All right. Good morning, everyone. Welcome to the 21st Needham Technology Media and Consumer Conference. With me here right now is Wally Rhines, Chief Executive Officer of Silvaco, and Chris Zegarelli, Chief Financial Officer of the company. Maybe let's just dive in. Walden, thanks for joining us today. I think we did a fireside chat not so long ago in January at Needham Growth Conference in New York City, N.Y. It's been about four months since then and, you know, Silvaco has delivered two beat and raise earnings reports. Fair to say a lot has changed about Silvaco.
Before we get into Silvaco discussion, I would like to zoom out a little bit and ask you where you think the EDA industry is at today, since you're still involved with the ESD Alliance, the EDA Industry Association, which reports EDA market statistics on a basically quarterly basis. Wally, what's the most incremental thing happening in the EDA industry since January, from your perspective?
It's another exciting year for EDA. You know, we came through this decade where EDA was propelled by non-traditional customers. That is people who hadn't traditionally designed integrated circuits. Google, Facebook, Amazon, Alibaba, all these companies entered the silicon chip design business, and that added a substantial amount of revenue from customers who were less price sensitive than the traditional semiconductor companies that had been the EDA buyers. What's happened since then? Well, right now it's all AI. It's each of the traditional tools for design are being examined in terms of how AI can change the way they're used, the results they produce, how they can be a co-partner in design, and a variety of other capabilities.
Each of the major EDA companies is introducing products first to revamp their product line, add capabilities, add accuracy, add better performance, and then secondarily, to explore new areas that EDA has not traditionally covered and to introduce new products. That's led to a sustainable double-digit growth, although now it's low double digits, but still, the EDA industry has been a great place to be the last 15 years.
Let me ask you, since you brought up sustainable low double-digit growth for EDA. I remember back maybe around 2022, back then you were not the CEO of Silvaco, right? You were semi-retired. Is that the right way to characterize that period of time?
No, I was a full-time CEO for a fabless company at that time, but I was out of EDA for five years.
Yeah, semi-retired from EDA, let's put it that more precisely. Back then when EDA industry was probably growing at high teens, that was a pretty good couple years, right? During the COVID years. I invited you to speak to Needham institutional clients in a virtual conference call. Back then you said something like you think EDA would decelerate a little bit because high teens kind of was not, you thought that was probably not very sustainable. It seems that you were probably right all along. There was a few years since then for EDA companies, large EDA companies, their top line growth decelerated a little bit. Yes, still in the low teens, but they decelerated definitely from the high teens.
What's your latest thought, the sustainable EDA growth sounds like it's low, low double digit, kinda can you elaborate a little bit and do you think Silvaco can outperform? That's the first question very specific to Silvaco.
This is great to have a leading analyst say that I was right. It doesn't happen often. Let me trace why the growth rates and what the future is for Silvaco. First of all, the reason it popped up to the mid to high double digits was the entry of this new set of designers I mentioned before. All of a sudden, you had new people coming in. They had no tools. They had to buy the whole tool set, so an enormous bump to revenue, and they constituted about 25% of the revenue during that period.
The reason I said at the time is this can't be sustained is eventually they get a full tool set, and then they're simply upgrading it and renewing it year after year, so it settles down into a little more stability, which is why I was predicting it would go down to low double digits. Traditionally, if we go back, EDA was a business that grew about 1 point faster than the semiconductor industry. The semiconductor industry, up until the early 2000s, was really a 5% grower over a long period of time. It was not a double-digit growth industry. It's more than double digits. It's the real star. I don't open The Wall Street Journal any day without reading about all my friends in the semiconductor industry.
We're, we're now at a point where we don't have new people coming in the market, and the problem, the reason I believe that the overall industry drops back to low double digits or even high single digits, is because budgets are somewhat fixed. People in the semiconductor industry have a fixed amount of R&D they spend. Over the last 40 years, it's averaged 14% of revenue. Just because you have a new and great tool to offer, that budget doesn't expand. Something else has to be squeezed out. The total amount can't grow that fast. Now, it has grown fast recently. As you mentioned, I handle the statistics for ESDA, and we're now up to a $22 billion run rate. That's phenomenal.
We went a long time in low double digits for the total. The total industry was a $10 billion industry for quite some time, $10 billion to $12 billion. What about the future, and what about Silvaco? For the future, I think, because of the renewals and the new customers and so on, it'll stay high single, low double digits. I think Silvaco will do better than the industry. Why? Several reasons. One, Silvaco's growing from a smaller base, it's fairly easy to have higher growth percentage growth rates, but that's just an arithmetic thing. The real reason is that Silvaco's in new and emerging markets where there is no incumbent, particularly AI-driven manufacturing, where the sky is the limit.
All of the industry will move to AI-driven manufacturing process development, yield improvement, and all those things today are done with a very small amount of EDA, basically TCAD for $200 million a year. This will become a much, much bigger market. Silvaco is at the leading edge of it. We have a teacher customer, Micron, that has helped us develop the capability, and that's the biggest single reason I believe Silvaco will grow at a faster rate. As we'll get into later, our IP business is growing at a very high growth rate as well, so it gives us a lot of confidence in being a double-digit grower.
Thanks, Wally. Let's dive into more details. You have three product lines, right? TCAD, EDA, and IP. TCAD, as you mentioned, $200 million per year, but it's a small within EDA, but it's the largest product line for Silvaco. A lot of folks here probably are not as familiar with TCAD and probably wonder why it's a separate category from EDA. Can you explain to us what TCAD does, why it's different from, let's say, the traditional EDA, and why Silvaco, a company with a much smaller scale than, let's say, Synopsys, the main TCAD competitor for you, can survive for this long and hopefully can do very well going forward?
Basically Silvaco created the TCAD industry starting 40 years ago, and developing tools that could simulate processes by basically solving partial differential equations. The computation was very complex, very lengthy, as typically, if you wanted to do a simulation to predict, the concentration of diffusing elements or an oxide thickness or something, it could take 24 hours to do a one single run. That's the way processes were developed 40 years ago, and more recently even. You developed a model, typically these were BSIM models, and that model was the basis that a designer used to test or to design a particular transistor with a new process. That was not an attractive market for most of EDA. It was attractive for Silvaco.
For most of EDA, they concentrated on designers, because there are probably 100 or maybe even 1,000 designers for every process development engineer, so there weren't that many people to sell to. Synopsys came into the business totally by accident. They acquired Avanti had acquired TMA, and, you know, Aart de Geus was fairly dismissive of the whole TCAD market for a long time because it was a small number of customers and not that much revenue compared to EDA for designers. They did a brilliant thing. They acquired the rest of the competition, other than Silvaco, and consolidated the industry, so now it's an industry that is just Synopsys and Silvaco. Because they acquired more companies, they're bigger than Silvaco.
Basically, the historical stability of a customer base has allowed Silvaco to continue to maintain and grow that business throughout its history. It's still growing at healthy, single-digit kinds of growth rates, and continues to be used by the largest base of customers. It won't be replaced by the next generation, but the leading edge will move and adopt the next generation development, which is, we refer to as FTCO, but it's basically AI-driven process development. TCAD will continue for decades because not everybody has to go to the next generation of process development. Basically, it's a great opportunity for companies like Silvaco to lead the way in the next generation.
It's attractive also because it's not an enormous market. The, the larger competitors really don't show that much interest 'cause it doesn't have that much effect on their business. For Silvaco, a company that's in the $60 million-$70 million range, it's a big impact. It can grow our revenue rapidly, it can be very profitable. It's something where there's limited competition and enormous leverage with customers. Customers, if they can save a few months in process development, that's a few months that a $30 billion capital investment is sitting around waiting to ramp up its wafers. If they can use less prototyping wafers, that's an enormous cost, a cycle time advantage. Big leverage, small business, big business for Silvaco.
Thanks. That, that's a very good way, I mean, to segue into the next question. You talk about, you know, there's a lots of stuff happening in the not so advanced process nodes. I think one of the things, right, that's the recent development in AI data center really is the power semis, right, gallium nitride, silicon carbide power devices, optical networking, right? There's the silicon photonics, and there are lasers built on some pretty exotic compound semiconductor materials like indium phosphide. Maybe this, let me direct this question to Chris. Chris, I looked up your LinkedIn.
You know, you were the CFO of GaN Systems and then joined Infineon when GaN Systems was acquired by Infineon. I think this is probably your wheelhouse in some way. You know, all those areas I just mentioned, right, power devices, optical networking, gallium nitride, silicon carbide, indium phosphide, these areas feel like what Silvaco actually can participate in in some meaningful way and should be the sweet spot for Silvaco. Am I looking at this the right way, and do you see any strong pull from those kind of customers?
Hey, Charles. Great question. Thank you for looking back into my background. That's exactly right. I was for years CFO with GaN Systems, stayed through the acquisition by Infineon. I can tell you, yes, we have really good exposure here in Silvaco to power semis. It's a meaningful percentage of the business, mostly directed at the kind of GaN and silicon carbide players. I'll tell you, Charles, what's interesting, the first time I really heard about Silvaco, I mean, you know me, 25 years of semiconductor hardware finance experience, you know, mostly on that side of it. It was with the GaN Systems developers that I heard about, a lot about the Silvaco tools, 'cause they have a good reputation for managing and modeling GaN, you know, better than the competition.
I'll tell you, there are a lot of unique things about GaN, for example, that make it, like, ideal for simulation and modeling. There are lattice mismatches, there are thermal mismatches. You know, getting the epi layer just right is crucial, and I was surprised to see just how many development wafers are constantly run to model different iterations on process, different thicknesses of epi, just different growth models. If that can be done virtually, you speed it up, you save money, we have good exposure to that. I would also comment on the photonic side, you know, we have exposure through our Tech-X acquisition as well. In my GaN Systems days, we were always excited about the industry moving from silicon wide bandgap, and GaN in particular, we are very bullish on the future of GaN.
I'll tell you, as more and more designers move from legacy silicon to wide bandgap, that is a tailwind to our business, 'cause the more development design licenses that you need from a Silvaco to design the next generation of GaN, and there's meaningful savings and efficiencies you get from modeling it more accurately, more correctly, and has just been investing in a lot of optimized tools specific to bandgap, in order to deliver differentiated value to customers. I agree with you, it is a nice tailwind to the business, something that'll play out over time, and I'm excited to see that progress.
Yeah, I would add that, Charles, that investors get excited when they see things connected to the data center, and they don't always make that connection. I can tell you, photonics is taking off like a rocket ship because of the infra data center communications for photonics, and wide bandgap semiconductors are taking off at a very rapid growth rate because they offer lower power and high performance in a whole variety of applications. I think, to some extent, these two are in the sweet spot of where the high growth is occurring today, in addition to, of course, digital graphics, processors, and things like that are the heart of the data center.
Great. That's a very good color. Let's talk more about the products going from here. Let's talk about FTCO. That's the AI product, like infusing AI into your own product, right? That's the idea. Since I know you for not too long, Wally, but long enough, I recall you were AI skeptic going back to 2021 to 2022 when the large EDA companies were really spinning the AI narrative, pretty, I mean, pretty intensively back then, I think. That was kind of like a narrative inflection point from many of the large EDA companies, with regarding AI. Now, I mean, as the CEO of Silvaco, you are also selling AI products to customers like FTCO.
Can you tell folks in the audience what FTCO is? Why is it AI? Why is it gaining tractions with your customers? I recall you have three already, right?
Yeah
Why should we expect more deals, FTCO deals to happen going forward?
In the 10 years ago, 15 years ago, AI was being applied to EDA, but basically, it was just making our existing products better. It wasn't expanding the market. It was meaning that we provided our customers with greater capability, but their budgets still remain, you know, basically a 2% of revenue spent on EDA. It wasn't creating the market growth. The things you need to look for are where are there opportunities for new AI capabilities applied to EDA, where it taps into a new budget. I mentioned earlier the Google, Facebook, Amazon coming in, that was a new budget that allowed the industry to grow. The semiconductor manufacturing is a totally different budget from the budget that is used for design EDA. It comes from the manufacturing group.
These are manufacturing people who think nothing of spending billions of dollars on capital every year. If you can make that capital more efficient, then the cost of EDA, TCAD, FTCO is a very minimal expense and has very high leverage. To me, this is an area where a lot of new dollars can come into EDA as we move from basically physical prototyping to virtual development of processes, doing it all through simulation. FTCO basically simulates each of the major process steps. It allows the process integration engineer to put them together, to test them, to see what the capabilities are, and not just electrical capabilities, thermal effects, stress, other things, and engineer the next generation process to optimize and pick the best mix of features.
On top of that, it also brings in a whole new set of customers, the fab engineers, the test engineers, others who wanna examine faults, and they see a variation in the process, they wanna analyze what should that variation do to the actual results that I see on the tester. So it expands far beyond what TCAD did with just process development engineers who are very specialized, not very many of them. Now we're going to a much larger market. We have much more leverage on the dollars. It's a real opportunity for AI to increase the total revenue of the EDA industry.
Got it. I think, Wally, that's very insightful. You said, we should probably always focus on the new budget. Does something unlock new budget, right? Loud and clear, you are unlocking the new budget from the manufacturers. Now let me ask you a question on the IP business. You talked a little bit about IP, you know, when we were speaking about silicon photonics. IP is expected to be your best product line in terms of growth rate for this year. We know some of that is inorganic, but still, I mean, you have sounded quite bullish about the overall IP growth.
You're very fond of the mix of IP that Silvaco acquired before you joined as a CEO. Can you tell us why, and explain to us once again, how can Silvaco's IP business differentiate, and how can it grow maybe at the market growth rate or above that, and more importantly, profitably?
Yeah. It's, you're right. I view that as a very exciting near-term propellant of growth for Silvaco. Several reasons. One is quality is king. You cannot afford to take out a chip if it's gonna have bugs in the IP that you used in that chip. Mixel has probably the best reputation in the industry for never, ever having had a customer bug, that or a bug in a customer's IP. They are highly respected. The existing IP business was relatively small at Silvaco, but as we put the teams together, there's a lot of shared learning that has improved our efficiencies, and we've effectively doubled our capacity just through those efficiencies. On top of that, we had, at Mixel, we had one salesperson producing all of their revenue.
We have a whole sales force at Silvaco that's excited. They've got a great product. They're highly respected. It's a very specialized market, about a $300 million market. Clearly Mixel is the king of customized IP, and we're selling more of standard IP as well. It would not surprise me at all to have our total revenue for IP double in the current year.
Great. Thanks. Maybe some near-term P&L dynamics, question for Chris. You guys spoke about cutting costs, increasing gross margin, reducing OpEx, and what has been done so far. It definitely has shown up in the results, right? The gross margin expanding, OpEx is coming down. What should investors expect through the balance of the year? Any thoughts on maybe a longer term, how we should think about all those P&L metrics, Chris?
Sorry. Great question, Charles. Thanks so much. You're right. When Wally and I joined the company kind of in late Q3 of last year, it was clear to us, you know, the total exceeded revenue, so we needed to get spending in line, we needed to focus on margin expansion and growth. I'll tell you know, we think about spending as total spending, so combining operating expenses with cost of sales, because in our EDA business, most of cost of sales is really the cost of our people and colleagues supporting customers. If you look at that metric, total spending, it was down 9% sequentially in Q4, down about 6% again in Q1, we're guiding for that to continue coming down in Q2. We've made great strides in getting cost reduction in place.
Most, we announced a $20 million cost reduction program. First it was 15, we upped it to 20 in Q4. Most of that has been implemented. We still do have costs though, Charles. Some of that just takes some time. Looking forward, there will continue to be cost that comes out. We are making some targeted investments in things like AI tools to accelerate internal development. On the spending side, it's come down a lot. I expect it to be kind of flattish or down a little bit going forward based on those dynamics. What's encouraging, Charles, is our guidance into Q2 implies non-GAAP operating profitability being attained in Q2, which is a great milestone and consistent with the commitments that Wally and I talked to when we joined.
Looking forward, though, Charles, to Wally's point, it's really a focus on growth. That's really where we're gonna get leverage in the model, so we're gonna maintain discipline on spending. Wally talked about IP. It's been a great grower of 270% year-on-year in Q1. As Wally said, we see that momentum continuing. Andy, who runs that business, is really focused on growing our pipeline, increasing efficiency, and positioning us for growth. I'll tell you, the pipeline there, you know, doubled in the last year. That's just on the organic side. We're seeing a lot of opportunity to expand IP. TCAD's layering more FTCO under TCAD, but that business is, as well, up 22% year-over-year.
It's really up almost 50% in the last six months from a revenue perspective, and that's driven by this FTCO momentum that we've been talking about as well. Those two key growth drivers should propel top line. We maintain discipline on OpEx, and we are focused on reaching profitability in Q2 and building from there going forward. A lot of progress in a short amount of time, and I'm looking forward to continue building on that momentum.
Thanks, Chris. We have a little bit more time, but I do wanna make sure we have time to ask Wally. The last question I prepared here, Wally, you were the head of TI Semiconductor division back in 1980s. I've read books about how you got into the DSP business or kinds of good stuff. Then you went to Mentor Graphics in 1990s and led the turnaround, served as a CEO through mid-2010s until, like, the company got acquired by Siemens, right? Well, you went a different place. I was writing you went semi-retired for a few years, but, you know, you were in different, a different part of the semiconductor ecosystem.
You were in the fabless world for a while. Now you're back in the EDA game, you're leading Silvaco, but it feels like, you know, you have a wealth of knowledge and decades of experience. Well, that is being channeled into one small company called Silvaco. I heard you do have a, quote-unquote, "term limit." I don't know if that's the right word to call it, but, what legacy do you plan to pass along to the next Silvaco leadership once this turnaround is accomplished over the next few years?
Yeah. Okay. Well, you described it well. Turnarounds are a lot of fun, and, while I don't have any real term limit, I think what you're saying is, look, I had over 20 years to take Mentor from a troubled state to a prosperous one. I don't have 20 years. It's gonna be less than that. There's a little more sense of urgency here. I think the things that really make for fun is, first of all, you're coming from a low base, some really disappointed customers and disappointed investors, and everything you do gets a lot of gratitude, but it's also very visible. The first thing was, "Hey, we have some real financial issues here.
Let's get the ship righted in terms of the financial base, you know, cash, revenue growth, fundamentals, but particularly take out a lot of costs. The cost reduction is pretty much well into the final stages, and so now we're moving into the next phase of growth. The great thing about these kinds of turnarounds, in EDA in particular, you need to hunt around. There are always jewels somewhere, or at least the roots that can create them. You pick one or two, not five or 10, one or two, focus a lot of resource, and take it to the point where it enters that exponential takeoff that occurs whenever a new, totally new capability is introduced. Slow at first, then it takes off.
From my point of view, the first, the financial stability, and I think we've seen initial results of some financial recovery, both in the stock price and certainly in our, the financials that Chris talked about. The second thing is build a basis for future growth and stability. Find one or more products or product groups where you can have substantial major growth going forward, and that's the kind of thing that takes a company from sort of a people don't care that much to an exciting one. Mentor did it with just a couple products, really. Calibre just took off so tremendously over that period of time that it became enormously more valuable company, more than 10x an increase in value just from that alone.
I think the same thing's possible at Silvaco, maybe even more so 'cause it's coming from a much smaller base than the Mentor case. That, that's what I'll be looking for. That's, I've gotta do it quickly, but I think Silvaco has the potential, and I can assure you that as we do it'll get more and more fun.
All right. That's a great way to maybe wrap this up. Before we all exit, let me see if any questions from the audience, anything that, let's see if Wally, you can answer some of the questions live here. I believe you will have a Q&A box on your interface, and see if you can type some questions there for folks on the line. You can also send me an email if you would like to. Let's give it a few minutes and see if anything here.