The Zhitong Finance App learned that Synopsys (SNPS.US) announced the results for the third quarter of fiscal year 2026 on August 26. Revenue, profit margins, and earnings per share all exceeded the upper limit of the guideline. Management emphasized during the conference call that EDA is accelerating towards double-digit growth, design IP has returned to the growth path, and the synergy effects have begun to be realized after one year of integration with Ansys. CEO Ghazi pointed out that the complexity of AI-driven chip design is not only a threat, but has led to an exponential increase in demand for EDA software and IP, and the joint launch of Multiphysics Fusion has been verified by leading customers. The company simultaneously raised its annual performance guidelines and announced that the “Factory 2” customized IP strategy is progressing smoothly. More details will be disclosed on Investor Day in September. Management is confident of continued growth in the 2027 fiscal year.
Synopsys delivered strong responses in the third quarter of fiscal year 2026. Revenue, non-GAAP operating margins, and earnings per share all exceeded the guideline limits. Management sent a clear signal during the conference call: EDA growth is accelerating towards double digits, design IP has returned to the growth path, and the synergy effects are gradually being realized after a full year of integration with Ansys. CEO Sassine Ghazi attributed this season's performance to “the overall strengthening of product portfolio fundamentals” — EDA's hardware-assisted verification revenue reached a new high thanks to intensive investment by AI and high-performance computing customers in the 3DIC advanced packaging field; the IP business continued to pick up sequentially, driven by AI infrastructure demand. Among them, the die-to-die business is expected to double year-on-year, and the PCIe 7 chance bid rate is over 95%. What is particularly noteworthy is that Multiphysics Fusion, jointly launched by Synopsys and Ansys, has been verified by leading customers such as Nvidia, MediaTek, and Samsung, and can accelerate design convergence by up to 10 times. Management expects the product to contribute substantially to EDA growth starting in FY2027.
Management focused on the effects of AI in reshaping the chip design paradigm. Ghazi made it clear that AI is not only a disruptive threat to the EDA industry; instead, it has spawned greater software demand — customers are increasingly relying on “ground reality” level physical verification when building autonomous design workflows. This season, Synopsys showcased autonomous EDA agents in collaboration with Nvidia, Microsoft and AMD, which can shorten the verification cycle by 50 times and reduce debugging time by 40%. Ghazi revealed that there are currently more than 30 active agent-based artificial intelligence customer cooperation projects, and as agents undertake more engineering tasks, the call consumption of underlying EDA tools will increase exponentially, forming a new consumption-driven growth model.
Facing the future, management proposed an IP growth strategy in parallel with “Factory 1” and “Factory 2.” “Factory 1”, the standard “build once, sell multiple times” IP model, still benefits from the steady demand for chip startup in the automotive, mobile and other fields. The automotive ADAS platform has maintained a bid rate of over 90% for three consecutive quarters. “Factory 2” targets customized chip opportunities and extends upstream of the value chain through the “license+gold” model. In-depth negotiations have now been carried out with a number of hyperscale cloud vendors and ASIC suppliers. More details will be revealed on Investor Day on September 30. CFO Shelagh Glaser simultaneously raised the full-year guidance, with a mid-year revenue increase of 50 million US dollars, non-GAAP earnings per share in the range of $0.31 to 15.04-15.10 US dollars, and a sharp increase of 600 million US dollars to about 2.6 billion US dollars in free cash flow guidance, fully demonstrating management's confidence in the end of the year and continued growth in the 2027 fiscal year.
The following is a transcript of Synopsys' Q3 FY2026 results conference call:
Executive speeches
Tushar Jain
Investor Relations Supervisor
Good afternoon everyone. Welcome to Synopsys's Q3 FY2026 results conference call. Today we were joined by Synopsys President and CEO Sassine Ghazi and Chief Financial Officer Shelagh Glaser.
Before we begin, I'd like to remind you that during this conference call Synopsys will discuss forecasts, goals, and other forward-looking statements about the company and its financial results. Although these statements represent our best judgment on future results and performance to date, our actual performance is affected by many risks and uncertainties, which may cause actual results to differ materially from our expectations. In addition to any of the risks we highlighted during this conference call, important factors that may affect our future results are described in our recent report to the SEC and the results press release issued today.
Additionally, we will be referring to certain non-GAAP financial measures during the discussion. Reconciliation to the most directly comparable GAAP financial measures and supplementary financial information can be found in our earnings press release, financial supplement, and 8-K file issued earlier today. All of these documents, as well as the latest investor presentations, press statements, and Investor Day information, are available on our website at www.synopsys.com.
Next, I'll hand over the phone to Sassine Ghazi.
Sassine Ghazi
Chief Executive Officer, President and Director
Good afternoon everyone. Synopsys achieved outstanding results in the third quarter, with revenue, non-GAAP operating margins, and earnings per share exceeding the upper limit of the guideline range. These results reflect a wide range of growth drivers, including the excellent performance of EDA and ANSYS, as well as continued growth in the IP business. We are increasing our full-year revenue, non-GAAP operating margin, and earnings per share guidance. Additionally, we expect EDA growth to accelerate in the fourth quarter and achieve double-digit growth throughout the year.
The key conclusion of the third quarter was that the fundamentals of our product portfolio are strengthening. EDA is growing at an accelerated pace. Design IP has resumed growth, and Ansys has performed strongly while starting to create new growth opportunities in the combined product portfolio. This quarter marks the first anniversary of the completion of the Ansys acquisition. In the third quarter, we launched Multiphysics Fusion, our first Synopsys joint solution with Ansys.
I want to thank our global team for integrating our world-class capabilities with focus and agile execution. This merger strengthens our competitive position, expands our market opportunities, and enables us to provide differentiated solutions to address the physical challenges of modern chip design. Industry trends are highly consistent with our strategies and strengths, and we are a leading provider of engineered solutions from chip to system. Unprecedented design complexity, driven by AI requirements, is driving the need for IP and design solutions to deliver next-generation AI computing, infrastructure, and physical AI systems. These trends are reflected in our third-quarter results.
Starting with Design Automation (Design Automation), the business experienced healthy growth in the third quarter, thanks to EDA's strong performance, including record hardware revenue. We expect this trend to continue, with EDA growth accelerating to double digits in the fourth quarter and achieving double-digit growth throughout the year.
Design activity is most active among AI and high-performance computing customers, who are developing increasingly specialized chips using multi-core architectures and more complex packaging and system requirements. These are all areas where Synopsys is leading. AMD's recent Instinct MI455X GPU is a good example. To deliver this highly complex new product line, AMD used Synopsys' 3DIC Compiler, the industry's only multi-core and advanced packaging co-design and optimization platform from exploration to signature.
AI computing's requirements for complex software and systems are also driving the demand for our hardware-assisted verification solutions. We received orders from 12 new customers and 66 repeat customers for hardware-assisted verification during the quarter.
As I mentioned, the release of Multiphysics Fusion was an important milestone in the EDA sector in the third quarter, creating new growth opportunities for Synopsys. Multiphysics Fusion combines Synopsys and Ansys technology into the industry's only solution that fully integrates thermal analysis into the chip design process. Customers including Nvidia, Cisco, MediaTek, and Samsung Foundry have verified up to 10x faster design convergence and 3x faster uptime. This brings greater value to both our customers and our products. We expect these additional features to contribute to EDA growth starting in 2027.
Agentic AI is another growth vector for Synopsys, and we showed strong progress in the third quarter. At the DAC conference, we and Nvidia jointly demonstrated a fully autonomous long-term operation design verification agent that can orchestrate the entire chip verification cycle, accelerate RTL verification time by up to 50 times, and achieve an additional 20% increase in coverage. We partnered with Microsoft and AMD to launch the first autonomous EDA workflow on Microsoft Discovery, which can automate debugging, implementation, and design convergence. Early collaboration showed that commissioning cycle times were reduced by up to 40%, saving weeks of engineering work while improving design quality.
We have seen strong customer interest in our agent-based artificial intelligence platform, and there are currently more than 30 active customer collaborations underway. Early feedback clearly indicates that as these agents take on more engineering work, they will orchestrate our underlying EDA tools at a significantly higher rate. This allows customers to run more design and verification workloads, and as we capture a fair share of the value these agent workflows and foundational tools bring to customers, it creates opportunities for incremental growth for Synopsys.
Turn to Ansys. A year after the merger, Ansys continues to see strong demand. From semiconductors to aerospace to industry, companies are embracing digital engineering. Across industries, Ansys simulation is accelerating innovation while reducing development risks and costs. For example, a leading car manufacturer is using Ansys SiMai to achieve approximately 98% prediction accuracy and advance collision analysis to near real-time.
A leading heavy equipment manufacturer has accelerated motor design by more than 10 times. We are applying AI to extend our leadership in simulation and further automate the simulation of complex systems. This includes expanding our GPU-accelerated Ansys application portfolio. In the third quarter, our biggest ANSYS deal was GPU-accelerated ANSYS CFD to support enterprise-grade digital twins for a multinational electronics manufacturer.
Shifting to the design IP business, the business achieved month-on-month and year-over-year growth driven by extensive AI infrastructure requirements. As AI drives the demand for higher bandwidth, faster connectivity, and more complex system architectures, our portfolio of interfaces, memory, and die-to-die IP products is at the heart of the entire technology stack. Our third-quarter results also reflect this.
We won over 95% of PCIe 7 opportunities, including a subsystem order from a well-known enterprise storage customer. In terms of LPDDR6, our products have been silicon verified at multiple nodes and foundries, and we have received 25 design orders so far this year. Our die-to-die business is expected to double year-on-year, and has now accumulated more than 100 design orders. The industry continues to rely on Synopsys to deliver silicon-proven quality and unrivaled scale.
Our standards-based “build once, sell many times” IP model remains the cornerstone of our growth strategy. We will continue to invest and grow this business, which I call “Factory 1” (Factory 1), which benefits from strong chip launch activity and solid traction from various industries. For example, in the automotive sector, we have maintained a design bid rate of over 90% for three consecutive quarters as ADAS platforms have been upgraded to 5nm and 3nm processes. In the fields of mobile, consumer, and edge AI, our cumulative lifetime orders for USB IP have exceeded 2 billion US dollars, and design orders from front-line customers have migrated to leading process nodes.
As AI expands from digital infrastructure to physical products, demand for silicon will continue to expand, providing a smooth wind for our standards-based IP business.
Opportunities for higher growth in the IP business lie in the growing number of AI customers requiring deeper collaboration and IP solutions optimized for their specific workloads and architectures. To meet this demand, we are expanding to differentiated IP subsystems and enabling customized chip solutions. From hyperscale cloud vendors, ASIC vendors, and foundries to traditional semiconductor companies, customers of all types want to partner with Synopsys to accelerate their chip development efforts and use our IP and engineering expertise to build increasingly differentiated custom chips.
This is our “Factory 2” (Factory 2) model for customized IP. It extends us from a simple licensing business to a licensing plus royal gold model, enabling us to seize the rapidly growing opportunities in the custom chip market. This is an important strategic priority, and we are making strong progress. We are actively negotiating with several “Factory 2” customers, and I look forward to sharing more information on Investor Day.
All in all, I want to thank the entire Synopsys team for their continued focus, innovation, and execution. The third quarter solidified our strategic strengths and our confidence in a strong end to the fiscal year. AI is driving the demand for advanced chips, system-level engineering, and AI-driven design. Our leading product portfolio allows us to capture a larger share of R&D investments across industries. We remain focused on translating technology leadership into sustainable growth and margin expansion.
Next, leave it to Shelagh.
Shelagh Glaser
chief financial officer
Thanks Sassine. We achieved excellent results in the third quarter, achieving revenue of $2,477 billion, non-GAAP operating margin of 41.6%, and non-GAAP earnings of $3.91 per share, all exceeding the upper limits of our guideline range. Thanks to the broad growth drivers of various businesses, the performance of revenue exceeding expectations was driven by the strong performance of EDA and ANSYS businesses.
The backlog remains very strong, reaching $10.9 billion, and declined slightly from month to month due to the completed asset divestment of the processor IP solutions business in the third quarter. With strong performance in the third quarter, steady cash flow generation, and continued momentum into the fourth quarter, we are increasing full-year revenue, non-GAAP operating margin, earnings per share, and cash flow guidance.
I'm now reviewing the third quarter results. All comparisons are YoY unless otherwise stated. We achieved total revenue of US$2,477 million, an increase of approximately 42% year over year, including Ansys revenue of approximately US$711 million.
As stated by Sassine, Ansys continues to perform strongly one year after the merger. We are also ahead of schedule in terms of cost synergy commitments and have repaid our term loans ahead of schedule.
GAAP total costs and fees were $2.119 billion, and GAAP earnings per share were $2.84. Third-quarter GAAP earnings per share include earnings related to the sale of the processor IP solutions business completed during the quarter. Total non-GAAP costs and expenses of $1,446 million are at the lower end of our guidance range as we continue to improve operational efficiency and achieve Ansys cost synergies ahead of time to achieve a non-GAAP operating margin of 41.6%. Non-GAAP earnings per share of $3.91 exceeded our guidelines and highlighted our strong operating execution this quarter.
Next, let's look at each business segment. The Design Automation segment generated approximately $2 billion in revenue. As a reminder, this data excludes Optical Solutions Group (Optical Solutions Group), which was divested in the fourth quarter of fiscal year 2025. In the design automation sector, EDA revenue increased 8.5% year over year in the third quarter, reflecting strong EDA software performance and another record quarter for hardware-assisted verification solutions. The operating margin after adjustment for design automation was 45.2%.
The Design IP (Design IP) sector resumed growth, with revenue of US$474 million, an increase of about 11% over the previous year. In line with our expectations, this represents continued month-on-month growth in the IP sector as we repositioned our product portfolio to focus on the highest-value market opportunities. The adjusted operating margin for design IP was 26.5%.
About cash flow. Free cash flow for the third quarter was $746 million, and cash and short-term investments at the end of the quarter were $3.6 billion. Total debt at the end of the third quarter was approximately $10 billion.
About the year-round guide. We raised the midpoint of our total revenue guidance by $50 million, mainly due to the strong performance of the EDA-led design automation sector. As stated by Sassine, EDA remains strong, and we expect double-digit growth in organic EDA revenue for the fourth quarter and full year of FY2026. We continue to expect the IP business to grow sequentially in the fourth quarter. As a result, the annual revenue range was $9.69 billion to $9.74 billion. Among them, Ansys' annual revenue contribution is estimated to be approximately US$2.98 billion, an increase of US$20 million from previous guidance.
About fees. The total GAAP costs and expenses are estimated to be between $8.667 billion and $8.742 billion. This includes an increase in anticipated FY 2026 expenses associated with our previously announced restructuring plan as we accelerate the delivery of promised synergies.
Total non-GAAP costs and expenses are expected to be between $5.67 billion and $5.70 billion, with a midpoint non-GAAP operating margin of 41.5%, up 50 basis points from previous guidance. GAAP earnings per share are expected to be between $3.84 and $4.08. Non-GAAP earnings per share are expected to be between $15.04 and $15.10, up $0.31 at the midpoint, due to higher revenue and improved operational efficiency.
We raised our operating cash flow guideline by US$500 million to approximately US$2.8 billion, and lowered our capital expenditure guidance to approximately US$225 million. As a result, free cash flow was approximately US$2.6 billion, an increase of US$600 million over the previous guidance.
Regarding the fourth quarter target: Total revenue between $2.53 billion and $2.58 billion, total GAAP costs and expenses between $2,225 million and $2.3 billion, total non-GAAP costs and expenses between $1.45 billion and $1.48 billion, GAAP earnings per share between $0.60 and $0.85, and non-GAAP earnings per share between $4.10 and $4.16. Our press releases and financial supplements include more goals and GAAP and non-GAAP reconciliation information.
Thank you to our global Synopsys team for another strong quarter. These results reflect strong execution across our business lines, continued demand for our technology, and strict operational management, and we are laying the foundation for the next phase of growth. We look forward to meeting you on Investor Day in September to discuss our long-term development opportunities as a mission-critical partner for our clients.
Next, there will be a question and answer session.
Q & A session
Jason Celino
KeyBanc Capital Markets Research Division
The results were excellent. I think the most prominent one is the 8% EDA growth, which is the same as the previous quarter despite having a high base. You mentioned that the fourth quarter and full year are expected to accelerate to double digits. What are the factors driving this acceleration? Is it a design launch event? Is it agent-based artificial intelligence? Or is it a better monetization strategy? Please help us understand.
Sassine Ghazi
Chief Executive Officer, President and Director
Thank you Jason for the question. Yes, we are very excited about EDA's overall performance and ending the year with double digit growth, which is exactly what we promised to grow in the business segment. There are many factors driving confidence in the EDA business: the complexity of chip design, the migration to advanced packaging and 3DIC — the example of AMD using the 3DIC Compiler mentioned in my remarks, and many other customers who are designing advanced packages are using our technology.
AI is definitely a tailwind factor. As customers rethink how to redesign their chip design engineering, this requires a different engineering approach, which brings us another tailwind factor. On the hardware side, our hardware business achieved record annual revenue. Overall, we are very excited about EDA's organic growth rate of over 8%.
Shelagh Glaser
chief financial officer
Yes, Jason, I'd like to point out that the 8.5% EDA growth you saw in the third quarter was achieved under a very high comparison base. The third quarter of fiscal year 2025 was 16%. So, as Sassine said, this just shows the strength of the business—being able to achieve this kind of growth under a high base and maintain double-digit growth throughout the year.
Jason Celino
KeyBanc Capital Markets Research Division
OK. Also, Sassine, you mentioned that customers need to redesign their engineering processes. We're seeing a lot of innovation happening. People are always talking about new AI models. Yesterday, as an example, OpenAI talked about the development of their new chip and mentioned using its own AI model to speed up the design process.
Can you talk about what it takes to redesign the engineering process when you hear examples like this — because many of your customers have also used their own models in the past? Is this a complement or addition to existing processes? Or is it an alternative? I don't know if this statement is accurate, but...
Sassine Ghazi
Chief Executive Officer, President and Director
Of course, of course. Thanks for the question. For at least the past year and a half to two years, we've been discussing how AI is reshaping engineering. Synopsys has been investing in and leading the way in providing “intelligent engineers” to our customers to rethink workflows, including AI models — these models will definitely participate in and drive engineering redesign. In every case, the underlying requirements are increasing use of our software, because whether the customer is using an AI model or an agent, you still need a physical level of “ground truth” (ground truth) to enable the model to operate with confidence and deliver the best results.
So, the example you mentioned is actually a very good example, and a huge opportunity for Synopsys in the EDA field. Not to mention, this is also a huge opportunity for IP, as it is a rethinking of the entire architecture when customizing the chip.
Joseph Vruwink
Robert W. Baird & Co. Research Division
I'd like to look back at the situation a few years ago. On Investor Day 2024, you shared the forecast that 30% of EDA software demand by next year could come from multi-core projects. I'm curious how that number is currently progressing and what your thoughts are after 2027. I think we've all noticed a lot of attention recently and even this week about HBM. When you think about migrating the DRAM process to the logical process, how will the content of EDA change? Is this actually stronger than you expected a few years ago?
Sassine Ghazi
Chief Executive Officer, President and Director
Yes, Joe, that prediction has actually accelerated. As I mentioned in my prepared statement, our die-to-die orders have doubled in the past 12 months. The reason is the rise of advanced packaging and 3DIC architectures as a whole, which is driving significant opportunities in the IT field. Die-to-die is just one example; of course, there are all the other interfaces needed to splice the system together as an end product.
In terms of EDA, the 3DIC Compiler is the key, and this is also the importance of a joint solution with Ansys. You can't build these systems without taking physical, thermal, structural, and fluid factors into account during the chip design phase. At the time, we thought the 30% figure was a challenging target, but given all of our customers' investments in building these efficient custom chips, that number has definitely accelerated.
Joseph Vruwink
Robert W. Baird & Co. Research Division
OK, great. I would also like to ask if EDA's double-digit organic growth is split into two parts: software and hardware, will the performance of the software business reach the state you want to see at the end of the year? Hardware seems to have always been a very strong driver. I'm more concerned about the performance of the software business.
Sassine Ghazi
Chief Executive Officer, President and Director
We are very pleased with the contribution of the software part, very, very satisfied. I'm actually most excited about the delivery of our joint solution with Ansys. This provides us with a platform to address the direction of future engineering challenges. As we just mentioned, the need for advanced packaging and the introduction of physics into electronic design makes us more confident to continue on the path of double-digit growth.
Yu Shi
Needham & Company, LLC Research Division
First, Shelagh — or maybe this question is for Sassine. I have a high level long term problem. Sassine, we know that going back about 10 years, you played a critical role in introducing AI into Synopsys tool processes. I know it probably wasn't a big language model back then; it was more like a reinforcement learning type of AI, such as excellent products such as DSO.ai.
However, AI has developed rapidly in the past 10 years, especially in the past 3 years. So the question I often hear from investors is, is there a risk that AI will disrupt commercial EDA businesses? One way of thinking is there an end-to-end so-called “AI native chip design” that can bypass all commercial EDA tools, especially some AI models that may be trained using synthetic data generated by commercial EDA tools? In your opinion is this a real threat to the entire EDA industry? Or in what ways do you think AI might replace or complement commercial EDA businesses?
Sassine Ghazi
Chief Executive Officer, President and Director
Yes, thanks Charles. I don't need to go back as far as 10 years ago. You're right, around 2017, we introduced DSO.ai and invested in reinforcement learning, with great results. If you look at the past 3 years, the focus is on copilots (copilots), generative AI, and agents (agents). Now we're talking about independent design. When you start looking at autonomous workflows, accuracy and certainty are paramount. Customers won't invest hundreds of millions of dollars in a product without confidence that it will work.
We have a product portfolio based on sign-off leadership, which is critical to building these autonomous workflows. We're participating with our customers on how to enable autonomous workflows, and this isn't happening without our participation. We're actively working with our customers to redesign how they see the future of AI-centered engineering.
I'm not at all worried that at some point this model will be able to complete an end-to-end design without our participation. Because you have to keep in mind that these models aren't static; they're constantly changing and need to be learned. So the opportunity is just the opposite — this isn't a threat, but a huge demand for our software — for training, reasoning, and continuously empowering faster design to deal with complexity, and we're at the center of this.
Yu Shi
Needham & Company, LLC Research Division
Perhaps, Sassine, with regard to opportunities for agent-based artificial intelligence, how should we consider improving overall revenue growth, especially EDA growth? Can you describe the progress of shifting the business model to a subscription plus consumption model (at least for smart entities)? How are discussions with customers going?
Sassine Ghazi
Chief Executive Officer, President and Director
Yes, we already have several collaborative projects. Regarding your first question, when our customers are exploring whether to use Synopsys agents or use Synopsys agents plus their own intelligent—to keep their own “secret recipe” in the workflow—the core requirement for customer workflows or customer agents is that more licenses are required. We're defining a number of ways to work with our customers, including subscribing to our agents, subscribing to our workflows, and consumption measurement models as these agents and new workflows consume more software.
In a few weeks, on September 30, we will further elaborate on how we think about the long-term growth model in this context. But to be sure, we're having in-depth discussions with many customers about the different ways — using their smart devices, using our agents, a mix of the two, and the multi-model options they're considering.
Lee Simpson
Morgan Stanley Research Division
OK, maybe a few quick questions about the IP business. Maybe before Analyst Day I was wondering if Sassine could you outline the speed at which the “Factory 2” opportunity (Authorized Gold Model) you described is changing?
And then the second question, I think it's been about a year since we saw Intrinsic ID's acquisition. I remember when you mentioned secure IP as a new growth vector for the IP business. I'd like to know if you could outline the scale of this opportunity, how you see the collaboration progressing, and where you think the deployment will take place in particular.
Sassine Ghazi
Chief Executive Officer, President and Director
Thanks Lee. Regarding “Factory 2,” you actually only need to look at the trend where every hyperscale cloud vendor is investing and at various stages of delivering their own custom chips. These chips would not be possible without our interface IP. These customers need our interface IP to build their own chip and connect to the ecosystem, whether it's a memory vendor or — if they're investing in their own XPU but need a network chip from the vendor, connecting them all together is inseparable from our interface IP. That's where the opportunity lies.
We started talking about “Factory 2” because we saw a need to tailor these standards and accelerate delivery of those customizations. Synopsys is uniquely positioned to deliver this customized and accelerated service due to its scale, knowledge, skills, and market position. We are having in-depth discussions with many of these customers to shift the business model from traditional IP licensing plus some NRE to licensing plus gold.
I'm looking forward to helping you understand what this looks like in a few weeks. But I'm very excited about the investment we have made and the flexible adjustments in this direction, and of course we will continue to lead and invest in “Factory 1,” as it is also an area that will continue to bring significant opportunities.
Regarding security, the reason we began reviewing our product portfolio and divesting our processor IP business was to focus on areas of growth. Safety is one of them. As you said, we haven't listed them separately, but we have a strong market position in the security sector, and this only becomes more important as the need to protect chips not only at the software level but also at the hardware level becomes more important — this is where our security product portfolio comes in.
Arsenije Matovic
Wolfe Research, LLC
This is Arsenije asking instead of Josh. First, Sassine, you mentioned seeing strong early customer interest after Multiphysics Fusion launched. Historically, chip design and simulation have often been handled by different engineering teams using different workflows and tools. What I'm asking is, when you bring thermal analysis directly into the design process, how are customers structurally coping with this integration? Are you seeing these teams begin to work more closely and integrate around more common workflows? Or does adoption still require spanning diverse engineering teams and possibly independent budgets?
Sassine Ghazi
Chief Executive Officer, President and Director
Yes, you are absolutely right. These engineering teams or engineering skills and expertise used to be separate fields. As the design steps go from synthesis to physical design to signing, etc., there are different ways to hand them over. The requirement for collaborative design is critical to reduce margins and deliver these competitive products. Therefore, absolutely, leading customers require that thermal considerations be taken into account in the design implementation process, when constructing these 3DICs, and when incorporating structure and stress into the chip architecture, not just in the implementation phase of integrating multiple cores.
So yeah, definitely. This is the direction we are investing in — enabling implementation engineers at the front end of the design process to introduce sign-off accuracy analysis early in the design process with minimal additional effort. This is exactly what we see as an opportunity to combine product portfolios and deliver a fusion of physics and electronics.
Regarding the budget — sorry, about the second part of your question — from a budget perspective, as we promised “1 plus 1 is greater than 2,” this means that even if the budget comes from EDA, the joint solution will capture incremental revenue compared to existing independent point tools.
Arsenije Matovic
Wolfe Research, LLC
Got it, that helps a lot. Then, Shelagh, I'd like to clarify a specific question about Ansys — an increase of $20 million to $2.98 billion. You helped us understand that accounting dynamic that contributed $12.5 million last quarter, or $60 million for the whole year. Is it still $60 million for the whole year? Or is Ansys' $20 million increase simply a substantial increase in good execution?
Shelagh Glaser
chief financial officer
Yes, we talked about accounting changes last quarter. We made that accounting change, and the guidelines we have given include this change. So we're seeing strong performance in the Ansys business, including the channel business.
Arsenije Matovic
Wolfe Research, LLC
Understood. So what were Ansys' accounting changes in the third quarter? Is it still $60 million for the whole year? Or is it just a different number?
Shelagh Glaser
chief financial officer
It remains within the same range throughout the year.
Arsenije Matovic
Wolfe Research, LLC
Understood. Was that quarter $12.5 million or...
Shelagh Glaser
chief financial officer
We have not separately disclosed quarterly figures.
Joseph Quatrochi
Wells Fargo Securities Research Division
You talked about the “two markets” of AI and non-AI in EDA. I'm wondering if the acceleration you're seeing is AI taking a bigger share of the whole and offsetting non-AI parts? Or are you seeing an acceleration in the non-AI part too?
Sassine Ghazi
Chief Executive Officer, President and Director
Yes, Joe, the reason we initially made this judgment—an internal measurement we do every quarter—was because we track chip starts (chip starts). The reason we were able to cover the design launch so well is thanks to our IP product portfolio. Any customer who is planning to launch a new chip will be reached very early on through our IP product portfolio. Of course, EDA will follow.
The observation is that in the non-AI sector, the past two quarters have stabilized. This means we previously observed a slowdown in design launches in non-AI fields, but over the past few quarters, it has stabilized and is no longer declining. On the other hand, in the field of AI, we have been and will continue to see an acceleration in design launches, which provides us with a good balance.
Joseph Quatrochi
Wells Fargo Securities Research Division
This helps a lot. Also, as a question, I was wondering if you could provide some analysis of the additions and subtractions in relation to pending performance obligations (RPO). There was a slight decline this quarter, affected by asset divestments? Did it meet expectations?
Shelagh Glaser
chief financial officer
Yes, as expected. As I stated in my statement to the press release, the minor changes were mainly due to the divestment of the processor IP business assets completed during this quarter. As you may recall, we were close to completing the transaction at the time of the last earnings report, which was completed within a few days of the earnings report being released. So this is what you saw this quarter.
S itikantha Panigrahi
Mizuho Securities US Research Division
Congratulations, Sassine, for a truly outstanding quarter. Back to Ansys and Synopsys' integrated products, you launched these products. I think you mentioned that Multiphysics Fusion isn't expected to contribute to EDA growth until 2027. So can you talk about adoption and pipeline trajectory from now to then? And is the 2027 contribution more likely to be reflected in incremental EDA growth? Or is it a share acquisition from your Ansys simulation customer base? Any relevant information on pricing advantages and value to customers would be very useful.
Sassine Ghazi
Chief Executive Officer, President and Director
Yes, from a customer value perspective, 10x design convergence acceleration or 3x SPICE accurate multiphysics timing acceleration is a significant value for customers. What does this mean for customers? Fewer iterations, better design, more speed. As a result, in early partnerships, our customers have validated these values and are in the early stages of deployment. Once these customers enter the production phase, we'll start to see upside in revenue.
As we said from the beginning, the FY2026 Joint Solution will not contribute much. This is a year of execution, a year of delivery of these products. Looking ahead to fiscal year 2027, it will definitely contribute to our EDA growth. We are also absolutely committed to reaching our $400 million synergy target in the fourth year. So when we meet in September, we'll be able to begin discussing the contribution of differentiated solutions in 2027 and beyond.
Sitikantha Panigrahi
Mizuho Securities US Research Division
This helps a lot. Another quick question. I know Mike Ellow has been with us for a few quarters. How does he drive the sales organization? What specific changes is he considering or is already driving to launch an integrated product as you launch an integrated product?
Sassine Ghazi
Chief Executive Officer, President and Director
Yes, Mike has done a great job using what the organization is doing very well and where we need to increase our investment — in terms of product portfolio, marketing, etc. As we look to fiscal year 2027 and beyond, the priority is how to interact with customers in areas of differentiation and ensure we have the right investments in our customers. We're empowering our customers to successfully seize monetization opportunities. As a result, Mike spent a lot of energy internally and with the team to prepare for “IP Factory 2,” AI monetization, and joint solutions. All of this is taking up a significant amount of Mike's time as we look to 2027 and beyond.
Sitikantha Panigrahi
Mizuho Securities US Research Division
He's an amazing talent.
Sassine Ghazi
Chief Executive Officer, President and Director
Yes, we're excited to have him on board.
Jay Vleeschhouwer
Griffin Securities Research Division
Sassine, let me ask you a question first. This might not be a perfect analogy, but how do you see the potential migration or adoption benefits of the new multiphysics product portfolio and subsequent products compared to the transition from ICC to Fusion Compiler back then? That time apparently didn't include Ansys components at the time, but that was your last architecture or generational change. So how do you compare this ongoing transformation now to that time?
Sassine Ghazi
Chief Executive Officer, President and Director
Yes Jay, thanks. This is actually a really good question. What we're doing at Fusion is introducing sign-off capabilities and advantages to the physical implementation stage, because at that time, customers had to iterate at a later stage to associate or sign off chips. The situation is exactly the same here, only physics has been added now.
RedHawk, HFSS, and the rest of the ANSYS product portfolio—how can you bring not only an approximation but an actual engine into the design phase to reduce iterations, enable convergence processes, etc.? Our investment in the Fusion platform — remember that the Fusion platform doesn't piece tools together through a common interface or user interface — we've invested heavily at the data model level. This allows our R&D team to write code and optimize on the same infrastructure and data model. This is exactly where Shankar and his team have invested heavily from day one to deliver the exact same value we've built on Fusion.
Jay Vleeschhouwer
Griffin Securities Research Division
OK. Second, the “engineering reengineering” concept you've been talking about since last year's Converge conference is still a very interesting and promising one for you and your customers. The question is what elements does the customer need to make it happen? Is it a series of changes or variables that need to be made? Or is there a magical single solution? For example, your joint presentation on DAC with AMD and Microsoft around Discovery looked very interesting and could be a catalyst for these changes. So please talk about what actually needs to happen to achieve the entire “reengineering” goal?
Sassine Ghazi
Chief Executive Officer, President and Director
Yes. We talked about Fusion as a key component of having a complete asset set from specifications to sign-off. This is a critical part of Synopsys' complete asset set. Now, integrating workflows that can use the speed of AI, from co-pilots, generative AI to autonomy, is a direction we are investing a lot of resources and working hard to advance. The two demos we did — thank you for mentioning Microsoft and AMD, the fully autonomous EDA workflow shown on Microsoft Discovery — think how powerful this is: you can start from one specification, and a cognitive layer can orchestrate multiple task agents to deliver results that meet the specifications.
At the same time, what we showed with Nvidia at DAC was a demonstration of the ability to run smart devices autonomously for a long time. This is a critical part of the redesign workflow. What excites me most is that when you dig deeper into what I've just described, it's not just the assets we own; these assets are being consumed exponentially to allow autonomous processes to deliver on the promises our customers expect — efficient design, better power, better performance, better cost — and that's what we're doing.
Gary Mobley
StoneX Group Inc.
Sassine, you mentioned in your press statement that Ansys' biggest deal this quarter was a GPU-based digital twin. Is the reason for this large deal because it uses accelerated computing (as opposed to CPU-based computing)? What I want to know is whether the benefits from the efficiency of the accelerated computing digital twin system go to Nvidia or to you?
Sassine Ghazi
Chief Executive Officer, President and Director
Thank you Gary for the question. Simulation is the perfect application of GPU acceleration because these tasks can largely not only be parallelized, but can also be significantly accelerated. This acceleration isn't limited to 10x or 15x. In the field of CFD, we have seen acceleration of 40x, 50x, and 60x. For our customers, the bottleneck is the time to get results, the time to get accurate results. In fact, investment in GPU acceleration began even before Synopsys bought Ansys. After the acquisition, we only further accelerated our commitment and investment. Of course, cooperative investment and collaboration with Nvidia on this opportunity also helped.
About who gets the value — we sell and capture the full value increment of the GPU. Of course, Nvidia is in the background — you need GPUs to switch from traditional CPUs to GPUs — and they benefit in that way. Therefore, the reason for the large size of this deal is the benefit to the customer and the acceleration we are able to deliver to the customer.
Gary Mobley
StoneX Group Inc.
My question is, I would like to ask about the situation in the Korean market. Revenue from South Korea appears to be on the rise, approaching 20% this year. This is truly an outstanding performance. Is this a reflection of the strength of the memory market? Or is it a reflection of how much more market share you are retaining at Samsung and what many people are speculating about? Maybe you can provide some relevant info.
Sassine Ghazi
Chief Executive Officer, President and Director
I'm not sure what those speculations are. What I can tell you is that our relationships with Samsung, SK Hynix, and the wider market in Korea have always been excellent partnerships. This region is leading the way in key parts of AI infrastructure, and we are leading the way among these customers, and we are not limited to these two, but also covering a wider range of regions and the entire product portfolio. So you need to look at everything from IP (including our custom HBM and HBM partnerships with leading customers) to EDA and ANSYS. We are very pleased with our performance.
Wei Chia
Citigroup Research Division
It's great to see the IP business back on track. I remember that a few years ago, the team proposed that the long-term growth of the IP business would be in the middle double digits. Since then, we've seen a significant acceleration in chip design activity, particularly among hyperscale cloud vendors. Also, it appears that the non-AI part is stabilizing as well. In this context, is it unreasonable to think that the IP business is likely to achieve stronger growth than the medium double digit set in the next few quarters? If so, what factors might hinder the upward adjustment of this growth algorithm?
Sassine Ghazi
Chief Executive Officer, President and Director
Kelsey, thanks for the question. Your assumption is reasonable — more chips are launched, huge opportunities to customize chips (I call it “Factory 2”), while “Factory 1” continues to deliver and expand. Currently, medium double digits are our long-term guideline. We look forward to sharing more information and providing any updates as necessary in September.
Concluding remarks
Sassine Ghazi
Chief Executive Officer, President and Director
Thanks for all the questions. On the first anniversary of the completion of this transformative acquisition of Ansys, we are executing, gaining momentum, and expanding our leadership position from chip to system with a focused attitude. I look forward to seeing you all on Investor Day in September. Thank you very much.