Ladies and gentlemen, welcome to the Synopsys Earnings Conference Call for the Second Quarter Fiscal Year 2025. — Operator Instructions — Today’s call will last 1 hour. As a reminder, today’s call is being recorded. At this time, I would like to turn the conference over to Trey Campbell, Senior Vice President, Investor Relations. Please go ahead, sir.

Good afternoon, everyone. With us today are Sassine Ghazi, President and CEO of Synopsys; and Shelagh Glaser, CFO. Before we begin, I’d like to remind everyone that during the course of this conference call, Synopsys will discuss forecasts, targets and other forward-looking statements regarding the company and its financial results. With these statements – while these statements represent our best current judgment about future results and performance as of today, our actual results are subject to many risks and uncertainties that could cause actual results to differ materially from what we expect. In addition to any risks that we highlight during the call, important factors that may affect our future results are described in our most recent SEC reports and today’s earnings press release. In addition, we will refer to certain non-GAAP financial measures during the discussion. Reconciliations to their most directly comparable GAAP financial measures and supplemental financial information can be found in the earnings press release, financial sup- plement and 8-K that we released earlier today. All of these items, plus the most recent 1 investor presentation are available on our website at www.synopsys.com. In addition, the prepared remarks will be posted on our website at the conclusion of the call. With that, I’ll turn the call over to Sassine.

Good afternoon. We had a strong second quarter with revenue up 10% year-over-year, exceeding the midpoint of our guidance and non-GAAP EPS was above our guided range. We are reiterating our revenue guidance for the full year as these results demonstrate the strength of our products, which are mission-critical to our customers’ innovation, the resiliency of our business and relentless execution by our global team. I’ll provide more details about the quarter, and then Shelagh will delve deeper into the financials. At the macro level, the tale of two markets persisted in Q2. Despite market fluctuations, the AI and HPC sectors remained robust. And while we’re seeing signs of stabilization in industrial and automotive, non-AI end market demand remains subdued. For Synopsys, a slowdown in China was offset by strong demand from customers in other regions. The megatrends of AI, software-defined systems and silicon proliferation continue to drive our growth. These trends are increasing design complexity and costs while also increasing compute performance and energy demands. Synopsys benefits as a missioncritical partner in addressing these challenges and as the industry leader in applying AI to help customers innovate faster. Our pending acquisition of Ansys will address the need for new AI-powered silicon to systems design solutions, integrating electronics and multiphysics. We have regulatory clearances in all jurisdictions other than China. We are working cooperatively and actively negotiating with SAMR to secure China regulatory clearance, and we continue to anticipate closing in the first half of this year. 2 Now let’s move on to the business highlights. Design Automation demonstrated resilience with revenue up 6% year-over-year. Our new hardware-assisted verification products, HAPS-200 and ZeBu-200 are off to a strong start. These systems offer the highest performance and ultimate flexibility between prototyping and emulation. In EDA, the industry’s growing adoption of multi-die architecture plays to our strength and leadership position. In Q2, we supported multiple active production deployments with a leading HPC AI chip maker, including delivering what we consider the most complex 3D heterogeneous integrated design with over 40 chiplets and advanced packaging technology. We also displaced manual high-bandwidth memory layout flows with Synopsys automated 3DIC compiler implementation at a top-tier Asian semiconductor customer, achieving best-in-class productivity and quality of results improvements. The race for performance is driving adoption of leading-edge process nodes and Synopsys is proud to help usher in the Angstrom era with our foundry partners. In Q2, we enabled the industry’s first 2-nanometer-based HPC design and delivered multiple successful test chips across sub 2-nanometer process technologies. And the leading AI capabilities we’ve pioneered across the full stack are generating wins among both semiconductor and systems customers. In Q2, DSO.ai momentum continued, driving multiple design wins for flagship CPU and GPU cores, while a major AI infrastructure customer began large-scale deployment of VSO.ai across five projects. Additionally, our AI capabilities are winning analog designs with major automotive Tier 1 in Japan adopting Synopsys ASO.ai after extensive evaluation. Turning to design IP, where revenue increased 21% year-over-year as customers rely on Synopsys IP to minimize integration risk and speed time to market. Our leading foundation and interface IP also expedites customer adoption of the latest protocols and leading- edge process nodes. 3 Driven by AI and the need to transport more data faster, we are seeing strong demand for high-speed SerDes IP with Synopsys 224 gig PHY securing multiple competitive wins in Q2. AI accelerators and GPUs necessitate ultra-efficient networking infrastructure and Synopsys is the first mover in PCIe 7.0 with seven unique customer wins and the clear leader in UALink with over 20 customer engagements. Our recently announced support of NVIDIA’s NVLink Fusion ecosystem will further enhance scale-up optionality for AI factories. Before wrapping up, I want to thank the many customers and partners who joined us for SNUG Silicon Valley and our inaugural Executive Forum in March. There, we showcased our generative AI-powered assistive and creative capabilities, which are unlocking new levels of efficiency for customers. For example, what previously took hours searching documentation or waiting for expert help can now take minutes with Synopsys.ai Copilot assistive capabilities. And using Gen AI to generate design collateral like RTL or Testbenchs is helping early access customers accelerate design and verification cycle times from days to hours and hours to minutes. Synopsys leading Gen AI capabilities are necessary foundation for the paradigm shift that comes next. Agentic AI will transform engineering workflows, allowing R&D teams to focus on important architecture and design decisions while tasking agent engineer technology with implementation details. It’s an exciting time. Engineering is undergoing un- precedented transformation and Synopsys is seizing the opportunity to reengineer engineering. A few closing thoughts. Our business model is resilient, and our solutions are essential to our customers’ innovation. We have steady momentum across the business, supported by growth trends. Synopsys is leading AI for chip design, and we are investing to maintain and extend this leadership position. Thank you to our employees, customers and partners 4 for a strong quarter and for your continued commitment. Now over to Shelagh.

Thank you, Sassine. We delivered a strong Q2 with revenue of $1.6 billion, non-GAAP operating margin of 38% and non-GAAP EPS of $3.67. Backlog came in at $8.1 billion, up $400 million quarter-on-quarter. These results reflect our leadership position, consistent execution and resilient business model in a market fueled by the secular megatrends of AI, software-defined systems and silicon proliferation. Despite a dynamic macroeconomic environment, we are reaffirming our full year 2025 targets for revenue and non-GAAP operating margin and updating our EPS and free cash flow guidance to account for our Q2 results and bond issuance. I’ll now review our second quarter results. All comparisons are year-over-year unless otherwise stated. We generated total revenue of $1.6 billion, up 10% with strong growth in Design IP. Regionally, we saw strength in Europe and South Korea, offsetting China headwinds. Total GAAP cost and expenses were $1.23 billion and total non-GAAP costs and expenses were $995 million, resulting in non-GAAP operating margin of 38%. GAAP earnings per share were $2.24 and non-GAAP earnings per share were $3.67. Non-GAAP earnings included a $0.28 benefit from the sale of a building as well as approximately $0.06 of net charges associated with the $10 billion bond issuance in Q2. These items were not included in our prior guidance. Now on to our segments. Design Automation segment revenue was $1.12 billion, up 6% against a strong compare. Design Automation adjusted operating margin was 40.9%. Design IP segment revenue was $482 million, up 21%, with strong performance from Interface IP. Design IP adjusted operating margin was 31.2%. 5 Moving to cash. Free cash flow was approximately $220 million. We ended the quarter with cash and short-term investments of $14.3 billion and debt of $10.1 billion. Now to guidance. Full year revenue and operating margin targets remained unchanged from the prior guidance. For fiscal year 2025, the full year targets are: revenue of $6.745 billion to $6.805 billion, total GAAP costs and expenses between $5.01 billion and $5.07 billion, total non-GAAP costs and expenses between $4.05 billion and $4.09 billion, resulting in non-GAAP operating margin of 40% at the midpoint, non-GAAP tax rate of 16%, GAAP earnings of $10.14 to $10.34 per share, non-GAAP earnings of $15.11 to $15.19 per share. Our non-GAAP EPS targets have been increased from the prior guidance to reflect the Q2 outperformance, partly offset by net interest expenses associated with our bond issuance. Cash flow from operations of approximately $1.5 billion and free cash flow of approximately $1.3 billion, lower than the prior guide due to financing and acquisitionrelated costs. Now to targets for the third quarter. Revenue between $1.755 billion and $1.785 billion, total GAAP costs and expenses between $1.27 billion and $1.29 billion. total non-GAAP costs and expenses between $1.06 billion and $1.07 billion, GAAP earnings of $2.63 to $2.74 per share and non-GAAP earnings of $3.82 to $3.87 per share, which includes a $0.13 impact from bond-related costs. Our press release and financial supplement include additional targets and GAAP to nonGAAP reconciliations. In conclusion, we delivered a strong Q2 and are poised to deliver a strong second half. Our confidence reflects our relentless execution and leadership position to take advantage of the secular megatrends driving the semiconductor industry. With that, I’ll turn it over to the operator for questions.

6 — Operator Instructions — Our first question today comes from Lee Simpson from Morgan Stanley.

Operator, before you turn it to questions, I would like to make a quick comment because I’m sure it’s on everyone’s top of mind. I want to acknowledge that we are aware of the reporting and speculations, but Synopsys has not received a notice from BIS. So our guidance that we are reiterating for the full year reflects our current understanding of BIS export restrictions as well as our expectations for a year-over-year decline in China. So I want to put this upfront as we go through the questions. And given the news happened in the last few hours, just to acknowledge that we have not received the letter as well as our guidance that we shared with you reflects the current understanding of BIS export control. And with that, please open it up for questions.

Thank you, Sassine. And we’ll go to Lee Simpson from Morgan Stanley.

Great. Maybe just vectoring off of what you said there at the end. We note that China sales now are 10% of the mix, I think 12% in Q1, and you’re now talking about decline in China for the year. So are we to see this as probably a high single-digit percent in the mix? And I think the added question here, I suppose, is if that’s the sales portion, do we think there’s a group average similar margin for the China business? In other words, the impact bottom line could be the same if indeed the rumors of a BIS change in export controls come through? 7

Yes. Thank you, Lee, for the question. Recall as we started sometimes in FY ’24 communicating that we are seeing both a cumulative impact of the restrictions in China as well as the macro situation inside China have caused us to continue on communicating that this deceleration will continue and that headwind has gotten stronger as we go through each quarter over the last year, 1.5 years. Now halfway through our fiscal year by us reiterating the year, we are taking into account that China will be declining year-over-year. So FY ’25 to FY ’24 there will be a decline in China’s revenue. But despite that decline, we are reiterating the full year guidance. And that’s due to great execution in other regions and strength in the portfolio that we feel confident with reiterating the full year guidance. As far as the other part of your question regarding impact on operating margin, et cetera, et cetera, we cannot speculate about any potential impact to a notice that we have not received.

Okay. That’s pretty clear. Maybe just a quick clarification question maybe for Shelagh. Obviously, we’ve seen the $10 billion raise – and I think as part of that, your first payment appears 1st of October of this year. But I suppose rather than being a clean semiannual coupon, it may include some of the catch-up for the first half month because the bond settled on March 17, but we didn’t make the payment on the 1st of April. So it would be 6.5 months payment, if that’s the right way to look at this?

Yes, it’s a bit of a catch-up, and then we’ll get into the regular rhythm of the biannual payment.

And the next question comes from Jason Celino of KeyBanc Capital Markets. 8

Thank you Sassine for the clarification upfront. If it’s okay, just a few prodding type of questions. Maybe one, as it relates to your revenues in China, are you able to speak to the mix? How much is software, how much is IP and how much is hardware?

It’s – we don’t split per region the revenue, but you can say it’s similar to what we have in the rest of the world.

Okay. And then secondly, in 2019, when we had the original ban on Huawei, there was no expense impact because I imagine whatever variable expenses you had were redeployed to other customers. Is there any way to conceptualize how variable your expense base is, kind of a poorly worded question, but that on like the overall business, not specific to a region?

Jason, it’s difficult really to do that because, as you know, we have the global operations. Our team, not only in China, around the world is a mix of R&D, field, et cetera. And at this point, there are no changes per se to our operations anywhere, not only in China. So it’s difficult to really answer that question.

The next question is from Vivek Arya from Bank of America.

This is Liam Pharr on for Vivek. On your largest customer, they have been reducing R&D spend. Is that still a growth area for you? And how do you see market share evolve at that largest customer versus peers? And are there any areas that are more open to 9 competition versus others?

Yes, I’m assuming you’re talking about Intel in this case. The as we have communicated again that Intel has a mix of EDA software, IP and hardware. These are multiyear committed agreements. And same as with other customers, as their road map may be fluctuating or there are some rethinking about the road map, it does not impact generally the EDA software. There might be some impact on hardware and IP pull down, even though those agreements are committed, noncancelable, there might be quarter-over-quarter fluctuation, but that’s really about it at this stage.

And then just a quick follow-up. On the Ansys deal that is still pending, if in the scenario that the deal gets pushed out or doesn’t close, what’s the plan B? And how will you address the need to have more of the system design capabilities organically compared to your peer Cadence that’s kind of been developing that for the last couple of years since you announced that Ansys acquisition?

Yes. This transaction, as we have communicated, is so consequential not only for Synopsys and Ansys but also for the industry. The support we got from our customers in – related to this transaction directly to regulators that it’s a great opportunity to accelerate innovation is something that is essential for us and for our customers and the industry to focus on the completion of this transaction as really the only scenario we are considering and assessing at this stage. And the reason the confidence is high that this is the scenario for us to consider is if you look at the approvals we have received across all jurisdictions and currently, as I mentioned in my prepared remarks, we are in active negotiation with SAMR as well as antic- 10 ipating closure in the first half of this year. This is really the option we’re focused on at this stage.

Siti Panigrahi from Mizuho is up next.

Now I’ll switch back to the design activity. You talked about strong demand on the AI side. Sassine, I’m wondering what kind of design activity you are seeing on the non-AI customer base, especially have you seen any kind of changes there?

Yes. Thank you, Siti. Good question. As we have been communicating the tale of two markets, on one hand, you have – and I’m talking in particular, semiconductor customers, the AI, HPC infrastructure build-out remains very strong. For a while, anything industrial automotive was declining in terms of road map at our customer, even though their R&D is stable, but the new chips or accelerating design has been muted for a while. Now in industrial and automotive, we actually see more stabilization and new energy and vitality happening in that market. And where we see it first is in our IP portfolio because typically, this is when the customers are thinking about the next chip. next customer, they come to us to talk about, here’s what I’m designing for this particular foundry for this particular application, and we’re seeing that pickup in industrial and automotive, which is a very good sign.

Okay. And quickly, when do you think all the AI customers that’s slowly looking at building their custom chips, when do you think they will come to the label of what you say, hyperscaler? 11

You’re talking about hyperscalers building their own chips?

Or even the AI companies? Or do you see them building their own chips? Do you see that as an opportunity basically expanding design beyond hyperscalers?

Absolutely. Yes, yes. Absolutely. It’s been a fantastic opportunity actually because there are – think of them as three categories. The first one you’re building a data center, you get a merchant chip, you know who to go get it from. There is a category two, which is driving a significant ASIC model type of business. They define an architecture and to some level, they have an investment in designing the chip. And with multi-die, by the way, there are many, many options there. They may design the AI accelerators themselves but they get the GPU, CPU, the I/O somewhere else. So think of that as the ASIC opportunity. Then there’s the third one, and it’s at different stages of maturity, is designing the whole AI system themselves. For all three of them, it’s a great opportunity for Synopsys because you buy a mix or all of the above for EDA, IP and hardware. The opportunity for hardware there is fairly significant. because even if you’re not designing your own chip and you’re using an ASIC model, you still need to verify that chip in the context of your software. And this is where Synopsys has a sweet spot with HAPS-200 and ZeBu-200, which is performance-based to bring the software up before the chip is available. So absolutely is an opportunity for Synopsys, Siti.

Gianni Conti has the next question from Deutsche Bank. 12

So the first one would be, could you share any color at this stage looking ahead to the next renewal date on whether you’re seeing more of an increased opportunity or a potential risk from the Intel overweight? Has there been any recent developments that point to a more optimistic scenario where R&D spend increases or concentrates more towards EDA tools versus people? I guess not just for Intel, but broadly even across customers and across the range in semis.

Yes. Gianni, remember, where does Synopsys see an opportunity? We see an opportunity when customer is pushing their road map, designing more complex systems and chips because the more you’re pushing that limit, it opens up an opportunity to sell new stuff to the customers. If a customer has a stagnant road map is one thing, they still – when a renewal comes up, there is always an opportunity to increase the run rate for that customer for various reasons. But when a customer is investing by going to the next technology node, by developing a new chip for a new market, by adopting a new methodology, multi-die, et cetera, that’s a new opportunity for Synopsys to say – to sell the new EDA software, IP to connect these chips together or to serve these new markets and of course, the hardware. So any time a customer continue investing in their R&D to serve these road map items is an opportunity. And that’s what deliver to the resiliency of our business.

Okay. That makes sense. Just a quick follow-up. Sorry to reiterate this question, but I wonder if you could share a few more words on what exactly is giving the confidence in closing the Ansys deal in H1, given that your only roadblock is China. I mean, can China not reply and keep this going on for quite some time in a similar fashion to — indiscernible — 13 with NXPI. I’m just curious because we’re about a month away from the deadline. And of course, everyone agrees that this is a fantastic opportunity for Synopsys. So – and there’s no overlap, of course. So I’m just curious to hear – you seem very confident and it’s great. So I just want to hear your puts and takes on that.

Yes, Gianni, the best way to answer it is because we’re in the middle of it, meaning we are in the middle of the discussions, the negotiation. And you just simply look at Synopsys clearing other jurisdictions in a phase one approval due to the merits of the deal. It’s no different in our discussions with SAMR and in China. The back and forth in our negotiation with SAMR is giving us the confidence that we are committing to the first half closure based on purely the merits and the current conversations we’re having.

The next question today comes from Jay Vleeschhouwer at Griffin Securities.

Shelagh, let me start with you. Taking the numbers in the Q for RPO and current RPO, it looks like your current RPO increased by about $100 million or so, maybe more sequentially, implying roughly high single digit year-over-year versus mid-single coming out of Q1. So does that sound reasonable? And do you think that high single-digit current RPO growth is sustainable or perhaps something you can even improve further upon? Then a question for Sassine.

Sure. So I think as you – as we talked about even in my prepared remarks, we built significant backlog. We’re up $400 million quarter-over-quarter. So as Sassine was outlining 14 what we’re seeing in the industry, that’s really why we’re seeing the strength in the business and continuing to build the backlog, part of which is in the shorter-term duration that you’re talking about. So we do see a strong business environment that we’re operating in.

Okay. So Sassine, referring back to a principal team of yours at the forum 2 months ago, you made some very interesting comments regarding the need for engineering workflows to change, and you alluded to six areas specifically. And it seems to me that for most, if not all, of those areas, those would be so irrespective of AI. But perhaps you could talk about how you’re thinking about the pace or impact of those 6 areas that you spoke of in terms of having to change and that the implications would therefore be for you? And maybe I’ll make the question even more cumbersome. You’ve spoken of a tale of two markets. Perhaps there’s a tale of two AI to date has been largely about optimization, which is classical EDA productivity. But I’m wondering if with Agentic AI, customers’ considerations for managing that process will change significantly, and it’s a whole new way of managing the design process and perhaps at the risk as much as an opportunity?

Yes, Jay, excellent. Thank you for the question. We really covered two concepts at SNUG. One is the reengineering engineering and the second one is the move to what we call agent engineer solutions or technology. For reengineering engineering, it applies at multiple level in both the two tails of the semiconductor market or system companies. And there, if you take a multi-die advanced package system, there, the challenge is not only electronics, as you know very well, the challenge becomes how do you take into account the thermal structure, fluid to cooler off, et cetera, during the design phase. And this is a new approach that our customers need 15 to adopt a new methodology to ensure that they are designing the system right the first time, not running into issues after the system is in the field, et cetera. You’re designing a car, a drone, a robot, how do you bring the multiple aspects of physics with electronics to engineer that end product. There, the complexity, the pace, the cost, it’s really the opportunity that we see to deliver that digital twin of the future of these products. That’s the reengineering engineering. The next aspect is around how to simplify that complexity to our customers. And you’re absolutely right. The first wave of AI was focused on optimization. And this is not only unique to Synopsys. If you look at any company that is delivering an AI solution or a different applications, the first wave of AI is how do I optimize and provide a modern, simplified way to interact with the product. And in the EDA place, given it’s all about optimization, we’ve made great progress with fantastic customer adoption of the DSO, the VSO, et cetera dot-AI products because it always deliver better results faster. Where we are right now with agent engineer technology, we’re seeing some great opportunities for specific tasks say, RTL, hormone verification, et cetera, that the human engineer will define a target, a goal and you can pass it to an agent to deliver on an outcome. Now once these agents, the next phase, we see these agents communicated, learning from one another, having an orchestration layer, planning layer, decision-making layer. And that’s where we see the opportunity in terms of changing the workflow, which will give us an opportunity to change the monetization from the traditional EDA to a new opportunity as we deliver that value to our customers. Long answer, but it was a big question.

We’ll take the next question today from Nay Soe Naing from Berenberg.

16 I’ve got two, if I may. If I could start with the China performance, please. I think you mentioned that you’re now expecting revenue from China this year to be down year-onyear. If I remember correctly, I think you started the year with growth in China to be at corporate average, then you take that down to below group average. So I was wondering if there has been any changes in the business environment in China. Anything that’s changed in the last 3 months, that would be great. And my second question is around the recurring revenue in Q2 coming a little bit soft. I think it was similar dynamics in Q1 as well. I was wondering if there’s any color – additional color that you could provide around recurring revenue development? And maybe if you could also comment a little bit on your pricing power with your customer base today, please?

Yes. Thank you for the question. You are correct. We – as I said, in FY ’24, at the beginning of ’24, we started communicating the deceleration we’re seeing in China due to two factors. One is the macro inside China; and two, the cumulative impact of restrictions in China. So halfway through this year, in FY ’25, we could see clearer that these headwinds will not only bring China below the corporate average, what we’re factoring into our guidance is a decline in China year-over-year, and despite that decline, we are reiterating the full year. So there is nothing per se different happened in China, but that cumulative effect, we thought it’s prudent for us as we look at the full year guidance to say, can we reiterate that guidance with the decline happening in China? The answer is yes. Now weaving into the next part of your question, that’s due to the strength that we are seeing and have seen in other regions as well in the complete part of the portfolio. Recall, 17 we announced our hardware systems. and very well received by our customers, strong momentum. And we have always planned the year in a 45-55 fashion, 45% the first half, 55% in the second half. And the reason the shape is that way is with the anticipation of a bigger second half as we continue on rolling out our hardware system and of course, having visibility over renewals and our IP demands and EDA software based on our knowledge and insights with customers’ road map. And when you – the point you made regarding Q1, Q2, we actually were at or above the midpoint of guidance. So there were no surprises. We delivered to exactly what we said we’re going to deliver, and that’s the expectation for the second half. Lastly, on pricing, we always approach pricing based on the value and impact we deliver to customers as the challenges are higher for our customers to design these chips, it’s absolutely an opportunity to improve on our run rate either at the time of the renewal or when the customer is looking for new technology in order to address these challenges.

Yes. And just one add that I would have to Sassine’s answer is if your comment was about the percent of recurring revenue in Q2, just to help kind of put some color on that, that’s really normal based on timing. And obviously, our IP business is becoming a bigger and bigger part of the business. And so that tends to have an upfront. So as you saw, that was up 21%. So I would say it’s just a normal part of the business that it’s going to fluctuate quarter-on-quarter and particularly when you see a strong IP quarter.

We’ll go next to Joe Vruwink from Baird.

Maybe on the topic of China, if the Commerce Department was nearing a point of restrict18 ing a broader swath of your sales into China, do you think you would have heard about it by now or received evidence before today? Because in the past, I think the EDA vendors have actually had pretty good foresight on commerce deliberations and what we’re now learning today seems a bit different, I suppose.

Joe, based on history, we had different flavors of either having insights, having a heads up or not. The reason I wanted to address that point upfront because I saw the reporting and the swirling of the speculations. And all I can say at this stage is we have not received anything from BIS. And I cannot speculate why, why not, et cetera, but there’s nothing at this stage that Synopsys has received.

Fair enough. The early signs of pickup in auto and industrial, and I think last quarter, you were already seeing pickup in like PC and mobile. Maybe I’ll frame this question as outside of China. But when you think across all the different subcategories of chip R&D, it does seem like you’re seeing outright improvement or maybe the potential for improvement. Do you think the overall R&D landscape is maybe getting close to what you saw in the 2021, 2022 time frame? Because obviously, that was a very strong period for your bookings and revenue growth. And it maybe seems directionally, we’re heading back to something similar.

Yes. In the non-AI part of the semiconductor, what is – what we’re clearly observing in the last quarter, quarter plus is a pickup in automotive, industrial. And by the way, when I talk about industrial, we put robotics type of application, et cetera, in that category. And those early signals, we see them in IP. With those same set of customers, what they are contemplating when it comes to their R&D investment and their road map is how much 19 AI do they need to have for these chips that they are building and the applications they’re building. So it’s not necessarily the same chip or derivative of the chip that they were applying for the same market is more sophisticated type of chip they’re building. Therefore, it’s expected that their R&D investment will either get modified to make sure they have the right skills to deliver on these chips or increase overall, which is, again, if you’re in these customers’ shoes, that’s the opportunity that they need to strive towards in order to compete and have an opportunity to grow in the new opportunity of AI being everywhere.

Your next question today comes from Gary Mobley from Loop Capital.

Sassine, I think you would agree with this that your customers do see some benefit as you shift more of your EDA tools to GPU-based compute and away from retrieval-based compute. And I guess the benefit there would be coming up with a solution on the fly for that accelerated compute. So my question to you is, what is the pricing benefit? And do you feel like you’re pricing that product correctly? And is there a cost consequence as well as presumably that type of compute is hosted in the cloud environment?

Yes, that’s an excellent question, Gary. The one product actually that has – was introduced first and has been talked about regarding the GPU acceleration is Proteus, which is an OPC product is in manufacturing, where it’s a very, very heavy compute step in the process development, and we’re talking about thousands of CPUs for every run in order to deal with the complexity of computation. 20 There with GPU, and again, this is what we talked about at Snug, et cetera, 15x plus speed up using a GPU. So it’s not a percentage speed up, it’s multiple factors of speed up. And we absolutely price it based on the total cost of ownership that our customer is seeing by moving from a CPU to a GPU in that case. And we’re actually very happy with the pricing delta that we’re able to get due to GPU acceleration. Very similarly, as we look at the rest of the portfolio, where are the compute-heavy applications, verification, sign-off when it comes to timing, et cetera, those are massive machines, big compute requirement, and we have the pricing associated with that value based on the customer perceived TCO and overall benefit that they’re seeing. And I believe the other part of your question, are we seeing – or do we believe there’s a downside for an acceleration? I don’t believe so because the complexity is so high. And when you have the right pricing model, it’s an upside.

Okay. Just a quick follow-up. As we look into the second half of the year with a high percentage of revenue coming from ZeBu and HAPS 200, do you have any supply chain constraint considerations, whether it be from your chip partners or from your EMS partners?

Yes. Gary, this is Shelagh. No, we don’t. The truth is that we would like to have more right now, but we’re able to fulfill customer demand. And as you kind of see in the shape of the year, we have a better supply situation in Q4 than we do at any other time of the year. So that’s a bit of the heavy Q4 weighting.

The next question today is from Joshua Tilton from Wolfe Research. 21

I have two. The first one is more of a clarification. And I guess what I’m trying to understand is – somebody else already kind of walked you guys through the time line of what you’re expecting for China going into the year and how you kind of walked that back a little for the last 2 quarters. And I guess what I’m trying to reconcile is you still hit numbers this quarter. So was the China revenue in 2Q in line with expectations? Or was that below what you guys thought you were going to do, and that’s why you now think that China is going to be worse than you thought it would be 90 days ago? That’s my first question.

Yes, Josh, as you zoom out and you think about how do we forecast and therefore, guide, there are the puts and takes that we have in place. And what we have hit in Q1 and Q2 for China was in line with what we forecasted internally. As we look at the rest of the year and some of the, let me call them, tailwinds that we have in the first half, and we look at the entire year, we felt confident to reiterate the overall guidance with an assumption that even if China is a decline year-over-year, we feel confident that we will hit the overall company’s growth that we have guided initially.

I guess maybe just – sorry to clarify that over time. So is it fair to assume the updated guidance, which now includes China to decline year-over-year, it would imply that you need to see China get worse from where it was for the first half of the year?

No. What we’re saying – no, there is the first half assumption in China that was lower than the second half assumption in China. So the first half, if I’m not mistaken, it’s a minus 28% 22 in China. We’re not assuming that the second half is going to continue declining at that level because when we forecasted the entire China growth, we knew that the first half is going to be tougher than the second half. What we wanted to communicate clearly is our reiteration of the guidance is affirming that guidance with an assumption that if China is declining year-over-year below the FY ’24 finish, we still feel confident about hitting the overall company’s revenue target.

Okay. And then maybe just my quick follow-up to that is, you guys put out the financial supplement every quarter in there. There’s a nice little line that says your long-term multiyear objectives are industry-leading double-digit revenue growth. And I understand that things are happening on the fly and you guys haven’t received the letter yet. But as investors kind of contemplating all the potential scenarios in the future, are you guys just as confident in being able to deliver that industry-leading double-digit revenue growth even if for some reason, sales to China have to be halted?

Put the sales to China being halted because that’s something we – I’ll be speculating. I have no idea what will be halted, what will not be. So put that aside, we’re absolutely confident with an industry-leading double-digit growth at company level for design automation, double digit; for IT, mid-teens. And that’s what we’ve been delivering to. And that’s organic growth. It’s important to emphasize. That’s an organic execution we have been able to deliver to these numbers.

And our final question today comes from Blair Abernethy from Rosenblatt. 23

Two quick ones. Just, Sassine, on your AI portfolio. So your DSO AI and other optimization products have done very well in the last 4 years. You haven’t spoken a lot about Design.da and Silicon.da and Fab.da. Could you speak to the adoption of those vis-a-vis the customers that have taken up the optimization products? And then secondly, do you have any sense of time line that you’d like to share with us in terms of rolling out Agentic EDA?

Yes, Blair, thank you for the question. Absolutely correct. Our dot.ai products with our flagship early introduction of DSO.ai followed by VSO, ASO, et cetera, great traction by the customers. And I want to say it’s becoming an easier go-to-market and engagement with customers because the customers are pulling us into these engagements and finding an opportunity for every one of those dot-ai technologies. As far as dot-da, you’re right, we may need to do a better job communicating what we are seeing in terms of customer adoption. But think of the .a as an infrastructure step to tie up. It’s like a continuum of data where you provide insights from step two back into step one, moving to step four, et cetera, in the entire design flow. Fab.da, for example, is limited to fewer customers because that’s limited to companies that they typically have their manufacturing capability in-house. But think of it as an infrastructure step that makes our optimization product even stronger and better because it provides insights. As far as the Agentic road map, today, we have a number of engagements with a limited set of partners to provide feedback on the application. And when you think agents, you have to be very specific to a customer workflow because each customer, they have dif24 ferent ways in bringing up that agent engineer based on their environment, their data, et cetera, et cetera. At SNUG, we communicated L1 through L5. Same analogy for automotive. And what we said today, we are in the L2, L3 phase. It doesn’t mean when you move to L4 that you stop in L2. We continue building those task agents in the L2 phase as we move to the L3, which is the orchestration between the agents. We’ll make sure to continue on communicating our progress there because it’s super exciting.

Thanks so much for everybody who joined our call. And Lisa, thank you for shepherding us through the call. And with that, we look forward to talking to you through the quarter.

Thank you all.

Thank you.

And once again, ladies and gentlemen, that does conclude today’s conference. We would like to thank you all for your participation today. You may now disconnect. Copyright © 2025, S&P Global Market Intelligence. All rights reserved 25