Synopsys, Inc. (SNPS) Earnings Call Transcript & Summary

Hi, everyone. I'm Joe Vruwink from the vertical software team at Baird. Thank you all so much for attending our conference. We all very much appreciate it. The big kickoff is Synopsys. So Synopsys is a leader when it comes to the design software that goes into developing chips, IP products that can be incorporated into designs and then security software that can help identify risk during software and application development. Very pleased to have with us from the company today, Shelagh Glaser, she's CFO; and in the audience, we have Trey Campbell, Senior Vice President of Investor Relations; and Phil Lee from the IR team. This is going to be a fireside chat format.

I will feed questions from the audience in along the way. You can e-mail Session 2 at R.W. Baird, and I can moderate those. But why don't we begin with just an overview of Synopsys, and we can take it from there.

Great. It's really great to be here. I was telling Joe, I remember coming during COVID in New York, and it was quite lonely. So it's good to see everybody here. So I'll just give you a little backdrop on Synopsys. So we are in service of helping support silicon companies, get silicon to market, and then we think about ourselves as software to silicon support. We just recently did our Q2 earnings. We had excellent financial results. We beat the top end of our range on revenue and EPS, and then we raised the full year. So our full year guidance 14% to 15% revenue. A 150 basis points improvement in our op margin and 21% to 22% EPS. We've got a very resilient and stable business model, we're indexed to semiconductor R&D which does not perturbate with some of these demand supply things that are going on, very resilient model. We've got $7.3 billion in backlog and we're off and running in the AI space. I think that may come up here. Just might be a topic of conversation, really bringing AI to automate design for this complex workload. So a really strong business, excellent progress in execution and financials this year. Glad to be here, Joe.

That's great. I don't want to be the analyst responsible for getting the time off very first meeting. So can we get the stop clock working just so I keep all of us on track. So we're going to get to AI. But before we get to AI, maybe to go back a couple of years, in December 2021, Synopsys issued new long-term financial guidance, and it entailed consistent annual double-digit growth, more than 100 basis points margin expansion that works out to mid-teens earnings growth. What were you seeing back then? I remember at that time, the thought that 10% plus could be a sustainable number was quite shocking, but last year was 20%. This year is going to be mid-teens. So what has been going on to ultimately drive the strength?

Well, I think underpinning it all is we all are seeing it. Software runs on silicon and the world runs on software now. Everything is -- every company is a digital company. Every company is a technology company. And at the heart of that is silicon. And we're really in service of helping people create silicon and helping them simplify that very complex task. And what you're seeing is you're seeing all sorts of different instantiations of silicon, more people building silicon than ever before. You're seeing, obviously, traditional semi companies, hyperscalers, you're seeing automotive companies. And the reason that they're on a race to build silicon is really workload-driven down silicon, in the olden days, it used to be that there were general purpose engines and then people design software. Those days are gone. Things are often co-optimized and people know what the workload is, and they're building the silicon specifically for that. And one of the most important vectors is really power management. And if you think about power is pretty nonnegotiable. So in a hyperscaler, power is pretty nonnegotiable. In an autonomous vehicle or an EV, power is not negotiable too. So we expect to see more and more of that. We expect to see a hybrid of different types of silicon. That means more design engineers. That's our key customer. More design engineers, more business for us, and we can increasingly play a role in our customers' success. Because as some of these new modern companies move into making silicon, they don't, for instance, build an IP group that does standard IP block. So not only will they advantage themselves of our design animation tools but they'll also advantage themselves on our very broad IP portfolio. We have greater than 7,500 titles in standard IP blocks, on 340 process nodes. So if you want to build a chip and you're a more modern company building a chip, which might be a hyperscaler or could be an automotive company, you're going to want to have that standard block of IP library. So we can serve a bigger role for our customers.

I get the question a lot. Synopsys sets an expectation at 10% and then does 15% or 20%, what drives the upside, I think, the growth you've seen with system companies, as you've been discussing as a contributor to that. Can you maybe just put some numbers around that? You've talked about shifting mix of revenue in your business? I mean is that the bulk of incremental growth or when you look inside your traditional semi audience, is that also growing faster?

So there's really 2 swim lanes there. One is, increasingly, we are servicing customers that are nontraditional customers or system customers, it's about 45%. So obviously, that's -- they didn't use to make semiconductors. So now they do. And it's incredibly important because for their business, that's what they sell. They sell cycles on those. Those engines. So they have very good understanding of what the workloads are, but we're also playing a role inside our traditional customers. The race is on. I was out of the industry for about 3 years, I was doing SaaS and came back. We've all seen Moore's Law, don't know how many times that's been predicted and that's going to end. But there's really no end in sight in terms of the need for more and more complex workload. So as there's more complex workloads and even the numbers of chips in the traditional customers explodes, that's opportunity for us, too.

I remember meeting with Sassine, President of Synopsys. And he mentioned that automakers are aware the hyperscalers were maybe 6 years ago and so when you think about these other industries coming up the curve, is that just as consequential as what Apple ended up doing and the cloud computing companies ended up doing, I mean, is auto, aerospace, medical, industrial machinery are all of those opportunities?

Absolutely. And we see automakers realizing that what they're doing in some of our customers talk about building software-defined cars. And so really thinking about it from a system level and making sure not only that the design is optimized for the workload, so a vehicle out on the road, which has to be saved, has to be secure and has to have ways to signal that something is wrong. So they think about that whole solution once it's out in the wild. And then also their whole supply chain is changing. Their traditional supply chain is trying to understand. So they're coming to us to try to have us help them in that endeavor because this isn't something they've really ever navigated before. So yes, it's a bigger opportunity for us because they're asking us to actually even play a bigger role with them.

AI? Anything to talk about?

Definitely. I think AI is -- I mean we are all witnessing sort of a revolution in both workloads and capabilities. And we've been pioneering AI and design automation. We started in 2017 with our capability called DSO.ai which we've now brought to market. And we announced in Q1 earnings that it had been used on 100 commercial tapeouts. And as of Q2 earnings, that's 200 commercial tapeouts. And these are products that people are going to ship for revenue. So that's very consequential. And we've got the broadest portfolio in the industry. So we've got design space automation. So think of that as being -- when you design a chip, making sure that the chip is optimized properly. And then we've got test, so making sure that the chip does what you want it to do and then verification, DSO.ai and we're calling it synopsys.ai and the results that our customers are seeing are quite profound. So in terms of the design space automation, which we've got the DSO.ai which we've got the most experience with, this is something where customers are getting better results, faster with fewer engineers and less qualified engineers. And there's a shortage of silicon design engineers across the world. So there's already a extreme shortage, more people designing chips, and so we're giving customers tools that allow fewer people and less seasoned people to get better results than the humans were getting. It's pretty profound and they're getting better power, performance, and area in the DSO.ai. So we think pretty beneficial productivity results for our customers. So we're really excited. We're very committed to -- we've already got the broadest portfolio. We're very committed to continue to build this up. We think it's an exciting opportunity for our customers. And then obviously, as we create more value for those customers we share in that.

I'm going to ask this question as admittedly a non-engineer, but the EDA industry has been on this journey where it started [ all point ] solutions. I think there's 60 different tool categories that really matter, and then Synopsys said, well, why does it have to be point solutions. And so you started bringing some of the steps in the flow together. And now you have something like Fusion Compiler where you can do a lot of different work without ever leaving a tool. And AI seems to be kind of the evolution of this strategy, but it's also the enhancement. So what is needed from an underlying standpoint, is anything getting cannibalized if you're a Synopsys in order to do AI or is AI kind of the overlay on a foundational set of things. So it's really only incremental for you as a business.

For us, it's incremental. So the way that, for instance, the DSO, the design space on automation.ai works. You've got your underlying tools, which is a subscription business for us. So our design automation tools are sold on via multiyear contracts, subscription to a customer. And then the DSO.ai rides on top of that tools. It's using data from the tools, and we charge for that separately per project. So that's not a part of the subscription. That's a per project charge. And then as that is running thousands and thousands of iterations and making suggestions of what a possible layout would be. It's actually taxing the underlying tools and driving up usage of the underlying tools. So we see it as driving sort of a multiple in terms of usage of underlying tools and then also the specific charging for DSO.ai on top. And it doesn't share data from the customer. So the customer data stays safe with the customer. It's not getting out into unsafe hands, which is incredibly important for us. Root of Trust is incredible, important thing for us as a company. So there's no concern that a customer would have that their data would end up being out in the public domain.

And then maybe one last question on this. I don't think it's unusual for a big semi company to think going into a year, we're going to have thousands of new design starts. And so because this is project based, that thousands of opportunities to monetize. That's the way to think about it?

Yes. So the way that we are monetizing at this point is per project. So it's not just included in the underlying tool. It doesn't sort of come with.

Let's shift gears and spotlight the IP business. Just for some context, IP has been the fastest-growing category of the EDA industry for the last decade. 9 of the last 10 years, it's grown at a double-digit pace, and you're much more levered to IP than your big peer. So why is IP kind of where you decided to plant your flag and say this is something we need to be the best at and even in these long-term forecasts we've been discussing it is anticipated to be your fastest-growing business going forward?

Well, IP is very sticky. So if you -- what IP is basically, think about it as LEGO building blocks that we make, and we make the standard IP block. So what a customer will make is they'll make the differentiated IP blocks and then we can serve to them standard, sort of battle tested, already proven on a technology and that helps them go faster and helps ensure their design is successful and you get incredibly close to a customer. Because they're building some of the LEGO IP blocks. You're building some of the LEGO IP blocks. It's a very tight working relationship because any one of those blocks not working means the chip doesn't work. So we view it as incredibly strategic and incredibly good opportunity. And we think it's a huge growth area. If you think about all the work that's happening. And every single standard is on a road map of improvement, and then it has to change when you go to a new process. So that's sort of multistep that you can do. We've talked a lot about multichip. Those interfaces, those are all different IP blocks. And even if you think about the multichip, there is different chips where you can sell an IP block into. So we think about it being incredibly strategic. It's very sticky, and it's a great business opportunity.

I think I mentioned, I wasn't going to ask a question pertaining to quarterly guidance, but [ alight, ] I'm going to go down that route. Just in terms of how the outlook for this year is set up. The next 2 quarters are anticipated to see a step up in growth and IP is a part of that. What provides that sort of visibility or what enables Synopsys to have insight into how customers are going to start designs, use IP that allows you to kind of build it into a forecast?

Sure. This year is a more normal year for us, which is a bigger second half. Last year was somewhat of an abnormal year where it was really evenly distributed. And just as I described on IP, our teams are deep in work mode with our customers. So we have a very strong understanding of where they are on their road map. And the way that IP works is you think about us as every day, we're building new IP blocks. And then when a customer hits a certain phase of their design, they pull down the IP block. So it tends to be a lumpy business. But we look at it in annual increments, not quarter by quarter. And we can very much see a lot of the pull down activity is going to happen in the second half of the year. And again, we're side-by-side with customers as we're building our standard IP blocks that they're linking together with their custom blocks that they're building.

I think the lumpiness is part of the answer to what I'm going to ask next, but entering this year, you actually kind of shifted reporting Design IP was more visible, not just revenue, but also EBIT margins. And I think a lot of us were surprised at how profitable IP was. The thought was, well, software, of course, is very profitable, and it's absorbing the hit from IP but IP, I think, had quarters a year ago, where it was a 40% margin business. What kind of effort did [ is in ] -- because I know it didn't always used to be that way. So what got Synopsys to this point? And can you kind of think about how IP should trend relative to the software business?

Sure. A big -- we have a big focus on making sure that we're on the cutting edge of IP. So when a new process is coming up or a new standard is coming up, that we've got that ready to go. If you think about the industry, a lot of the leading edge is where a lot of the innovation and a lot of the kind of premium pricing happens, so making sure that we're always on the leading edge is an important thing. And then the scale. So the more we scale out and the more we help provide bigger and bigger support to our customers, the more opportunity there is. We think about the IP business as being slightly below our overall corporate average, but it's an incredibly profitable business.

Okay. And then maybe spending some time on core design and verification. I guess I'll leave it open ended to start. Kind of what would you focus on as the key performance drivers there?

Well, in the design space, so -- and I forgot to mention the other one, I wanted to mention that we did this year start breaking out and providing more on segment reporting information. I do want to highlight that. So we've got our design automation, which we'll talk about here in a minute. It's about 65% of our revenue, our Design IP, which we've talked quite a bit about, that's 25% and our software integrity, which makes up the balance. And so in the design automation space, the underlying trends that we talked about is just the plethora of people designing silicon. So that's the underlying driver of demand there and the complexity that people are dealing with. So the need for more performance and smaller power envelope is putting a lot of pressure on our customers. So that's driving a big opportunity. And inside design animation, we have the software tools that we've talked quite a bit about, but we also have a hardware business. And so that business, we talk about our ZeBu business. We just launched ZeBu 5, and that's to help assist customers on verification. And when customers are doing verification, they're basically looking at the chip before the chip exists. So they're building a software model of the chip and they're trying to understand the interaction. And sometimes they're making sure that they've got the software right, they're making sure that the chip is right. They're making sure those 2 things are working in tandem the way they expect because it's very expensive to tapeout a chip and find out that something is wrong. So customers want to look at it ahead of time, and that's where our hardware business comes in. And Aart, who is our CEO and Founder, will talk about verification is literally an unbounded problem. Because the chip could end up in so many different situations, how do you make sure that you stress test the right set of use cases. But we think again, with more customers developing silicon, the complexity of what people are facing, verification being an unbounded problem, our hardware business has great opportunity to help our customers tame that complexity.

I think I saw a study which is a design engineer actually spends more time in verification tools than design tools at this point to give some sense of just the importance of verification. I guess related to this, we kind of look at nonrecurring revenue as a proxy for the hardware business, last year grew 40%. You've said this year is going to set a new record. Is there a limitation to how much customers can buy is a typical account saturated in the verification hardware they buy from you? Or is this a case that in a development process not every customer has even gotten to the point where bringing up software as part of an emulation process is something they've even thought about yet.

Well, I would say customers are in all states of maturity. So some customers, this is very much a part of what they build into their, I'll call it, the whole design flow, for other customers, they're not as far along on that journey. And the thing that limits what's bought in any particular year, obviously, was supply, we're trying to resolve that. You saw us build some inventory in the quarter to help alleviate that now with supply lines not as constrained we want to be able to have that available for our customers. And typically, there's a window in which people buy that equipment. So they buy it, they phase it out based on time line to the project. So it's usually but once they've got the project assigned, they've got capital budget for it and then the provision how much of the hardware would they like to buy and it's going to be lumpy because people buy it for a project, the project might be a multiyear project, and then they would be back again to buy it. But I think in terms of this opportunity, we see it as a strong growth opportunity because what you're basically doing is you're building a chip on chips. So the hardware is giving you a model, building a model of what that chip is going to look like. And as the chip development gets more and more advanced, you'll want a more advanced engine to build that model on because otherwise, it's kind of running out of capacity or running out of throughput time. And so that constant refresh is sort of built in just to align the constant refresh of all the things we have to constantly refresh.

The iPads having some WiFi issues. So I'll ask a question. I'm cutting it short because I'm just working off the preview of the question. But it gets back to AI. And the question is, ultimately, doesn't this lower the barriers to entry easier to code, easier to develop. And so this is going to impact all, all incumbents, all traditional software vendors.

So we think about kind of our history. So the history of the company, you would talk to our founder who is our CEO, Aart, is the first product that we created with Synthesis, which is part of where the name comes from, and it allowed engineers to automate Synthesis. And he says, when that came out, all of a sudden, it was done so quickly and it used to take so many manual hours. And about 5% of the engineers were worried, Oh! My gosh, my jobs are going to go away. And the other 95% said, Oh! My goodness, thank -- thank gosh, because there was so much work I was never getting to. So I think this is going to become like a lot of things where it frees up routine tasks and allows us to do even harder task. And even in our design tools, as we talked about, Synopsys.ai and DSO.ai. It's not coming up with the architecture for the chip. People are coming up with what the architecture for the chip is, what the workload needs to be for the chip, what are the features, the tools taking part of the job of people and making that more simple to give them suggestions, but humans are deciding what the suggestions are. So I think that's what we're going to see with AI. And by the way, our AI different from like ChatGPT, there can't be anything amiss with what the recommendation is because any bad transistor in a chip means the whole chip is bad. So while you might tolerate some strange results from ChatGPT or something like that, you won't from our DSO.ai or Synopsys.ai tools because it has to be rock solid. But we view it as kind of the next level of helping for, yep, complex workloads and then helping free up the engineers to even solve bigger problems than they've been able to solve.

Maybe towards the end, we can talk about software integrity. This was a business originally meant to expand your TAM, it does bring you in front of some different user personas, different industries, it's also a business that has a lower margin profile relative to semi designer IP. Maybe we can talk about the original ambition, why you decided to get in and then the opportunity going forward. Is it the expectation that ultimately, this business can be a 30% margin business?

Sure. So our Software Integrity business is application security testing. And as we started -- we're working with our silicon customers more and more, we're hiring more and more software people. The discipline around software is not the same as the discipline around hardware. And so one of the things that was really clear is having tools to help make sure that, that software is clean, that there's not issues in the software and so that was a kind of security and then that's morphed into safety over time as software is running bigger and bigger workloads. And so the opportunity there is a TAM expansion. So traditional customers buy that software, but customers way beyond our traditional customers could be transportation companies, could be financial services, really every company builds software and making sure that those -- that code is properly safe and secure is important for every single company. And it is in a growth phase. Our growth plans for that are 15% to 20% a year. As I mentioned in the Q2 earnings with some of the pressures that you're seeing across lots of enterprise software companies that will be at the lower end, and we are committed to drive that operating margin over time, and we are committed that we think it can get closer towards the corporate average.

Great. I think that brings us the time. So we'll leave it there, but please join me in thanking Synopsys.

Thank you.