December quarter, you're seeing substantial growth in comms and enterprise in 4Q, and I think this is some of the opportunities that you touched on with the computing and networking. It's still down about 50% for the year, though. So, what are the expectations for the longer-term CAGR? You know-
Yeah
... that was one of the earliest inventory corrections I think that we saw.
Yeah. So while I'll take the big jump in percentage terms, it's off a low base. So just want to say that it's modest in dollar amount, but it's substantial in percentage amounts. Having said that, the biggest market we would like to win in is the networking, telecommunications, data center market. With the oscillator business, it's about $800 million. With the clocking put into it, it's another $600-$700 million, for a total of $1.5 billion. And so, SiTime has about 150-200 design win opportunities in that space, and we intend to be in every one of them. So we are highly dependent right now on optical modules, we're highly dependent on servers in the acceleration NIC cards and use cases.
Where we haven't quite really gotten to, because it's early days, is in the AGI, in the generative AI space, which is, of course, where all the biggest hype is.
Mm-hmm.
But in the meanwhile, we think that we have, whether it's top-of-rack switches or it's ML cards, or it's servers, or inside the server, I think we have a huge number of opportunities with the clocking piece as well as the oscillator piece.
Excellent. Maybe moving on to auto, industrial, and aerospace. It's also been down substantially this year, but expected to be a meaningful driver for longer term growth. And I think, you know, you've talked in the past, specifically in auto, your lead customer there, where precision timing is somewhat mission critical now, if we get into some of the more autonomous driving or some of the ADAS systems.
Yeah.
Maybe talk about what you're seeing there.
Yeah. So it's instructive that four years ago, when we went public, people said: "So what's your automotive business like, potential?" And we said: "We don't see any.
'Cause we didn't, 'cause ADAS wasn't a thing, and one of the best things that happened in the pandemic for us was the pandemic accelerated electric vehicles and ADAS and all of that. So now we forecast that in a Level 4 car, which is probably coming out, I don't know, 2026, onwards, SiTime has a potential for $50 per car.
Wow!
$50 of timing per car, and it's driven by, as you said, the ADAS, the automotive driving system, the processing, the Radar, the LiDAR, the cameras, the domain controllers, the display of all this information. I think all of that, and there's some of it coming from the electric vehicle management, power management, piece of it. So I think that's an astonishingly high piece, $50, for SiTime in that period. Yeah, I'm very excited.
I'm gonna pause here because I know that this is an area that a lot of investors have been interested in. Does anyone have any questions before we move on? Okay.
In the area of cars, I will say that apart from, you know, large companies that we've talked about, like Tesla, and Rivian, not that large, Rivian, but still important. China EVs are a source of revenue for SiTime, whether it's BYD, whether it's Geely, and a whole bunch of other names. I think I'm quite happy to see that we continue to get very good traction. And we here in the U.S. don't get to physically see these cars, but if you were to see these, some of these cars, they're really quite impressive. And frankly, one could argue that they're even more impressive than some of the cars that we do see. They really are quite impressive cars.
You know, I wouldn't be surprised if China really goes screaming right past Germany and Japan, and certainly, you know. And I think the U.S. is under some threat, even though GM and Ford are sort of rallying pretty well, but their cars are really good.
It's great that you keep expanding these opportunities and growing into new OEMs. I think that certainly helps-
Yeah
... with the durability of the business.
Yeah.
Is it safe to assume that once you get the design win, barring any sort of quality issues, you're probably locked in for several years?
Yeah, never-
Give you that much visibility.
... never assume anything.
Yeah.
Right. But I think particularly in the automotive market, the military aerospace defense market, which also is one we, we have about 350 design wins, and the networking data center piece, you can imagine we go for those markets, because they're valuable markets. But there's another point to be made here, and that is, 15 years ago, when I started with the company, when it was pre-revenue, timing was not that important as a category because systems didn't want timing, of that precision level. What has happened in the 15 years, and I dare say in the next 15 years, is that timing as a category becomes more and more and more important. And so for customers, the fact that SiTime becomes a one-stop shop for all aspects of high-end timing, whether it's in consumer...
By the way, there's a watch company in China called COROS, that's also using us, along with some of the better-known watch brands here in the U.S., like Garmin and another one.
Mm.
So, SiTime just believes that if you believe in AI, if you believe in automated driving, if you believe in a connected world, if you believe in IoT, precision timing is a category that just comes along with it. So, we're happy to be this highly diversified business that sees the growth of a company like AMD or NVIDIA, or a growth of a company in like Qualcomm or Broadcom, and says: Well, wherever those chips go, so do SiTime chips.
I guess that leads to the question about your capacity planning.
Yeah.
You know, I think you've had the benefit, and this may be part of what's driven your success. You've got the industry leaders in the silicon and industry leader in the MEMS supply. As you're getting some of these, you know, $50 in content per vehicle is a very significant incremental step up in content, how are you working with your suppliers to kind of ensure that you'll have the ability to service all of these opportunities?
Right. So first of all, we, as you said rightly, we have TSMC, the world's, you know-
Leader in foundry, and we have Bosch in Germany, which is not at all a foundry, it's a merchant fab, that just happens to be providing only for SiTime, the products. So starting with Bosch first, Bosch was an early investor in SiTime, and we built up a long-term relationship with them. Of course, it's our technology, it's our process. It's not something invented by them, it's our process, and long before 10, 15 years ago, we used to use Tower Jazz, and before that, we used another fab. So we've been able to port this around, and we choose to bring it to Bosch... because, as you said, they're the leader in MEMS, in accelerometers, and gyros, completely different MEMS.
But we also have this great relationship, and frankly, we get to protect our IP, because there's a lot of know-how in IP and MEMS, and instead of taking it to a merchant fab, where people, engineers leave and go, we would like to be at a captive fab. Maybe we pay a little bit more money, which we do. We, the wafers aren't cheap, they're, you know, more expensive than normal, but we get to keep our technology secure. On the TSMC side, it's very easy. It's the third time that I've been single sourced with TSMC. My previous company was also single sourced, and that works for Broadcom, that works for Qualcomm, that works for NVIDIA, it works for, you know, Apple. So, those guys are one of the best.
They invented the model for fab, and they have a great relationship with them. So I feel very, very confident. But just in case things go bump, we also carry a rather large amount of inventory. We have thousands of wafers in MEMS. We have thousands of wafers on the analog side from TSMC, so that given that our customers are single source with us, 80% of them, 80% of our revenue is single source. So to give customers like car companies and Nokias and others the comfort, we keep these wafers just in case something goes wrong. And also, we couldn't call the down off the bottom.
We're certainly not going to be able to call the big snapback whenever it happens, so it's prudent to be ready with the one company which has lots of capacity so that we can capture the upside when it comes, which we know it'll come.
And there's limited risk of obsolescence with-
There's zero risk.
With this inventory.
Yeah.
Okay.
Yeah.
Excellent.
It's just basically these. First of all, we have never end-of-lifed a product, so we've never said this product is no more needed. So all our products have so far had 15 years of life, so these products will be secure.
Okay. All right. Maybe with just a few minutes left, after the Aura deal, you know, the transaction is set up, you know, to be kind of like an earn-out situation.
Yeah.
So it's not a big upfront cash outlay. I think SiTime is fairly remarkable in having a very sizable cash balance and actually earning interest unlike most of the other companies that I cover. What are some of the priorities that you think about and you've been working with Beth maybe and talking about?
Well, first of all, as you rightly pointed out, before the acquisition, we had $575 million in cash and 0 debt, so that puts out a lot of money, but I don't expect the investors to recognize that source of money, because you guys can go out and buy Treasury bills just like we do. So our job is to make that work for you in exceptional ways. We have done that. This whole deal, if we pay out every bit of the earn-out, being very conservative, would be worth $270 million. Even after that, we'd have a significant amount of cash. First and foremost, we want to just digest this. We want this to be our first acquisition. As we all know, acquisitions, nine out of 10, 10 of them fail.
We don't want to be in the nine, we want to be in the one. And so we want to make that work. Second, we want to use the money so that it can provide exceptional leverage. So we'll see. Yeah.
Are there any other technologies or areas of investment that, you know, potentially you could find another engineering team or another product portfolio elsewhere and kind of accelerate? Or do you think that this current portfolio is what your customers are looking for for the next several years?
Yeah. So these customers are probably well-served, but, you know, timing is a very, very important part of the of running the world. GPS, for example, which brought in about $1 trillion so far of GDP, runs on exceptionally good timing. So I can see that there's a value in selling SiTime and selling timing boxes. Sometimes... I'm not saying we're going there, I'm just giving you, these boxes can go for hundreds of thousands of dollars. There's value in selling timing IP. Again, I'm not saying we're doing it, but there's, it's a possibility. There's timing software, there's timing as a service, there's timing at the very, very, very precise atomic clock level. So I think that a company that decides to own timing can only do well. So we'll see where it takes us.
Okay. CapEx intensity has kind of stepped up recently, mainly just because the revenue was lower. Should we expect your CapEx to return back to that low double-digit range as the revenue grows?
I think so.
Okay.
I think we probably overbuilt a little bit-
Mm-hmm
... in 2022 and early 2021. I mean, to early 2023, but mostly in 2022. Right now, our goal, both on CapEx and OpEx, is to keep it flat-
Okay
... between 2023 and 2024. We don't see a need to particularly grow.
Okay. We have just about a minute. Any closing thoughts or quick questions anyone wants to sneak in?
Well, I'm really pleased to be here. Love to meet as many of you as is possible. Also, I think that it's very early innings for SiTime. I've been in the technology business in various forms, mostly in semiconductors, for forty years, since 1984.