The last time they let me on stage to host anything was the Citi Tech Conference in September of 2008.
No!
I was the opening keynote, opening the conference. I won't remind you what happened shortly thereafter. So, let's just hope, we have a better outcome here. It's the last time they gave me the mic. So thank you all for coming. I'm glad you're all here because I'm gonna ask your help in, in helping me with the Q&A, 'cause we have two amazing speakers and an incredible company. So this is your opportunity to ask questions. But before I introduce Sassine and Shelagh, I have to read this because Phil sent it to me and said, "You better read this." Today's discussion will contain forward-looking statements related to Synopsys' current outlook, expectations, and beliefs, which are subject to certain risks and uncertainties that could cause actual results to differ.
Please refer to its most recent SEC filings for a discussion of the risk factors that may materially affect these statements. I can't read my own handwriting. So with that out of the way, I want to thank both Sassine Ghazi and Shelagh Glaser for coming here and presenting at the Citi Tech Conference, day two. Hopefully, everyone's having a good time. And so we're gonna start off with this. First of all, let me introduce Sassine. You're actually COO, but soon to be CEO-elect, so let's give him a round of applause for that-
Thank you
... in January, so we'll do it again in January. And Shelagh Glaser, who's CFO, and we have Trey Campbell and Phil Lee from Investor Relations. So let's get into it. Just given that it has been such a great growth story, let's just level set. Maybe, Sassine, I'll ask you to just give us an overview of Synopsys. What are, kind of from a high level, the different three businesses you're in and how you think about them?
Sure. First, thank you, JZ, for having us here. So we have three business segments. The first one is design automation, then design IP, and software integrity. Design automation is roughly 65% of our business; design IP, 25%; software integrity, about 10% of the business. Think of the first two segments, that anything related to semiconductor chips uses Synopsys. So as you are designing a chip for mobile, for data center, for automotive, you need automation tools in order to design the chip, and you need IP blocks. So if you want to connect the chip to the outside world or chip to chip, there are protocols that connects these chips, be it a USB, PCIe, DDR. That's the IP business. In all three segments, we are in a very fortunate position to be the leader in that segment.
We're the leader in Design Automation, in IP, and Software Integrity.
Let's drill down a little bit. Just at a high, again, a high level, let's talk about, given that design automation is the biggest business you're in, over the last several years, the business has really consolidated. A few leaders have emerged. So just talk about the competitive landscape, how you see it developing, and where Synopsys kind of fits in.
Yeah, the reason the market consolidated in design automation is the complexity of the chip. Of a chip design on an advanced node is so complex that you cannot have a design environment. So when our customers' engineers developing a chip, they cannot use multiple vendors to develop a block on that chip because it gets too complicated. So we are one of the very, very few supplier in the industry that has a complete design automation flow. The market has consolidated due to the nature of complexity, and this is where Synopsys is the leader in design automation, because we have the complete design flow for our engineers.
Okay, so let's take this as an opportunity. Sassine, you've been at the company 25 years. You've just gotten this great honor. So maybe just for the people who don't know you, a little bit about your background, your history at the company.
Sure. Actually, my career is fairly straightforward. Started my career at Intel. I was a design engineer at Intel, designing chips. When I came to Synopsys, I came as an engineer. Then Aart, the founder and CEO of the company, saw something in me, I don't know what it was exactly, and he started betting on me in different jobs. So I moved from an engineer to be part of the sales and marketing for a decade-plus, then moved to become the general manager of our EDA business. So I was managing the development of our EDA business for about 5, 6 years. Then the last 3 years, moved into a COO and president role, managing many functions of the company.
Well, I just have to say this: If you look at Aart's track record over the last 50 years, if he sees something in you- ... that's a very good thing. So that should be in everybody's notes. Okay. Let's talk about a little about what your priority is, because part of the challenge, it's funny, we had the new CEO of Texas Instruments up here yesterday, and, you know, Rich Templeton's a hard act to follow. Aart de Geus is a hard act to follow. So how do you view your priorities? What do you... You know, what's business as usual? Where, what do you want to put your stamp on the business?
So I'm not saying it because I'm the president of Synopsys. We really have a very special company.
Yes.
We have a great, great company. It's a great company for many reasons. One, the technology position we have in the industry is so unique. If you talk to any of our customers, and I'm talking about the semiconductor chip customers, they know us from the design engineer who's running the software all the way up to the CEO, because we're very consequential to their success and their business. As our customer base is expanding from the traditional semiconductor companies to the hyperscalers, to automotive OEMs, et cetera, the relevancy of Synopsys in that chain is very clear to the leadership of each one of those market verticals. So that should continue in terms of priority of myself and the leadership team as we look at the next decade of Synopsys. Our DNA has been built around technology, innovation, and leadership.
Aart was a disruptor himself when he created Synopsys. It was something new called synthesis, which is a new method, a new way to design a chip, simplifying the design of a chip, and that was 37 years ago. Since then, we've been on the leading edge in everything we do in terms of technology innovation. That will continue. The priority, and as you're seeing, our performance over the last number of years, has been focused on growth based on where do we make our investments, where is the new market trends, and two, how do we scale the company efficiently? Just as a quick side note, we doubled the size of our company in both number of employees and revenue in the last four years.
When you're growing that fast, you have to really think about how do you grow your own processes, systems, et cetera, to scale the company more efficiently. From a priority standpoint, it's about continued momentum on growth, on scaling the company, and the technology, innovation, and leadership.
That... It, it's a lot-- it's, it's a great place to start. Which brings me to the next great place to start, and that's AI. I know most of you probably have heard about- enough about it in the last 48 hours, so we're going to make more of it, because here's a great way to talk about it. On the recent earnings call, you talked about there are actually three ways that Synopsys sort of can get paid, both internally, externally. Maybe just to remind everyone, refresh everyone, what are, what are those three sort of levers you see for, for AI in the company?
Yeah, the way we articulated it in our earnings call, the first priority are the customers who are developing and designing an AI chip.
Right.
So any semiconductor company that are thinking or already designing an AI chip, they typically need a design automation products and IP that goes on that chip to accelerate their development. Those companies use Synopsys EDA and Synopsys IP to design their AI chips. That trailing twelve months revenue from that particular market, the chip guys developing AI applications, have contributed to about $500 million of our trailing twelve months revenue. And as you look at the forecast over the next number of years of more companies designing AI chips or beefing up their existing roadmap, that trajectory for Synopsys should continue on both design automation and design IP.
Okay.
So that's the first category. The second one is how can we, as being the leader in, in the market of design automation, provide an AI system to our customers where they can accelerate their own chip design effort? And that applies to any chip effort, does not have to be an AI chip. It can be mobile chip, automotive chip, et cetera. Just to put things in context, to develop an advanced 5-nanometer, 4-nanometer chip takes few hundred engineers working on that chip for many, many months. Even if it's a derivative from a previous chip, it typically takes 12 months-18 months to develop that chip. And resources are scarce for our customers to just put more resources on these chips, and the complexity is very high.
So what we've done, we've invested in AI technology for chip design to help accelerate the chip design cycle for our customers, and this is where we brand it as Synopsys.ai. And we have a slew of-
I was gonna ask about this. Go ahead, yeah.
So that's where we're branding Synopsys.ai, and we have slew of products and technology to speed up and improve the chip design process and development. The third effort is using AI to help us scale our own company internally. How do we use AI in our own development? How do we use AI in our own internal digitization and processes inside the company to drive the scaling? So those are the three categories. The first two, we can monetize them from growth perspective. The third one, it really reflects in our operating margin and profitability.
... And I should also point out, on the recent call, you've actually been very involved and integral to Synopsys's AI effort since, I think, going back to 2017.
Right. Right.
So you're not new to this?
And no, not at all. And actually, when we say Synopsys pioneered AI for chip design, this is not just a nice tagline. You look at when we, when we started AI development in that category number two I mentioned to accelerate chip design, it was in the 2017 timeframe, where not too many people were as excited as they are today about AI, and they wanna use AI. And we were really leading the market in that effort, by at least couple years, compared to where our customers were or the rest of the EDA ecosystems at.
Fair enough. Well, Shelagh, we're gonna bring you in now. So let's talk a little bit about operating margins, 'cause this is a big topic for investors. You've talked about growing operating margins at least 100 basis points a year. This year, you're talking about 200 basis points. Let's talk about what are the puts and takes that you see in achieving that?
Sure. So we're committed to operating margin improvement, both short and long term, and I think as Sassine said, over the history of the company, we've been driving improvements. This year, we're excited we're gonna drive 200 basis points improvement, so we'll end the year at 35%. And really, the way we're thinking about it is, just as Sassine outlined, we've grown the company very rapidly over the last four years, in both the top line and both our resources. So we're really looking at how we modernize systems, processes, and really improve the leverage. And so this isn't just a one-time, this is a long-term commit of at least 100 basis points improvement every year.
Could you talk a little bit about how the IP business impacts that?
Sure. So as Sassine outlined, we're really excited this year to share our three largest segments of the business, our design automation, our design IP, and our software integrity. And one of the things we're most excited about is each of those businesses has a slightly different customer base, has a slightly different kind of rhythm of the business. The IP business, we have the largest IP portfolio of any company, and that business tends to be what we've always described as lumpy. And we all know the word lumpy. It's hard to, you know, have an exact percentage of what lumpy means, but now that we've got that as a segment, we're able to share kind of the quarter-to-quarter variation. And we're not managing that business on a quarter to quarter, we're managing that business for the long term.
The way that business works, it's a very resource-intensive. So if you think about, we build standard IP blocks, those blocks regularly get updated, and so we have to apply resources to update those blocks. And then we have the standard blocks built on essentially every foundry in the industry, all the different nodes. So you need to constantly be building IP blocks. We sign long-term, committed deals with our customers, but the only uncertainty is the timing of when they pull that IP block. So we know that they will pull the IP block, but they pull it based on their schedule. So that's how we see some quarterly perturbations.
But long term, the margin structure of that business is slightly below our corporate average, and it's an incredibly sticky business, because once you're in the middle of somebody's design and they're depending on you to deliver a block in their latest, greatest design, you're a partner for life.
Okay, I just wanna come back on one last thing on margins, and we'll move on to another topic.
Sure.
And that is, Sassine touched on the sort of three drivers that AI have, the third one being this sort of internal efficiency. Is any of that embedded in this margin improvement outlook going forward?
So-
or how do you think about it?
So, we are in the early stages of really comprehending what that means. Sassine gave each of the organizations in the company a homework assignment to determine how each of us, and I mean G&A, sales and marketing, engineering, how we're gonna ingest AI to actually make make what we do for the company better, faster, cheaper. So we think that's all still in our future. We've not yet realized the full gains of that. We're in experimentation, and we see some really, really early exciting signs across every element of the company.
Perfect. Well, I'm gonna take a pause here for a second, go out to the audience, see if you have any questions. I can tell you the experiment I tried earlier today was, I asked the first question in a meeting, and that got the ball rolling, and then, like, three or four more questions came. So I can't—I'm asking the questions here now, but let's start with raise your hand if you have any questions you wanna ask. See, it's such a simple, straightforward story that you're covering it all. Okay, well, we'll move on, but think about questions that come up if you have any.
I think there was a question over there.
Oh, sorry. Okay. This is the guy who's gonna get the ball rolling.
Yeah, just super interesting on the one of the second topic you're talking about, the revenue flow we can see for the AI, how to provide a AI software, like a system, to your customers. Could you maybe give us a, like, a example, one of your customer maybe asking for it, or you are designing something similar to them? Like, how does that work in the real world?
Yes. So as I mentioned, we started the development effort in 2017. Then we started early customer experimentation to see are we hitting the mark or not around 2019 timeframe. In 2020, we had about a handful of customers in early adoption of the technology. At that time, frankly, they did not know what to do with it because they structured their team with a certain level of expertise, with a certain skill sets required to develop a chip and the way they've been doing it for a number of decades. So at that time, and we did not know how to even sell it. It's something exciting from a technology point of view. We did not know how to deploy it to the customer.
And what I mean, we did not know how to deploy it. It's the requirement of the AI system. It changes the way our customers run their data center, like how much compute do they need to run certain tasks? How much EDA software does the AI system will require to run and scale the jobs on a scalable way? In 2020, we started having customers with early deployment of the technology. At this point, we have about 270 partitions, meaning when you think of a chip, it has multiple partitions, that they have used that AI system in production, where that means they integrated it as part of their workflow and their new methodology of running a chip, and they took it all the way through manufacturing.
It's no longer an experimentation, let me see if I trust the tool or not. It's in production at many, many customers and is validated by the 270+ production tape-outs, they're called, using the technology.
Hi, hi, Sassine. So clearly, AI is a pretty exciting opportunity. So as you look to invest in this space, how are you looking to balance both the investment and also the revenue synergies with the Software Integrity segment? And you have given the, you know, metrics on the IP side, and it's slightly below corporate average. So, I mean, and you have given the metrics on Software Integrity as well, but would those margins continue to evolve and go beyond the 20% target you have laid out?
So the way we started that investment in 2017, it was fully organic because you don't need only an AI data expert. You need a domain expert. You need somebody who understands the chip design and someone who understand how does your product works, what is the algorithm, the optimization inside your own product and technology. So that was fully organic investment from a development standpoint. As we expanded into Synopsys.ai, which is beyond that early product, the early product was called the DSO.ai, which is design space optimization for a specific function of the chip. Then we're building along the complete stack of chip development, which is the Synopsys.ai, and it's all done, again, organic, self-funded, part of our investments in new technology.
As we look ahead, where can we make investment that's going to move the needle from a revenue point of view? It's contributing already to our revenue because we've been selling the product for since 2020. The way we're selling it right now is a subscription-based license, the same way as EDA is sold. A customer who's buying our EDA product, if they want to use the AI capability, they layer extra spending on top of their baseline spending in order to get access to our AI technology. I think of it, even though we have 270+ tape-outs, as still very early stage because many customers, they still have a significant opportunity to deploy that technology moving forward.
So that's a very exciting opportunity from an ROI on the investment we made for designing the chip. As far as the operating margin in the three different business segments we have, I'll make a couple comments, then turn it to Shelagh. We look at our business as a portfolio. We have three segments. Each one of them has a different growth potential and trajectory, and a different investment requirement in order to maintain a leadership position and investing in the future trends. In IP, for example, we made significant investments the last few years in IP for security. And because security is going to become significantly important as you have more and more AI applications and more software content, how do you have security not only at the software level, security at the hardware level?
So we made security investment in our IP portfolio. We made investment in AI in our IP portfolio. We made investment in chip-to-chip integration. You know, as the world is moving to what's called multi-die or chiplets, you need a way to connect these chiplets together in a system. So that's another investment. Automotive IP investment, they require different security safety protocols. So all these investments are made with a balance of by when do we see the return and revenue for these different opportunity, and when do you make that investment? In that context, when we think of the company trajectory, and Shelagh talked about the 100 basis point, as the floor as we're forecasting forward in terms of, improvement, we are calculating the opportunity of where do we invest in these segments and where's the growth for each segment?
That was really the core for our communication in terms of three market segments for the company. Because otherwise, it would be too difficult to communicate to our investors, where are we making the investment? Where's the ops margin, lumpiness is coming from? So now with the three segments, you will be able to see clearly over time, where is the revenue and the operating margin. Shelagh, anything you'd like to add?
Yeah, and if you think about the segments, a lot of what I talked about, the leverage and scale that we're driving, that accrues to each of the segments. That would be not market specific. But if you think about our segments, what we've talked about a Design Automation, operating margin is slightly above the corporate average. It's a software subscription business, so it tends to have that kind of economics. Our IP business is slightly below the corporate average over time because, again, that's a much more resource-intensive business, and those two businesses are quite complementary to one another. We're selling design tools and IP to the same customers. And then our Software Integrity business, we've been driving operating margin improvement as we've been driving scale on that business.
But that business is at a much lower scale than the other two. So the ambition over time is to drive that towards the corporate average.
We have another question? I told you, one gets it going, it's like dominoes. Go ahead.
Thanks for taking my question. So as the world is transitioning to accelerated computing, many are transitioning from an x86 infrastructure to Arm infrastructure. I know that you've worked or collaborated with Arm for decades now, but could you speak to how that might be, how that transition might be a tailwind for you, and how your relationship with or partnership with Arm may be advantaged relative to, say, like, Cadence's relationship to Arm? Ultimately, my question is, are you in a better position to capitalize on the Arm infrastructure transition?
Yeah, thank you for the question. And, our relationship with Arm dates back to when Arm started, because, in order for Arm to develop their own, CPU, they need design automation to help them develop their own architecture and implementation of it. As Arm grew to what it is today, which is a significant presence in the mobile market with good traction in data center, client application, and an ambition to expand beyond these two markets, we have a fantastic relationship with Arm, both from their own use of our software in order to help them develop their next generation product. And as importantly, the customer who's taking the IP from Arm, they need to figure out, "How do I integrate it inside my SoC? How do I verify it?
What kind of IP do I need to connect the Arm CPU to the rest of the chip?" This is where the Arm ecosystem becomes very relevant and important to a company like Synopsys, because we work with our joint customers to optimize both the integration of the CPU, connecting it together to the rest of the system, et cetera. So, we see fantastic opportunity there and continuing in that particular architecture. Now, you can take the same argument with an x86 or other CPU architecture. The same thing applies. Those customers need design automation, need IP in order to integrate that CPU into the rest of the SoC. Another vector that we started a number of years ago was today, our software runs on an x86 architecture.
So meaning, when our customers are running Synopsys design automation product, their cloud infrastructure, be it on-prem or in the cloud, it runs on a x86 CPU. Over the last few years, we expanded this to an NVIDIA GPU. We're working with other GPU provider to run on their GPU. We have it running on an Arm architecture, given our customers' desire, based on what their data center strategy is. So we have Arm from an ecosystem partner, as well as Arm from a compute in a data center type of an investment as well.
Okay. Oh, I had some questions-
Hi. Can you hear me okay?
Yes.
Yeah, Sassine, were you involved with the Software Integrity portion of the business? Because it seems like you guys have done a lot of acquisitions there. How do you view that portion of the business, and what kind of synergies do you see? Because it doesn't seem like there are a lot, but maybe you can actually highlight what synergies are, and how are you going to view that as you take the CEO role?
So we started the investment around the 2013, 2014 timeframe. The thesis behind it was, as we talked to our customers, every customer was telling us the number of software engineers, and those are the chip guys we were talking to at the time, because that was our customer base. The number of hardware, the software engineers are exceeding the number of hardware engineers. The hardware engineers meaning the chip developers, the chip designers. And as we were seeing that trend, that more and more software, expertise and investments were made at our customer, we entered into an investment and an acquisition of Coverity, which was a static checker of your software. It checks the quality of your software.
We were a user of that technology because we develop software as well, and we use that technology in-house for the quality checks of the software. Since then, I wanna say we bumped into security as the next strategic thing in software. Because more and more software content you have, vulnerability of the software becomes very critical. So how do you check the security of the software? We went from quality to security of the software, and we consolidated quite a bit. We made many, many M&As and consolidated the market, and the market in that space is called application security testing. It's very specific to the part of the life cycle of the software that is related to quality and security of the software.
Now, since then, we, even though we've had a certain ambition for growth and for operating margin of that business, the last year-ish, year and a half, I wanna say we've been disappointed with the performance of that business, and the disappointment's coming from two factors. One, there's a headwind that's facing us and everybody in that sector. The IT budget, as you know, anybody who's selling to the CIO of companies, they're under pressure to squeeze in the budget in the last couple years, and that's impacting our growth opportunity in that business. The other factor that we're not blind to, and we're, we're watching very closely, is the AI impact on software development.
If you think of an offering like GitHub, where historically it was a human-developed software that goes into that repository, right now you have significant amount of AI-generated software that goes into that repository. How will that change the viability and the growth in the context of AI-generated software for application security testing? So therefore, the performance over the last year-ish, year and a half, has been impacted by both the headwind, and some customers are taking a little bit of a pause in assessing how do they change integrating their software into the system. Now, that being said, the thesis of the importance of software security and software quality is only going to increase in importance, given the content of the software is increasing.
So that's, that's how we're thinking of that part of the business in terms of, part of the portfolio. The first two are very much tied to the chip design aspect. The third one is a higher level, and a different market that we're in.
Very clear. Okay, so I'm gonna—we've got about five minutes left. So I got a couple of questions. And really, I wanted to ask more about capital allocations. This is actually for both of you. You recently announced a $300 million ASR. You also announced the acquisition of PikeTec. So talk about how you balance priorities currently, and Sassine, also, how you're thinking about that going forward.
So, we're constantly, and I think this is our job as we look at the market trends around us. If we look at our customers in the chip world, they're moving from what used to be called scale complexity, which is following Moore's Law-
Right
... to a systemic complexity. They're trying to optimize at the system level. If you talk to any one of our customers, they're no longer thinking, "I'm creating a general purpose chip, and hopefully it will get sold.
Right.
They're trying to go up the chain to optimize the software stack that comes with their chips and how to integrate inside each application they're selling to. That's what we call system. We, like, the PikeTec acquisition is along the lines of Synopsys expanding to what we're referring to as Electronics Digital Twin. So if you think of a car, you may buy the chips from different chip suppliers, but if you are the OEM, how are you architecting the electronic system well ahead of you deciding where to buy the chip, which chip do you need for the car? Which functionality are you looking for? What do you want the electronic system to look like and be architected and designs, and designed?
We believe it's a great opportunity for Synopsys, given we have some technology in that space in terms of virtualizing a system, and how do we go from a chip to a system in our offering? And the PikeTec is one step in that direction.
Okay.
Maybe Shelagh can expand a little bit further on how we think.
Sure. Yeah, so I think obviously we've got huge growth opportunities, so our first and foremost, and when we think about capital asset allocation, is investing in the business, investing in the growth. And that manifests itself both in organic and then inorganic. And we've done a tremendous amount of tuck-in acquisitions over the years... and assessing just an outline PikeTec, that's gonna plug into capability we've already been building organically. So, since we've got such a vast, you know, platform, it's pretty straightforward for us to do inorganic investment, inorganic acquisitions and plug them in, and we'll continue to do that because, as there's so much, kind of, you know, innovation in the industry, we wanna take advantage of marrying our internal innovation with some external innovation.
The second thing we think about after investing in the growth of the business is really how do we provide some cash back to the shareholders? And we do that through our buyback program. We just announced the $300 million, as you said. So we have a regular buyback program that effectively offsets the dilution in the company. And then finally, you know, we are at a phase of the company where we're extremely high growth. I know I've had people ask me about dividends and things like that, but obviously, cash back to the shareholders through buyback program is our primary.
All right, well, I'm gonna sneak one last question in. It's a quick one, and it's more just as I think about the history of the EDA industry, historically, your customers were Intel and the big chip makers, and clearly, that's changed dramatically in the last certainly 10 years. And you mentioned systems approach, like, like, the systems makers, whether it's Apple or, or Google or Amazon, they're all developing their own chips. They're all developing their own systems. From a Synopsys standpoint, is there really that much difference when you're engaging with those kind of like, let's call them systems or cloud hyperscalers, when you're dealing with those customers versus dealing with the chip maker? I mean, is there much real difference?
It's different. It's different and not different.
Okay.
So, on one hand, the resources, the people that they're developing those chips at the hyperscalers, they came from the chips guys.
Okay.
Right? So, so it's the same talent pool, so they just move around to create that new initiative. However, the approach in which they're architecting and thinking of the chip, because they're not selling a chip-
Right
... they're consuming the chip.
Right.
So they're developing the chip with the intention to be very specific to their application, to their workload. Because when a chip company is developing a chip, they're thinking of the architecture in a way that they can sell it to more than one customer.
Okay.
When a system company is developing a chip, they're thinking about, "How do I wanna consume that chip for my specific workload and application?" So from a technology engagement point of view, it opens up a new opportunity. From the ones the chip design gets started, it's the same workflow pretty much that the chip semiconductor company have used.
It's nice to see the expanded TAM. It's just-
Oh, that's beautiful-
Yeah
... because that's a whole new set of customers for the company.
Well, we're at time. Let me thank Sassine, and congratulations-
Thank you.
—and Shelagh, and Trey, and Phil. Thank you all for coming, and I'm sure they'll be around later if you wanna catch up with them. Thank you.
Yeah, thank you. Thank you.
Thank you.