Good morning. Welcome back to this morning session. I'm Vivek Arya. I lead the U.S. Semiconductor and Semiconductor Equipment research team, and absolutely delighted to have Shelagh Glaser, CFO of Synopsys, joining us this morning. We'll go through a list of my questions, but, please feel free to raise your hand if you'd like to bring any other point in between. Very warm welcome, Shelagh. I really appreciate you joining us this morning.
It's great to be here.
I guess six months on the new job, you know, still the new office smell, just, you know, shiny, new things. Give us a sense of how it has gone so far. You know, what's been the, you know, positives and, you know, things that can be improved on from your perspective?
Sure. I mean, it's been a whirlwind. I knew this company. I worked at Intel for 29 years, then did SaaS, came to this company, so it's kind of like the mix of those two things because we're in service of silicon engineers, but we sell them software. The pace has been incredible. The pace of innovation in the industry is extremely exciting. The number of chip designs, the number of different flavors of chip designs is super exciting. I think the, you know, world now understands chips more than they used to before the pandemic. It's been a really exciting time. The company's on a great growth trajectory. We just finished our Q2 earnings. We beat the high end of our revenue range. We beat the high end of our EPS range.
We raised our revenue guide for the year, we're 14%-15%, and we raised our profitability guide 150 basis points year-over-year, and our EPS, 21%-22%. We see a lot of opportunity in the market, and our role is to help our customers tame all this complexity that's coming at them and allow them to be able to serve up the chips to their customers. It's a great business. We have a big subscription part of the business. We have $7.3 billion in non-cancellable backlog, so very robust, very durable business model. It's an exciting part to play in a great industry, and it's a great company to be on the team with.
Excellent. One thing, Shelagh, that I find very interesting about the EDA space is how it has consistently accelerated the growth, right? What used to be 7%-8% growth, went to 10%-11% growth, and, you know, as you just mentioned, now you're putting up 14%-15% growth, even though the semiconductor industry growth rates have kind of been in that, right, 5%-7% range. What has helped Synopsys and the EDA industry consistently accelerate growth rates over time?
Sure. We're indexed to the R&D investment, which I think, even though there's a lot of perturbation on finished goods demand supply, that R&D does not change dramatically, 'cause it takes three to five years to build a chip. The role that we're playing as we're innovating on our EDA tools and our IP, we have an enormous IP library, about 7,500 titles on 340 process nodes, is how do we help our customers simplify the work that they have to do? They're getting bigger and bigger requirements, but they're not necessarily having more people to do it. As you said, their budgets are not expanding by probably their requirements, that's where we step in. Over time, how do we help them automate some of their work?
I'm sure we're gonna talk about AI here, so we've got exciting things to talk about there. How do we sort of become a multiplier for them? As we become a multiplier for them, we actually share and have a bigger and bigger role in the R&D spend, because we're helping them do more work. We think there's more that we can do, and there's more opportunity and as we do that, then we share a bigger part of the pie.
Got it. Do you think these kind of growth rates are sustainable, or what needs to happen for these growth rates to be sustainable? That, or is this just, you know, a temporary, you know, one-off, and the sustainable growth rate is lower than this?
Well, the underlying pinnings of demand are not going to change. I think now everything is digital. Every company is, you know, software company, underlying software is silicon. The trend on silicon, I think is, you know, at a new pace because that silicon needs to be really purpose-built for a power performance envelope. Gone are the days where it's just one piece of silicon and everybody go figure out your workloads. Now, the workload is deciding what the silicon is. That trend is not changing. In terms of the requirements for the teams to go off and solve these big, hard problems they need to solve, that trend is not changing, and we're the tools that help them be able to tame this complexity. Our commitment is for continued double-digit growth.
We think that that is certainly possible over the horizon, again, as that underlying demand and that needs for more and more compute and specialized compute continues.
Got it. You mentioned those two magic words of AI, let's dig into that. Give us a sense for how is AI helping you both from a revenue capture and growth perspective, and then also in terms of optimizing, right, your tools, your DSO.ai, right tool, number of design type, almost doubled sequentially. What's, what is the value proposition of that? You know, why are you suddenly seeing that acceleration of that tool?
Sure. We pioneered AI in design automation, so we started it in 2017. We had some experiments we showed in 2018, and we were able to commercially show it with a commercial tape out in 2021. We started in the front end of a design, and our first tool was DSO.ai. That's Design Space Optimization. That's a super consequential part of designing a chip, deciding how that chip is going to be configured, because that affects a lot of steps down the line. It affects what the power of the chip is, what the performance of the chip is, and what the area is. That's a very consequential workflow. People don't have enough people to throw at that problem, and don't have enough time to optimize it.
What our tool does, the way we monetize it, is you buy our underlying subscription, and then you buy the DSO.ai tool per project. That's not a part of the subscription. What we're finding with customers is, as they're running multiple iterations to optimize the design, they actually use more of the underlying tools. The thing that they wanna do is they wanna come with, you know, what is the optimal optimization for this design, and they only have a certain amount of time to do that work. If they have more tools from us, they can run more iterations. What we also announced in our software users group at the end of March, is that we're launching TSO.ai, that's the test, and VSO.ai. We're building AI tools out across the entire design flow.
Mm-hmm.
We, you know, each one of those is a per project, and it's driving, first of all, customers to wanna use our tools, because the, you know, data is only from our tool. Our customer's data is very precious, we can't be sharing that anywhere, so it's very important that we handle that with care. You need our underlying tool for the tool to run on top, and we think over time, this has enormous benefits. We're really focused on making sure that we're driving broad adoption, so we can have the successful use cases, and we'll experiment over time with, you know, additional monetization.
You know, as you mentioned, that even though, you know, I guess we are all kind of waking up to this growth spurt in AI, right? You guys have been engaged with it for a number of years. As you take a look back, is there a way to quantify how much AI has given the company in terms of the incremental? Like, would, you know, growth rates have been like 11% or 12% instead of 15%? Like, is there a way to quantify how much AI has helped in accelerating the growth rate?
Certainly, it's one of the reasons that we're growing faster than, as you said, the underlying R&D spend is growing, so it's a contributor to that. We think about our pricing as we're adding more value to the customer, and as a part of adding more value, we're collecting a portion of that value. It's certainly been a contributor.
Got it. Any other, quantification, Shelagh, which is on. Is there like a certain attach rate math on DSO.ai or VSO.ai or TSO.ai, right? Anything that helps us better understand where you are in the process of upselling that incremental tool on top of the baseline, and where that can go over time?
Yeah. We've been really excited at the rapid adoption rate. It obviously started in, like, POCs, because, you know, the design flow is so important, you can't be putting something in there. We announced in Q1 that we had 100 commercial tapeouts. Using DSO.ai, we announced in Q2 that we've got 200 commercial tapeouts. Obviously, that's still a fraction of the tapeouts, but you can see we're getting really rapid uptake. Our Fusion Compiler, which is, you know, kind of our number one seller, we think this adoption rate may be even faster than that, and that was one of our fastest adoption rates. We've been really excited with how much, the customers are, clamoring for this.
The reason they're clamoring for it is a lot of the things that we've been talking about: more complexity, more chip designs, not necessarily more, skilled people to be able to do those chip designs. This helps become a multiplier for their team. We've seen really rapid adoption, and as we see that more adoption, you know, collecting customer successes and what were their outcomes, is part of what we're really focused on, so we can continue to, you know, drive even more benefit to the customer.
Got it. You have not given a number, but if I throw out a number like mid-single digit percent kind of attach rate, would that be somewhere?
I think we're still.
Lower than that.
In early innings. We see ahead of us, really, most of our customers are gonna want to adopt this.
Got it.
We don't see sort of a barrier to adoption. It's more, when does it make sense for a customer to infuse it into their design?
Got it. One other thing that we have seen that's interesting, is that we have the hyperscalers who have their internal, you know, design teams, right, bringing some of their IP to bear. We have, you know, we just had the team from Marvell yesterday, we had the team from Broadcom, you know, helping them. Do both of them, not them specifically, but do, like, both the system companies trying to do internal designs and the semiconductor companies, buy these tools, or just one of them buys the tool? Like, do you get a magnification as more systems and semiconductor companies engage in this?
For us, it's an and. If we think about there's traditional semiconductor companies, there's new entrants, hyperscalers, automotive is also a new entrant. More people designing more chips, that's really our surface area.
Got it. Okay. If you look at just the growth in AI, do you think, you know, because if I look at it just, you know, the most kind of downstream, the spending pie is not growing the same level. Is something else not being done because AI is taking up so much of that CapEx and that oxygen, right? Do you see that push and pull in your business also?
What we see right now is, you know, the time to design a chip hasn't improved a lot. It's probably gotten longer, but the impatience with people getting chips out is high. What we see is people don't have enough resources to be able to get things done. What we're filling in the gap is. They don't have enough seasoned resources. By the way, I should have mentioned that the results customers are getting on the DSO.ai is better than the people we're getting, and it's fewer people and less trained people. The gap that we're sort of filling in is that gap in terms of resourcing that. In terms of, you're right, ultimately, you know, does the pie grow for everything, or does some portion of the pie get shared differently?
You know, at this point, we think we can play a bigger role in whatever size pie it turns out to be.
Got it. While, you know, AI, right, just the design starts on AI are going on, you know, you're also benefiting from just the progress on the process roadmap, right? Going towards 5 nanometer-
Mm-hmm.
3 nanometer as gate-all-around. Give us a sense for what that means for Synopsys. Let's say there was no AI, right? All we had was just this increasing complexity on the process side. What does that mean incrementally as we go from 1 node to another?
Sure. I think, you know, Moore's Law, as we knew it in the olden days, has changed. It used to be you got the benefit, you know, almost automatically. I know people worked really hard, but almost automatically going to the new process, and that's not exactly what's happening now. We see the race is still on with, you know, people building big factories, new, the new litho, and people wanting to continue that. We also see that people wanna get into what they're calling multi-chip solutions, so that maybe not all the die is moving forward, maybe only portions of it.
Right.
They want the chip to actually act like all the parts are right next to each other. We see both the race in technology as helpful for us, but then the solutions that people are starting to, you know, bring to market are also beneficial to us to try to tame that complexity. Aart calls it systemic. Our founder, who's our CEO, calls it systemic complexity because, you know, it was hard enough to design things on one chip, but now you're gonna have to design it on multiple chips, because maybe one or two of the chips stays in node back, but you want the multi-chip to behave as if it's all one instantiation, so no delay in communication across. We see that as an opportunity, too.
As people try to solve some of the, you know, barriers we're running into Moore's Law, some of the innovations they're coming up with is also a growth factor to us. Continuing the race, but trying some new approaches to mitigate some of those challenges that are on the, you know, standard Moore's Law race.
Got it. You brought up this interesting topic of, you know, multi-die silicon. You know, the traditional design is, right, just being, making larger and larger sized dies, right? That obviously increases Multi-die is supposed to reduce complexity, right? How is reducing complexity a positive for Synopsys?
Well, it changes the complexity, because now all of a sudden, when you have one die, one group of people is probably building that, you know, monolithic die. You've got sort of, you know, one set of parameters they're operating to. All of a sudden, when you change it, and maybe a chip's coming from one vendor, and a chip's coming from, you know, vendor A, vendor B, and then you're putting your chip on top of it, you changed where the complexity was. All of a sudden, you have to be worried, is that chip good? You know, where did that IP come from on the chip? We haven't talked about our IP business, but we obviously have, you know, an incredible IP business. 7,500 titles on 340 process nodes.
I mean, we're the land of standard IP versus Arm is CPU IP. We sort of do everything but that. You wanna make sure all those things are working well. Then you wanna make sure that that communicates. The complexity at the, you know, monolith die hasn't changed. Now the complexity in managing all those interactions across multiple parties, that changed pretty considerably.
Right. Oh, I'm glad you brought up the IP side. I mean, next to Arm, you guys, I believe, are the second largest-
Yeah
IP. It's a very different strategy than your main competitor, right, who has a much smaller IP business. Talk to us why Synopsys decided to have a much larger IP business, and what are kind of the pros and cons of having a much larger IP business?
Sure. Arm does CPUs. We do all the standard interfaces or memory. It's incredibly strategic for us because it puts us in people's designs. We are now side by side. You can think of the IP as sort of Lego blocks that they would integrate next to the blocks that they'll create. They'll create some differentiated block, and then they'll take standards from us, and it puts us literally right inside the design team. You can think about us being kind of hand in glove with the customer, so it creates a much tighter alignment. Over time, and certainly we talked about some of the newer players, they didn't even build the standard IP block teams because they know that they can pull them from us. They know they're battle-tested.
They know that if you wanna ramp on a new process node, we have that ready for you. For the traditional companies, I think it's an opportunity for us over time. As they have more and more on their plate, how do we play a bigger role? It's a profitable business. We think of that business, think of it almost as building an IP factory. You're constantly building IP blocks. Standards always evolve. You always need to move to the next node, and then customers pull the IP based on when their design is. We kind of manage it over a 12 month period, and we think that can grow in the mid-teens, and the profitability of that is slightly below our corporate profitability.
It gives us a much stickier relationship with the customer because we're literally right inside their chip building.
Got it. IP, you think, actually can grow faster than EDA?
I think, well, IP is, you know, those are the building blocks across everything. The race is on for all our businesses to grow, because we want the company to grow at double digits.
Makes-
you know, opportunity across all of our businesses.
To clarify, the majority of your IP is still subscription-based, right?
Well, we can sell IP multiple different ways, we sell, like, a Flexible Spending Account, we call it an FSA, that's got a specific term to it. The customer will pull the IP based on when they're in the, you know, stage of their design flow, when they need to pull the block down and integrate it, what they've signed up with us is for a specific amount of dollars over a specific amount of time.
I see. It's not as kind of cyclical, right? It's not as exposed to units.
There are some where they just buy an IP block, so buy it out of the shelf. Increasingly, we have customers buying in the FSA because they know they're gonna be building a chip, they know that over time, they're gonna need the IP block.
Makes sense. Then the third part of your business, the Software Integrity business, that's the business I find toughest to understand, because it doesn't do much with semiconductors. Give us a sense for what is the, kind of the strategic rationale for having that as part of the company, and what are the synergies between your EDA, with that business with EDA and the IP part?
Sure. The, you know, the customer for our core businesses, our design automation, our design IP, is the silicon design engineer. That's who we're in service of, or a manufacturing engineer, manufacturing test or verification engineer. The Software Integrity, really that's in service of, like, the DevOps, the software folks in the team. If you think about chips or software and hardware, the discipline around software is different than the discipline around developing a hardware chip. As we started on the journey, we wanted to build a comprehensive set of tools for those customers, and it was all about security, and increasingly, it's about security and software. Security and safety in software, because there's so many places where, you know, software is playing a major role.
Our core customers buy it, but we can also sell that to financial services, transportation companies, way beyond. Really, as we start to work with, like, auto manufacturers, they're certainly worried about from silicon all the way to software, because what's gonna, you know, when a car is out on a road, how's it gonna behave? How we're gonna do the updates, how we're gonna know that's safe and secure? As we think about smart everything, we think all of those things do end up coming together at a solution.
Right.
What we're in service of is making sure that, you know, application security testing, making sure that all that code is good, and that there's no, kind of, you know, bad code inside of there. We've got the broadest platform in the industry. We're the largest player in that part of the industry. We announced that, you know, on a, on a trailing 12 months, we're $500 million, so that's kind of our run rate now in that business, and we see that we can continue to scale that area. It's a very fragmented part of the market. We think, especially in a constrained environment, that consolidation play is even more important to people, and over time, it becomes increasingly important as a lot of the, you know, capability gets delivered in software, and people need to make sure that it's safe, smart, and secure.
Do you ever see it being accretive to corporate EBIT margins?
Our focus is on growing that. We've, our target is 15%-20%. This year, we mentioned it'll be at the lower end of the range, just a lot of the headwinds you're seeing in enterprise hardware, the elongated deal cycles, we're seeing all that. Our focus is on driving profitability, and we think over time, we can drive the profitability significantly and drive it towards the corporate average.
Good. One other macro type of risk that investors are concerned about is just the threat of restrictions, right? When we look at U.S.-China relationship. How much of an early warning system, you know, does Synopsys have? When the last set of restrictions came about, was the company informed about them, you know, well in advance, and you could take corrective steps? How should we handicap, like, if you were in our seat, how would you handicap that risk? Because it's a 14%-15% customer.
Yeah, China's 15% of our business across our different segments, and, yeah, we've incorporated all the entity lists inside our guidance. Everything that's known is fully incorporated, and I don't believe we get any more warning than anybody else gets. I don't believe that there's an early warning system. I mean, when, you know, the government makes a decision, then we certainly adhere with it. That said, the need for technology in China is still there, operating inside the, you know, confines and the restrictions, we still think that there's enormous opportunity for us, both on IP, EDA, and then also in our Software Integrity. We know that there's lots of uncertainty at play in China. We know that they want to become independent, they will, you know, endeavor to do so.
We view that as any competition. They're viewed, you know, our job is to, you know, continue to offer value so that we're the choice of the customer. I guess what I would say is, I don't have a crystal ball. I don't know exactly what's gonna happen there, but we still view in these current conditions that it can be a vibrant market for us.
Got it. just a few kind of near-term questions. When I look at the second half expectations, right, it's for like 10%-12%, half-on-half growth.
Mm-hmm.
How are you feeling about that? How is your confidence, visibility as you look into the second half of the year?
Sure. This year is a more normal year for us. Last year was a bit abnormal, and the second half and the first half were more evenly balanced. What we can see, and it goes back to some of the conversation we have, in IP, we know, we can see what's going on with customers because we're working so closely with them. I think I mentioned in the conference call. We see strength in the back half, but we see strength across our businesses in the second half. I think we feel good about what's going on in the business and feel good about what's going on in our end customers' markets.
Got it. Since, you know, the majority of your business, right, 80%+, right, is kind of recurring, so it's hard to change the revenues, right? Which is kind of both good, right, because it protects us from the downside, risk also. If there is any upside in the second half, what could be the potential source of that? Is it just hardware or?
Well, I mean, obviously, the, you know, focus is making sure that we're playing a bigger and bigger role in our customer success. For us, we've got the ratable business, as you said, the, you know, subscription business for our EDA tools. We've got IP, which is, as I said, increasingly we're selling it in FSAs, but that's always, you know, kind of period revenue. We've got hardware, which is period revenue.
Got it. Second thing on operating margins, I think you have laid an objective of expanding it at least, you know, 100 basis points. Historically, you have managed to expand it a lot more than that. What's different about this year that, you know, you can't, or is there a scenario in which you can actually be in line with the historical margin expansion?
This year, we had started with a greater than 100 basis points. We raised it to 150, obviously. Our long-term goal is greater than 100 basis points, so we're committed to that. You know, our focus is on profitability, driving that across the business. Thinking about how, as we continue to scale the business, we drive greater and greater leverage.
Got it. Then, free cash flow margins are very strong. Just given the predictability, right, in your business, why not consider paying a dividend?
One of the things that, you know, our priorities for our cash are, is first to invest in our business. That's our clear priority, and we obviously invest considerably in R&D. Everything we're just talking about, all the innovation, required is a significant R&D investment. We do quite a bit of tuck-in acquisitions.
Right.
We are able to bring capabilities in. We, in terms of cash back to the shareholders, we're really focused on this phase in terms of doing that through buybacks.
Got it. The last one, which, right, many investors, right? Investors have a choice, right, in terms of investing, right, between the two, right, the top two EDA companies. From your perspective, why should they consider, right, Synopsys instead of your competitor?
First of all, I'd say the space is a great space. The, you know, the old days of, you know, silicon's dead and, you know, only, you know, old lumbering companies do silicon. No, everybody does silicon. Silicon is the heart of everything. Software companies, what does software run on? It runs on silicon. The space is absolutely vibrant, and I see none of that changing. If you're interested in this space, you could be interested in either one of them. I would always choose us. I think that would surprise 0 people in this room that I would always choose us. You know, for me, there's a couple considerations.
First of all, I saw the power of this company as their top customer. That's obviously disclosed in our annual report, who our top customer is. The difference that it makes in terms of the capabilities that we're able to achieve as a customer, was profound. We are the leader in design automation. If you're going to go on a leading node, you come through us. You come through us on our design automation tools, you come through us on our IP, because you want battle-tested IP. What do you really want to do? You want the chip to work. When you want the chip to work, the stakes are really high. You want to make sure you've got a partner you can choose. We pioneered AI.
Right.
Just as you said, even before anybody was talking about any of this, the company innovated because we could see that we needed to do something to help our customers tame this complexity. The complexity was getting way beyond what any one person or any one team can actually deal with. That's doing nothing but accelerating. We've got already DSO.ai, VSO.ai, TSO.ai. We're calling it Synopsys.ai. We're gonna have it across our tool flow. That data requires you to have our underlying tools. We think it's an opportunity for us to further expand. Our IP, our IP library is unmatched, absolutely unmatched. We have the IP that can help you if you're on a lagging node, if you're on a leading node, help make sure that your chip is, you know, gonna tape out, gonna be a great success.
Then we're investing on the software side, because a lot of the value is coming over time from a chip integrated with software. So we want to play a role in helping tame the difficulties that people have there and making sure that their code is safe and secure. We see opportunity everywhere, and we are set to continue to evolve, maintain our leadership position, and serve a bigger and bigger role in our customers' success.
Excellent. Thank you so much, Shelagh. I really appreciate it.
Thank you.
Pleasure.
Thanks. Bye.