Okay, great. Hi, everyone. Welcome you to Mizuho Technology Conference 2026. We have a great pleasure to host Sassine Ghazi, CEO of Synopsys. Before I start, I was asked to read this. Today's discussion may contain forward-looking statements related to our current outlook, expectations and beliefs, which are subject to certain risk and uncertainty that could cause actual results to differ. Please refer to Synopsys' most recent SEC filings for a discussion of risk factors that may materially affect those statements. Good? Did I do a good job?
You did good.
Okay. All right. Before even I start, I just want to say that there's so much excitement in semiconductor right now. Just to give a quick intro, EDA companies, Synopsys, Cadence, they are kind of the backbone. They are the, like Jensen says, the indispensable companies or tools for building chip design. It's a great pleasure to host Synopsys today at our conference.
Thank you.
Sassine, I just want to kick off with this. You recently reported Q2 results, your fiscal Q2 results. Maybe it's great to refresh investors, like what's the latest with Synopsys and what's your puts and takes there, and how do you want investors to orient based on the results?
Yes. We just reported Q2. We had a fantastic quarter, where we did beat on our guidance on revenue, EPS, and OM. Actually, given the confidence with Q2, we raised the year on all metrics of revenue, OM, EPS, and free cash flow. It's our third quarter that we reported post the Ansys acquisition, which is a significant acquisition for Synopsys, and is expanding our opportunity dramatically.
Yeah. Ansys had a great quarter indeed. Let's get into each of this business. When you sift through your business segments, let's talk about some of the key investors' focus. One of the key metrics is core EDA growth. Last, first few quarters, it seems to be softening below 10%, like getting to single digits. What's the things that investors should keep in mind about the growth? The point I want to make that how do you get back to the durable double-digit growth that investors are looking for in EDA?
Yeah. The Synopsys portfolio, if you look at it, we have really three main categories of the portfolio. You have EDA, IP, and the Ansys portfolio, we call it simulation and analysis. We are the leader in EDA, we are the leader in interface IP, and we are the leader in simulation and analysis. In core EDA in particular, it's very difficult to compare a quarter versus a quarter. The way we look at it is a trailing 12 months.
The reason for that, you remove a lot of noise, because inside EDA, there's a hardware business. The hardware business is very lumpy because of the nature of the revenue accounting. You take revenue upfront versus EDA is a very ratable business. The first half of 2026, we did deliver double digits. It depends on the comps that you have. Since the acquisition of Ansys, I find myself, I'm doing a lot of explanations, which I don't like to spend the energy on explanations, especially when it relates to accounting. We had to divest number of technology.
Yeah.
We kept the Ansys part of the portfolio as a separate from EDA, because part of the Ansys portfolio has EDA in it. Therefore, when you're comparing to the market, you're not comparing apples to apples. Our commitment is double-digit growth for EDA, with opportunity to expand beyond with the AI monetization potential that is changing the workflow in EDA. At this stage, and at this point, our long-term, for EDA being double digits, we remain committed to that.
Is there any kind of growth driver that you can cite, like to give that confidence, double-digit growth, specifically design start activities?
Actually, in EDA, what drove our growth has been complexity.
Yeah.
If you go back a decade ago, core EDA was growing around the upper single digit. It expanded into double digits, and we've had years where it was well into the double digits, driven by many factors. The design starts, the complexity of design. Our customers cannot design these advanced chips w ithout the participation of EDA. Now these chips are turning into systems. They are not monolithic chips, they are 3D IC, that require bringing physics into the electronic design. That was one of the key thesis behind the acquisition of Ansys.
How do you accelerate the workload, as you're running a massive simulation, et cetera? There are many layers of monetization opportunities. The part that I'm very excited about is the impact of AI on the workflow of building a semiconductor chip. That's a significant change that is happening as we speak. Given the rapid exponential innovation, in foundation models.
Yeah. Certainly, we'll cover AI later on, I want to cover next other segment, which is interesting, is IP business. That has its own set of challenges over the first few quarters, as you know that. How can investors now feel more confident that worst is behind us and it's a normal trends going forward?
What we said is Q1 was the bottom.
Yeah.
What you need to expect for the rest of the year is sequential growth from the Q1 bottom.
Which you delivered.
Which we delivered in Q2.
Yeah.
We're committed -
Yeah
- to deliver for the rest of the year. The part I'm most excited about in the IP business is driven by the market dynamics. What Synopsys has done, we built that business organically. Think of it as a factory. We build an IP based on a standard. We build it once. We sell it many times, and it's a great business model. In the last three, four years, the big change that's happening at our customer base, especially hyperscalers, they're looking for different compute alternatives in order to optimize their TCO. There's a big trend not only to buy merchant chips or ASIC, is building their own chips, COT. The COT cannot happen without Synopsys IP. I know this is a big statement. COT cannot happen -
Yeah.
- without Synopsys IP. There's an opportunity for us to monetize COT at a completely different rate than we've had in the past, which is build it once, sell it many times. Why? In COT, in order to make that chip competitive -
Yeah.
- you need to customize the IP. It's no longer, even though it's a standard, you need to customize the standard and deliver it before the standard is finalized.
We know how to do it. We have the scale to do it. We have the skills to deliver to it. We're in active discussions with many of these hyperscalers to change the business model from a use fee only to a use fee and a royalty so we can participate in the upside. I cannot be more enthusiastic and excited about that part of the business that we have.
That's a great point, like you said, COT cannot happen without Synopsys IP. How should we think about this royalty model that you talked about, which is kind of incremental revenue to what you get right now, when should we start thinking of that flowing into your P&L?
Yeah, given the number of changes happening in our portfolio, and it's really the year one of the Synopsys Ansys integration, we have an investor day coming up September 30th, where we map out exactly the question that you mentioned. If at the highest level, think of the Synopsys IP will have two business models. Think of it as factory one and factory two. Factory one, build it once, sell it many times. That's a fantastic business. We do very well in that business. That will continue. We're building factory two for the customized IP. Any IP that's customized, which is pretty much every AI hyperscaler, COT requires that customization, will move to factory two, which is a new business model.
There, the license fee, that's how we've been monetizing. The royalty upside should be fairly meaningful in upside in order to justify even us talking about it or building that factory.
Yeah.
We'll share more what will that look like, not only for 2027, but for the next few years in terms of the revenue growth participation.
Yeah. I want to touch on other drivers as well, because one thing you mentioned that not only IP will bounce back this year, but next year it'll get back to the corporate growth rate. Royalty definitely one of the thing. What are the other drivers we should pay attention to?
The demand for IP is there.
Yeah.
Actually, the fundamental demand for IP is chip start.
Yeah.
In semiconductor right now, there is no shortage in excitement or the need to deliver customized silicon for different applications. That's for the applications of today. As you look toward the future, more physical AI coming requires more customization of chips, meaning more chip start. That IP opportunity is significant. We are the leader by number of factors compared to the number two in that space, and that gives us the opportunity to scale with that leadership position. That factory one- we've always communicated it will be in the mid-teens -
Yeah.
- and I have full confidence that will be in the mid-teens. You layer the factory two growth, which should be higher than factory one and achieve the growth at that level.
Okay, great. I think we'll now switch to the next level, next topic for investors at Intel.
Yes.
Intel commentary. I know you guys, it's a big vendor for Intel, or Intel is a big customer. Maybe talk about the background of your Intel relationship. I know you are the man behind it, and how it has evolved these years, and any meaningful changes in the Intel strategy or relationship post new leadership or even the restructuring. Intel has gone through a lot of changes last year and also a lot of excitement coming on the Intel side as well.
Yes.
Walk us through on that opportunity.
Intel has a special spot in my heart for me, because I started my career at Intel, and when I came to Synopsys, within four or five years of my early days at Synopsys, I was assigned to Intel to be the sponsor of the relationship for Intel. I remember around 2005 or 2006, we introduced the concept to the industry of a primary partner, Intel and Synopsys. I remember at the time, it was the first press release in the industry that Intel chose Synopsys as its primary partner. What that meant is consolidate from Intel internal tools, because Intel had a lot of internal CAD tools to Synopsys, over time, continue on consolidating from other suppliers to Synopsys.
That relationship has been an outstanding relationship. Over the years, as you can imagine, Intel had many management changes, many stress points. We've went through many renewals, almost every renewal, the discussion on the spend level, the market change dynamics, et cetera, was always in the background. Now I know with Lip-Bu as the CEO of Intel and brought in some management from Cadence, the constant discussion is there's going to be a bias, you're going to be out of Intel. That's not how the industry works. Customers don't use us because there's a relationship. Relationships and trust are important. They use us because of the technology we provide. You cannot design those chips without Synopsys.
Yeah.
Cadence has great technology, so does Synopsys, there is no advanced chip that will be done without Synopsys. When you now get back to the Intel dynamics, because I hear that question all the time, the relationship with Lip-Bu, the relationship with Srini, the relationship with the entire Intel management is very strong. There is no direction of nudging or pushing Synopsys out. When I get that question, I'm in the details. I look at market share, market share drives revenue share. There is zero change. The negotiation will happen whenever the time is up for the renewal to happen, based on the Intel dynamics, the negotiation will take whatever shape, which is no different than any other customer negotiation. There's the other part of Intel, which is Intel Foundry. There, I can even be stronger in my message. You cannot be in the foundry business
You cannot be in the foundry business without Synopsys.
Yeah.
You cannot on-ramp a customer to come to your foundry without Synopsys IP and foundation IP. We're the bridge. We're the on-ramp from customer to foundry. We have had, and publicly we announced a number of years ago, maybe three, four years ago, the relationship with Intel Foundry on 18A. We built our IP on the technology. We have an ongoing engagement with Intel Foundry on IP. I hope that clarifies -
No.
- the -
That's super helpful.
- the doubt that is always in the background when Intel comes up.
I remember your comment saying, "I'm not losing my sleep on Intel.
I'm not, because when I lose sleep is when we're not able to deliver to the customer.
Yeah.
Which is very rare. In the Intel case, not only we're delivering, we have a great partnership and engagement.
Yeah. Let's switching to the AI and the topic du jour. I know it took some time, moving to this bright AI future ahead of the industry, we're moving through three phases of AI compute. One is training at scale, training and inference at scale, now physical AI, all three. How does the progression translate into design start and even EDA tool consumption, and how Synopsys benefits in that whole AI transition?
Yeah, maybe quick comment on overall design start in semiconductor, because we track design starts very carefully, and we have the visibility pretty much on every chip that starts. In semiconductor, pretty much any company that is serving the AI infrastructure, we're seeing continued increase in chip start. What it means, if you're company A and you have four chips per year on your roadmap, you're seeing it going from four to possibly five, possibly six, which is great. You have the rest of the semiconductor industry that's not participating in AI infrastructure, we've seen a decline in chip start.
The last two quarters, based on our data, it seems like it hit the bottom. We have not seen yet the turn and the increase. Net in semiconductor, when you combine the both, the aggregate, is that chip start increase, but not at the same rate as the AI infrastructure, semiconductor for AI infrastructure. For us, chip start is very good because you sell a lot of IP, EDA, and hardware to design these chips. Now, when it comes to AI and lowering the bar for our customers to design these chips, because you have to remember, in our industry, one of the bottlenecks have been access to talent. The shortage of talent is real. When you look at hyperscalers with massive money. They're trying to build their semiconductor team. It takes them multiple years to do it, just because of the skills required to build that team.
AI is starting to change the workflow of how to build a chip, which is a huge opportunity for our customers to be able to build more chips, you can argue with the same number of people. Reduce the time to deliver a chip and the cost to deliver a chip. That will happen only with us evolving the workflow and the way our customers using our products. We have many initiatives there in terms of, not only initiatives, actually, roadmap items, delivery to the customer to bring in pretty much the agentic stack and the orchestration of these agents for our customer.
Yeah. The agentic AI engineers, agent engineers opportunity.
Yes, because there is definitely going to be a mix of human engineers running our product to design the chip, and agent engineers running our product to design the chip. From a business model point of view, for at least two decades right now, the industry moved to subscription license for human use. There will be another layer, which is consumption of agent use of the software.
Yeah.
That's going to be an incremental opportunity to monetize the agent use of the software. There's a third layer that is starting to show up where there are some AI labs are very interested in fine-tuning their foundation model for semiconductor applications. Because they believe there's a market opportunity for them, and there, the discussions with these labs is percent of token revenue upside, because the good news, if you look at it this way, our tools are going to be used under the hood regardless what model you're using because you need that solver, physics-based solver in order to train the model and inference the model.
Yeah. I'm glad you brought it up. Just bring a promotion here that yesterday we published a deep dive note on this thematic report, how agentic AI engineers expanding the EDA TAM from going from tools to the labor.
Yes. Exactly.
This is a big opportunity. Anybody wants to go dig deep into that, we have all the numbers that Sassine is not sharing here. Switching to the next thing is Ansys. Ansys, one of the key event happened last year and making the progress on that. The question always is we get, is one plus one better than more than two?
Yes.
That's kind of the questions. Help us understand, what's the early customer feedback? I know you launched some of the products, the combined product, earlier this year at Converge. What kind of feedback you are getting, and what's the opportunity there?
Our thesis on the acquisition of Ansys was anchored on the semiconductor chips, especially for AI applications, are moving from a monolithic chip to 3D IC. The moment you start stacking chips on top of each other, you're dealing with mechanical challenges, not only electronics. The first priority of the Ansys Synopsys integration is delivering a joint solution where we take the Ansys multi-physics simulation leadership position into the electronics digital design flow of Synopsys. By fusing that technology early in the electronics chip design, if you're a customer heading in that direction, you'll have a convergent flow, meaning when you're designing that sophisticated chip, you can predict what thermal issues you're going to have, what structure issues you're going to have.
We released the first wave of products in March at the Synopsys Converge. We are actively engaged with number of customers in evaluating the technology. I know for a fact there will be number of customers they'll move from an eval into a production use of the technology. Once the customer commit to production use, it means one plus one is greater than two from a revenue point of view. Most of these customers are already a Synopsys customer and they're already an Ansys customer. The new solution is a new product, new SKU that is priced differently regardless if you have one plus one.
Yeah.
Therefore, it will be greater than two.
It's not the first time you're bringing this. You did that Fusion Compiler before.
Yes.
How does it compare in terms of traction wide, what sort of pricing uplift you saw when you did that Fusion Compiler -
Yes.
- model?
Actually with Fusion Compiler, thank you for bringing this up, it brought in timing sign-off, power sign-off into the Fusion platform. The value we saw beside the uplift on pricing was a market share gain. We gained share because the customer could see better value than using a discrete product. They used the fused product. In this case, we're bringing multi-physics fused into the digital platform. We will win in two ways. One, one plus one greater than two opportunity, the uplift, and there will be a market share gain given the solution will be of higher value to our customers.
Okay. I have one more question. I will open it up for Q&A from audience after this. The path, I think the margin is the next kind of focus. How does the path from current 41% to your mid-40% operating margin plan look like? What are the specific levers that get you there in terms of priorities and ranking? You talked about synergy, scaling, AI, and all that. How does that give you that path to mid-40%?
On operating margin, actually this year.
I know I asked the question, Shelagh's questions.
That's great.
Yeah.
Shelagh and I were a partner on that mission. On the operating margin, this year we're going to deliver more than 300 basis point increase in our operating margin. It came through a number of areas of focus. One, portfolio. There are areas in the portfolio we decided to exit. We announced, around September timeframe, a 10% reduction in our workforce and accelerating the integration synergy of $400 million from year three to sooner, to earlier. Of course, the best way to continue on improving that ops margin is by focusing on the revenue growth opportunity and delivering on the innovation opportunity. The other point that often comes up in these discussions is your operating margin has been below your closest peer. It's true, but we have to remember, it took us 20+ months to close the Ansys acquisition.
During that period, we had very little flexibility to make changes in the portfolio or in the workforce. We are absolutely committed to get to the mid-40s and continue on expanding the margin as we drive the top line.
Okay. Let's see if there is any question. Yeah.
Thanks, Sassine. As you see your clients adopt agentic AI, what are the key challenges that they are talking to you about in the new paradigm here, an agent could design the entire chipset, entire system? Thank you.
There is part of the chip flow today that looks very similar to software before you move into the physical implementation of the chip. That first stage, which is called the front end of the design, where you provide a spec, then you do number of steps to create an RTL, which is the language before you move into the physical implementation. We announced a product in March that is multi-agent adaptive orchestration, where there's a cognitive layer calling multiple specialized agents to go from a spec to a verified RTL. That part of the design looks very similar to the software, we're seeing nice progress there. The moment you go into implementing the chip physically, that means which foundry are you using, which library from the foundry and PDK, et cetera.
That's where the physics solvers become essential to make sure whatever you're speccing will be able to manufacture. Almost every one of our customers is looking at the hybrid workforce of a human engineer with an agent engineer, and how to balance not only the work, the cost, because an agent engineer is not for free. There is a cost associated with the tokens that they use, the compute they use, the licenses they use, and the human engineer in the mix. In the Synopsys case, as you get to the final stage, we have the leadership of sign-off, be it from the Ansys portfolio or the classic Synopsys portfolio that is essential before you commit to the chip to go to manufacturing, regardless if it's designed by a human or agents or the combination of the two.
Sassine, a big announcement we heard recently was that with your partnership, if we look at Jensen Huang's words, physical AI could be a $40 trillion, $50 trillion opportunity. Synopsys is a participant in that through Ansys. You have a partnership with NVIDIA to accelerate workflows, and we saw a slide thinking within reasoning robots, that Synopsys support a huge budget along with that. It'll be great to see a vision around that in terms of how you're working with NVIDIA in advancing physical AI. At the same time, does that mean you're competing with NVIDIA? You're partnering with NVIDIA? How does that whole relationship with NVIDIA work out?
That's a great question. Thank you. The partnership with NVIDIA has three layers. The bottom layer is accelerated compute. Any place where we can use GPU to have our product run faster on a GPU, that's the base layer of the partnership. We've been working with them for a number of years on that base layer. The next layer up is the agentic orchestration layer. The third, the top layer, is the physical AI, where we're connecting our Ansys portfolio to Omniverse. Omniverse can be the design platform where it orchestrates multiple technology to envision that future physical robot or car or a drone. At the end of the day, no one will commit these products before doing tons of high-fidelity simulation in order to ensure it's going to be safe, secure in the real uncontrolled world.
That's our role, that's what NVIDIA saw in the value that Ansys brings into this partnership to accelerate. Definitely, we're not competing with NVIDIA. We're partnering with NVIDIA. Sometimes there's confusion when Jensen or NVIDIA talks about physics simulation. NVIDIA has had physics simulation for a very long time. If you see the impressive games, the car crashing into a wall, or I'm not a gamer, or whatever visualization you see, it looks so real because there's physics simulation in it. That's not the same physics simulation of if you're a BMW and you're building an actual car and you want to do physics crash testing, that's the highest fidelity simulation that Ansys owns in that market. That's where it's a complementary partnership. When you hear that NVIDIA or other, they have physics simulation, it's true. What we have is the sign-off physics simulation, which is the highest fidelity.
Okay. Since we have a couple of minutes, maybe my last question, Sassine. You have come up through engineering and sales at Synopsys decades, and you have seen Synopsys as a leader in EDA for decades right now. The last 12 months have been exciting as well as challenging. You closed one of the largest deals, Ansys, finally made it happen. We saw some IP challenges, we saw China, restriction there, Intel restructuring, all that. As you talk about the new Synopsys, new Synopsys, what are the most important things you want investors to take away with about the new Synopsys Synopsys, whether financially, strategically, or culturally, like where you are taking Synopsys to?
Yeah, our mission is to empower innovators to drive human advancement. I've been at the company now for 27, 28 years, I've been so fortunate to see the company going through multiple stages of evolution and growth. This year, by the way, is our 40th year anniversary. In 40 years, we've had a number of bumps along the way, the vision was very clear. How do we continue leading with innovation to drive human advancement, enable our customers with our technology to drive the product that they're building? Nothing changed there. As you're pointing to the challenges of the last year and a half, when we set into the new company, the new direction, we knew we're going to do something very big.
When you do something very big in an environment with a lot of stress, restrictions on China, changes in the industry, et cetera, we did hit some bumps. I have no doubt that we will continue on leading in the portfolio, both in revenue growth and be disciplined in our financials to have the highest return for our investors and, most importantly, for our customers to be dependent and leaning on Synopsys to drive their innovation.
Double-digit growth, margin expansion.
All the way.
All the way. Well, looking forward to the Investors Day in September, and looking forward to hosting you again next year at this conference.
Thank you.
Thank you so much.
Thank you.