Cadence Design Systems, Inc.

At that time, simply press the star key followed by the number one on your telephone keypad. Thank you. And I will now turn the call over to Richard Gu, Vice President of Investor Relations for Cadence. Please go ahead.

Thank you, operator. I would like to welcome everyone to our fourth quarter of 2024 earnings conference call. I'm joined today by Anirudh Devgn, President and Chief Executive Officer, and John Wall, Senior Vice President and Chief Financial Officer. The webcast of this call and a copy of today's prepared remarks will be available on our website, Cadence.com. Today's discussion will contain four looking statements, including our outlook on future business and operating results. Due to risks and uncertainties, actual results may differ materially from those projected or implied in today's discussion. For information on factors that could cause actual results to differ, please refer to our SEC filings, including our most recent forms 10-K and 10-Q, CFO commentary, and today's earnings release. All forward-looking statements during this call are based on estimates and information available to us as of today, and we disclaim any obligation to update them. In addition, all financial measures discussed on this call are non-GAAP unless otherwise specified. that non-GAAP measures should not be considered in isolation from or as a substitute for GAAP results. Reconciliations of GAAP to non-GAAP measures are included in today's earnings release. For the Q&A session today, we would ask that you observe a limit of one question only. If time permits, you can re-queue with additional questions. Now, I'll turn the call over to Anirudh.

Thank you, Richard. Good afternoon, everyone, and thank you for joining us today. I'm pleased to report that Cadence delivered exceptional results in the fourth quarter, capping off a strong 2024 with 13.5% revenue growth and 42.5% non-GAAP operating margin for the year. We exited 2024 with a record backlog of $6.8 billion, which is a testament to our compelling AI-driven chip-to-systems portfolio and its growing proliferation among marquee, system, and semi customers. The AI super cycle is entering a new phase with generative AI, agentic AI, and physical AI fueling an explosion in compute demand and semiconductor innovation. We see AI adoption unfolding in multiple phases, starting with the build-out of the AI infrastructure, where we are already deeply engaged. At the same time, we are actively integrating AI into our own products, and we are also exploring AI potential to create new markets. The infrastructure AI phase is well underway, and we have been closely collaborating with market leaders on their next generation AI designs across both training and inferencing. In Q4, we advance our longstanding partnership with NVIDIA through commitments across a range of our EDA, hardware, IP, and system software solutions. We are also using NVIDIA's latest NEMO and NEM microservices to build customized GenAI applications, delivering enhanced optimization and productivity. We deepened our collaboration with Qualcomm through a significant expansion of our EDA and system software solutions. Also, we meaningfully expanded our strategic partnership with Marvell through a broad proliferation of our portfolio of products. Our Cadence AI portfolio continued gaining strong momentum with market-shaping customers. Our AI-powered products, such as Cadence Cerebras, SimAI, and Allegro XAI are proliferating at scale. And our LLM-based design agents, powered by ZAI, data platform are showing promising results in early engagement. Cadence Cerebras is rapidly becoming an essential part of the design flow as it continues to deliver transformative PPA benefits to customers, with more than 750 tape-outs to date and over 300 in Q4 alone. Our Verisium AI-driven verification platform with its SIM AI and debug AI apps gain share at key competitive accounts as customers embrace the significant boost in verification quality and efficiency. Additionally, we are also applying AI to new market opportunities such as life sciences with our OpenEye drug discovery software. The growing Foundry ecosystem is driving increased design activity and creating significant opportunities for our products. In 2024, we strengthened our collaboration with existing Foundry partners and entered into new strategic engagement. We further our partnership with TSMC with AI optimized design flows certified for TSMC's N3 and N2P technologies. We strengthened our collaboration with Global Foundries and are partnering with Samsung on their SF2 gate all-around process. We also partnered with Intel Foundry to provide design software and leading IP solutions at multiple Intel advanced nodes and in Q4 entered into a strategic collaboration with Rapidus for two-nanometer enablement technology. We deepened our partnership with Arm through a broad expansion of our IP, hardware, and AI-driven design enablement solutions, and in Q4, successfully taped out industry's first Arm CSA standard-based system chiplet. Now let's talk about some of the specific product level highlights for Q4 and 2024. Our system design and analysis business delivered strong results, achieving over 40% growth in 2024. Our multi-physics analysis platform with AI-driven optimization is delivering superior results to a rapidly expanding customer base across multiple verticals, especially aerospace and defense, and automotive. Our Millennium CFD simulation platform ramped up over the year, closing two meaningful deals in Q4 in the aerospace and defense and energy verticals. With Beta CAE, we now provide a comprehensive multiphysics platform covering electromagnetics, electrothermal, CFD, and structural analysis. Beta CAE performed ahead of our expectations with major expansion at several marquee automakers, including at some of the fastest growing EV companies. Our Allegro X design platform with its significant productivity and next generation capabilities continued its momentum especially at aerospace and defense, hyperscale, and EV customers. Allegro XAI, the industry's first fully automated PCB design engine, is enabling customers to realize up to a 10X productivity gain. Our Integrity 3D IC platform, with its unified design, analysis, and sign-off capabilities for multi-chip-led architectures, expanded its footprint at hyperscale and memory customers, OSADs, and foundries. Our IP business drove a strong finish to the year, growing 28% year-over-year in Q4. Our AI HPC protocols, including our flagship HBM, DDR, PCIe, and UCIe solutions, propelled our business. with significant expansions and competitive displacement at top tier customers. We continue to broaden our IP portfolio with high growth star IP products and last month we entered into a definitive agreement to acquire SecureIC. The addition of their embedded security IP products will augment our rapidly expanding portfolio of leading edge silicon proven IP including interface, memory, AI, and DSP solutions. We announced a collaboration with Rapidus to provide a broad set of our advanced memory interface IP for their two nanometer backside power process node. And ASD Space Mobile committed to our AI-driven IP solutions and EDA tools on their AST5000ASIC, a custom low-power architecture for global space-based cellular broadband services. Our core EDA business comprising our digital, custom analog, and verification portfolios grew 15% year-over-year in Q4. Insatiable demand for more compute along with system complexity and the need for first-time right silicon continue to drive strong demand for our best-in-class palladium Z3 and protium X3 systems. Our hardware family delivered yet another record year, adding over 30 new customers and almost 200 repeat customers in 2024. with particularly strong demand from AI and hyperscale customers. Our digital portfolio had another strong year, gaining 36 new full-flow customers in 2024, including 17 in Q4. Cadence partnered with EqualOne on the development of its Alpha 5 quantum SOC that runs at cryogenic temperatures with our design and implementation tools being central to this revolutionary design. Our Virtuoso Inspector franchise solutions tackle the most complex design and simulation challenges in analog, mixed signal, and RF design. Virtuoso Studio, delivering industry's leading AI-powered layout automation and optimization, continued its strong ramp and is now deployed at over 450 customers. MediaTek adopted SpectreX running on NVIDIA's Hopper GPUs for its two nanometer designs, achieving up to a 6X performance boost while maintaining full accuracy. SpectreFX FastPy is now in production usage at more than 75 customers. including top memory vendors, as well as SOC and mixed signal companies. In closing, I'm pleased with our outstanding performance in 2024 and excited about the business momentum and opportunities ahead. As the AI era continues to unfold, our AI-driven EDA, SDA, and IP portfolio, powered by GenAI agents, and accelerated computing is delivering transformative results, uniquely positioning us to capitalize on these massive opportunities. Now I will turn it over to John to provide more details on the Q4 results and our 2025 outlook.

Thanks, Anirudh, and good afternoon, everyone. I'm pleased to report that Cadence delivered an outstanding Q4 in 2024 with broad-based strengths across all of our businesses. Robust design activity and customer demand combined with our strong execution drove 13.5% revenue growth and 42.5% non-GAAP operating margin for the year. Fourth quarter bookings were exceptionally strong and we entered the year with a record backlog of $6.8 billion and a record CRPO of $3.4 billion. Here are some of the financial highlights from the fourth quarter and the year starting with the P&L. Total revenue was $1,356,000,000 for the quarter and $4,641,000,000 for the year. GAAP operating margin was 33.7% for the quarter and 29.1% for the year. Non-GAAP operating margin was 46% for the quarter and 42.5% for the year. Gap EPS was $1.24 for the quarter and $3.85 for the year. And non-gap EPS was $1.88 for the quarter and $5.97 for the year. Next, turning to the balance sheet and cash flow, our cash balance was $2,644,000,000 at year end, while the principal value of debt outstanding was $2.5 billion. Operating cash flow was $441 million in the fourth quarter and $1.261 billion for the full year. DSOs were 48 days, and we used $550 million to repurchase cadence shares during the year. Before I provide our outlook for 2025, I'd like to share some assumptions that are embedded. Our outlook is based on our usual assumption that export control regulations in place today remain substantially similar for the remainder of the year. At the midpoint of revenue guidance, we are assuming 2025 China revenue will be flat year over year. And our non-GAAP EPS outlook continues to be based on a tax rate of 16.5%. In our outlook for 2025, we expect revenue in the range of $5.14 to $5.22 billion. Gap operating margin in the range of 30.3 to 31.3%. Non-gap operating margin in the range of 43.25 to 44.25%. Gap EPS in the range of $4.19 to $4.29. Non-GAAP EPS in the range of $6.65 to $6.75. Operating cash flow in the range of $1.6 to $1.7 billion. And we expect to use approximately 50% of our free cash flow to repurchase Cadence shares in 2025. For Q1, we expect revenue in the range of $1.23 billion to $1.25 billion, gap operating margin in the range of 27% to 28%, non-gap operating margin in the range of 40% to 41%, gap EPS in the range of $0.93 to $0.99, and non-gap EPS in the range of $1.46 to $1.52. And as usual, we published a CFO commentary document on our investor relations website, which includes our outlook for additional items, as well as further analysis and gap to non-gap reconciliations. In conclusion, I'm pleased that we finished the year with record backlog and CRPO. We are starting 2025 with a very strong outlook for non-GAAP incremental margin. As always, I'd like to thank our customers, partners, and our employees for their continued support. And with that, operator, we will now take questions.

Thank you. And at this time, I would like to remind everyone who wants to ask a question to please press star 1 on your telephone keypad. As a courtesy to all participants, we ask that you please limit yourself to one question. And we will pause for just a moment to compile the Q&A roster. And your first question comes from the line of Jason Salino with KeyBank Capital Markets. Your line is open.

Hey, thanks for taking my question. You know, when we look at the backlog number, you know, the 1.2 billion of sequential improvement, that's like the biggest we've ever seen, much higher than kind of what I think some of us were expecting. But then when we look at the guide for the full year, the 11.5% starting point is a little lower than where you started other years in the past. So maybe, John, can you help us just reconcile, you know, the the confidence in the backlog and kind of how you're approaching, you know, guidance for 2025. Thank you.

Yeah, sure, Jason. Thanks for the opportunity to clarify. Yeah, we're very pleased with the strength and momentum that we carried into year end and into this year. We did allude to it in a call last time and we thought the pipeline looked extremely strong and we just wanted the team to focus on converting that. We saw broad-based strength across, you know, all of the businesses on backlog Um, on bookings for Q4, but it, but it was a strong renewal quarter, uh, Q4 that, uh, in contrast, I mean, the, the, when I look at this year, second half is probably stronger for renewals in the first half. So we'll probably burn some of that backlog in the, in the first half of the year, but it'll, it'll come back, uh, in Q3 and Q4, uh, and the duration, uh, our backlog duration is normally somewhere in the region of 2.4 to 2.6 years every year. And we ended up at the high end of the, of, of that range for this year. Um, when I look at the year in total for 2024, we were super pleased with, uh, with the strength X China, the growth outside of China with high teams. Um, China, of course, declined just over a hundred million dollars in dollar terms, um, from, from 2023 to 2024. And we did back test that. I think I mentioned that on the, on the last call as well, we went back to 1999 and looked at the last 25 years of, of performance in China in dollar terms. And there were only three years where we had down years, and 2024 was the third time we experienced a down year in dollar terms in China. We've never had two consecutive down years in China, and we thought it would be prudent to assume in the guide for 2025 that China was flat, because it's notoriously hard to predict. And I want to manage an expense plan based on a prudent guide.

And your next question comes from the line of Joe with Wells Fargo. Your line is open.

Yeah, thanks for taking the question. I wanted to follow up on that note on the China piece. One, was there any change into the RPO balance from the export restrictions? And then two, I guess on that flat kind of expectation for 2025, should we assume that the hardware cycle specifically in China is maybe more of a 2026 driver then?

Oh, yeah, good question. I mean, basically, I was just trying to be prudent with the guide and make sure that we were managing expenses very effectively for 2025. But, I mean, when I look at the booking activity for Q4, it was quite strong. I think we're seeing strong design activity in China. Anur, do you want to comment on any of that?

Yeah, you know, actually, it's good to see design activity across the board, you know, and in all verticals. Now talking specifically about China, actually I was there in November and it's good to see design activity picking up, right? I think even our China results in 24 improved through the year from Q1 to Q2 to Q3 and Q4. And China, especially as you know, there's a lot of design activity in, for example, in automotive, where at least five or six like major auto companies, you know, driving the EV revolution and, you know, Almost all of them are, are designing chips now. And, uh, we have privilege of working with all the, all the major auto companies there. And also with our, you know, new systems portfolio, especially including beta, you know, not only we are engaged with them on the chip design side, but also on the packaging board and system side. So what all I do is think that the China activity is strong, like it is strong in, you know, a lot of other regions. I think in terms of the guide, it's just best to be prudent and then see how it goes. Given all these macro uncertainty, we thought it's better to be prudent on the guide, but the design activity seems to be actually picking up and we'll see. It's difficult to predict how that goes through the year, but so far, so good.

Joe, you mentioned the specific metrics in terms of CRPO. There's no real material impact on any of that from what we're seeing in China that you're right to call it hardware though. I mean, we had a strong hardware year in 2023 in China and it was kind of weaker in 24. And that was, that's why the, the, the year is down year over year in China. And I just didn't want to be predicting what would happen for 25. It's just very, very difficult for us to predict what happens in China for 25.

And your next question comes from the line of Ruben Roy with Stifel. Your line is open.

John, I hate to waste my one question on another follow-up on China, but I am getting this question from investors and just around the entity list, wondering if there had been any changes. It seemed like there were additions to the entity list kind of going into the end of the year and Is that informing sort of your prudent view on China, or is it really just kind of coming off of that down year and kind of assessing renewals and kind of timing of hardware ramp? Or has something changed, I guess, is the question around that detail list. Thank you.

Yep. Thanks, Ruben. No, I think I would put it down to general prudence. There's nothing really changed. in China, but we did see we had a down year, our third down year in 25 years, and we just thought it would be prudent to assume flatten that would de-risk the guys for everybody.

And your next question comes from the line of Vivek Arya with Bank of America Securities. Your line is open.

Thanks for taking my question. I'm trying to, you know, get a better handle on what's the right interpretation of the recurring revenue growth that is slowing down. I assume, obviously, China is one aspect of it. But if I were to just take, you know, the 24 recurring revenue, right, and then look at, I think you are guiding 80% recurring revenue. So that does suggest a bit of a deceleration. Why is that? Vivek Narayanan, The market for Ai and all these other products is improving so much, why is the recurring revenue sales for decelerating so much.

Vivek Narayanan, I be back thanks for the question the yeah when we look at recurring revenue growth, of course, the. Vivek Narayanan, The assumption that we have in for China being flat is a headwind for that recurring revenue growth, because, like if we look at the impact. of China decelerating by like 100 million from 23 to 24. The vast majority of that was in core EDA and mainly on the hardware side. But we're seeing that impact the recurring revenue numbers. Of course, the recurring revenue is the expected mix for this year is expected to go to 80% recurring, 20% upfront. But that's more a reflection of the strength that we're seeing in our upfront revenue businesses. They're growing faster than the average cadence business. Anirudh, do you want to talk to any of that?

Yeah, absolutely. I mean, of course, I think we're good to see strength across the board, but especially hardware and IP is performed really well in 24. And I expect that they will do well in 25. You know, there are multiple reasons driving that, which we can get into, but You know, we have a phenomenal product in hardware. You know, we're well ahead of, you know, what the customers will need for several years. And in terms of all our kind of, you know, engagements with the customers, they're very pleased with our hardware system. So we had a record 24 also. We had, you know, strong demand hardware going forward. And same thing on IP. You know, IP, I think, has been a weaker area for us over the years. But I think it did turnaround you know of course one year doesn't make a trend but it did turn around significantly in 24 so I think is up IP is up roughly 30 percent in 24 and is driven by multiple reasons so you know one reason our portfolio is just just better technically you know the PPA especially at Advanced Nodes you know we are delivering better power performance area for IP products Also, customers are, you know, we are engaged now with almost all the major customers in IP. We always engage with them on EDA and hardware. At this point, customers, you know, want more and more of our IP. And the third reason is, as you know, there is a lot more foundry activity. You know, not just, you know, the leading foundry, you know, from Taiwan, but also in other parts of the world, whether it's Intel or Samsung or Rapidus, global foundries. So that also requires more and more IP and EDA. So as a result, you know, we are, you know, I am more confident in the IP business going forward. I think it should have another good year in 25 from what I can see right now. And also we are investing more in the IP business in terms of expanding the portfolio. So as you see, we acquired, like last year, we acquired, you know, about a year ago, IP products from Rambus. And then recently we acquired security IP, which is another critical piece of the IP portfolio. So overall, I think ADIS has been always very strong in EDA, always very strong in 3DIC, always being very strong in hardware. And IP has been one area that we wanted to do more, and especially 24 is the turning point for IP. And together positions us well into the future.

And your next question comes from the line of Lee Simpson with Morgan Stanley. Your line is open.

Great. Thanks for squeezing me in here. I really just wanted to ask around the chiplet reference designs that you've clearly done very well with in China with EV makers. Is this the sole source of strength for the chiplet reference design? And maybe if we X that business out of China for this year, What does that do for the size of the client that we might see in China inferred in that flat number? Thanks.

Ali, that's a good question. I mean, first thing I would like to say is that chiplet design is happening across the globe and also in all verticals. It is not specific to a particular region or vertical. Of course, it started with, as you know, with high-performance computing with all these big AI chips uh and server products but now you know automotive and you know even laptops they're all getting disaggregated okay so we are always been focused on 3d ic and chiplet based design from i don't know maybe 15 years ago you know maybe it was too early we were i think our first reference flow with tsmc on 3d ic was like 2007 or 2008 okay but of course it has picked up steam in the last few years And we are particularly well positioned, you know, given we have EDA products, packaging products, system products, and also IP needs to go that way. So we have, you know, what I would call IP 2.0 is more focused not just on individual IP, but, you know, silicon solutions and chiplet architecture. So that's why you're seeing investment for us on chiplets, not just on the EDA side, but also on the IP side. But I think it's going to happen, I believe, across all geographies, across all verticals, and it's only going to accelerate going forward.

And your next question comes from the line of Charles Shi with Needham & Company. Your line is open.

Hey, good afternoon. Anirudh. John, I think I do want to ask a little bit about the core EDA. It looks like based on the segment information, the core EDA growth in 24 was somewhere in the high single-digit range. It seems like it has been growing at more like a double-digit. But at the same time, the system design and analysis continue to grow at a very strong double-digit growth. If you can give us a little bit of your long-term view, is this going to be the new normal going forward, like the more traditional EDA actually going to more like a high single-digit grower, but SD&A is more like where that can contribute to cadence outperformance going forward? I want to get your long-term thoughts there. Excellent.

Hi, Charles, that's yeah, generally in core EDA, I mean, we're always targeting double digits, uh, revenue growth and, and profitable and sustainable revenue growth is really important to us. Um, it would be very, very pleased with the progress that we're making with the hardware business over the last number of years that, uh, of course the, you know, the down year China from 23 to 24 impacts that, uh, that core EDA growth year over year from 23 to 24. Um, and then our assumption for 25 obviously impacts that as well. But we're delighted with the value we're providing to customers and with just the sheer demand for our products right across the core EDA space. Anirudh, do you want to add anything?

Yeah, in terms of product portfolio, I think we are better positioned than we ever have been. If you look at analog, digital verification in terms of core EDA and design activity is strong. Of course, in 24, there are two issues like John mentioned. One is China, and then also hardware had a transition year. So even though Q4 was phenomenally strong in hardware, we went through the transition in Q2 and Q3. So that's, of course, part of Core EDA. So we'll see how it proceeds going forward, but we are well-positioned. We are doing very well with customers in all these evaluations and new design starts. So we'll see how things progress going forward.

And your next question comes from the line of Gianmarco Conti with Deutsche Bank. Your line is open.

Yeah. Hi. Thank you for taking my question. So firstly, congrats on another fantastic quarter. Perhaps just a follow-up question on hardware. Given, again, the impressive backlog number and the increase in up-from-revenue we saw this quarter, is it fair to assume that Z3 and X3 is being adopted at a faster pace and that the mix of volume and pricing is what's driving a strong Q1 guide? And going forward, is it fair to assume an acceleration also in up-from-revenue given the closing of the air pocket? I think a fair few investors are looking at the strong, harder ramp and wondering if there is much of it inside that 6.8 billion backlog number. Any color on that would be great. Thank you.

Yeah, Gianmarco is a tremendously strong bookings quarter for hardware in Q4. We're seeing huge demand across all of our functional verification platform. So we were delighted with the progress that we've made there. We have strong backlog leading into the first quarter. So, yeah, I mean, you're seeing a more normal kind of shape to revenue this year for Cadence. We expect about 24% of the revenue to come in Q1. Similarly, about 24% in Q2. Second half of the year is going to be hard, or you expect Q4 to be higher than Q3, but we'll be able to tell you more about that in the second half when we see the pipeline for hardware in the middle of the year. Anirudh, anything?

Yeah, hardware demand is phenomenal. Of course, Q4 strength is not based on just hardware, even though it was phenomenal. Like I mentioned, IP is doing fabulous and overall 3DIC, overall our complete EDA portfolio systems growth as you saw in Q4. But specifically on hardware, I mean, as you know, we have now both new systems, Z3 and X3. They are in full production. We gained several competitive wins in Q3 and Q4. And it's the strength of our portfolio, which is a custom system. chip that we make which is in palladium z3 and fpga system for for protium and almost all like large designs especially all large ai designs are using palladium as as the platform of choice and i do expect that this strength will increase over time so because we are the only company that designs his own chip you know to do emulation so if you look at this you know some of the biggest uh chips in the world are designed through palladium and you know we have more than 100 chips in one rack which are liquid cooled connected optically and then you can connect 16 of these racks together so it's like more than 2 000 full reticle tsmc chips are used to emulate the world's most complicated designs and there is no other system that even comes close so the nearest competitor to palladium z3 is actually palladium z2 okay and then i think going forward You know, this is the nature of the architecture. Also, the inherent advantages it has over FPGAs. And we like FPGAs. It's useful for some software workloads in Proteum. But the gap between custom ASIC and FPGAs is only going to increase in the future. Because even Z3 is a more advanced node than FPGA. And you can see that in other workloads too, not just in emulation. You know, all the AI workloads, all acceleration, you know, custom ASIC is way superior to FPGA. And going forward, that will continue. And in terms of, you know, we have at least a 10-year lead in designing these custom ASICs for emulation. So hardware was phenomenal, you know, and we expect that to continue. Now, of course, the actual second half, we'll have more update. You know, we have good backlog for the first half, and depending on what happens in the first half, we can update the rest of the year, in the middle of the year, yeah.

And your next question comes from the line of Jay Fleashour with Griffin Securities. Your line is open.

Thank you. Good evening. Under John, your backlog came in at about a half billion above where we were for the quarter. So I assume that NVIDIA alone could certainly have accounted for a large portion of that. But your current RPO year over year growth progressively slowed over the course of the year. from 14% year-over-year in Q1, 11% in seven, now 6% year-over-year as of Q4. Maybe you could dissect that in terms of the composition of software versus hardware and IT, and then maybe on the geo side, what the current RPO growth might look like if you were to exclude China.

Yeah, Jay, thanks for the question. I think what you're focused on there is the annual value of the backlog. And yeah, we were very focused that we need to grow that annual value. And I think that would be an important focus point for us for 2025. We're very pleased with the amount of bookings that we recorded with customers in Q4. Very, very strong pipeline got converted. But we know we can do better on the annual value, and we're going to work on that for 2025. X China, of course, China is a headwind to that number. We would expect the CRPO to improve naturally in 2025. But I would expect the backlog number to decline a little bit in the first half. We'll burn some of it in the first half just because of the timing of wind renewals. come up in the year. Q4 2024 was a strong renewals quarter. Q3 and Q4 are kind of the strongest renewals quarters for 2025.

And your next question comes from the line of Gary Mobley with Loop Capital. Your line is open.

Hi, guys. Thanks for taking my question. John, you seem to imply that revenue will be split 48%, 52% first half versus second half of the year, but you're OPX guide or your operating margin guide seems to apply the opposite directional change for non-GAAP OPX. Normally, you would see a second half increase versus first half. What's the dynamic there in the OPX for the second half versus the first half?

Hey, Gary, thanks for spotting that and giving me the opportunity to explain. We changed to a We normally have a mid-year merit cycle where we increase salaries in July. We've moved that to January and started in 2025. So the profile for the year, typically in the past, you would have seen us improving margin Q1 to Q2 and then typically taking a drop off from Q2 to Q3. This year, we'd expect the profitability profile to improve right throughout the year with Q2 margins exceeding Q1, Q3 exceeding Q2, and Q4 exceeding Q3. We thought it was better to move to an annual cycle from the start of the year so as not to cause that kind of disruption in the quarter over quarter numbers from Q2 to Q3 going forward.

And your next question comes from the line of Harlan Suhr with JP Morgan. Your line is open.

Good afternoon. Thanks for taking my question. On the full year guidance, you know, if you drive within the upper range of the guidance for this year, your three-year revenue CAGR is going to drop below 15% to be more specific, more like 13.5% for the first time in a number of years. And if I look at the environment, right, you've got AI leading edge design starts and activity at an all-time high. You're going through a strong upgrade cycle on your hardware business. I know you're taking a more conservative view on China, which I think is prudent, but outside of China, you know, we know auto and industrial semiconductor trends are still fairly muted. We also know a couple of your large leading edge IDM customers continue to face competitive difficulties. So outside of China, can you guys just help us understand, like, where else are you taking a more cautious view on the 2025 outlook?

Yeah, I think what you're referring to is that, I mean, when you look ex-China, I mean, if we exclude China, that last year we were high team year over year growth. And what's implied in the guide is low team year over year growth. I mean, normally we start the year prudently anyway. And I think what you're highlighting is we believe we provide tremendous value to our customers. And it's important for our customers to extract that value from what we're providing. So we focus on proliferation and adoption of our technology. And we're confident we'll extract that value from customers over time. But essentially, that's all that's in the guide. China is the kind of headwind, I guess, the assumption that that's flat. But we're pleased with low team growth across the rest of the business.

Just to point out, yeah, one thing to point out is, of course, we are always looking at both, you know, revenue growth and profitability. So I think we expect not only what I would consider good revenue growth in 25, but also good improvement in profitability. So if we add them together, like rule of 40, I think this is the first time we are guiding more than 55 in the beginning of the year. So it's very important to focus on both those things like we have done for years now.

And your next question comes from the line of C.T. Panigrahi with Mizuho. Your line is open.

Taking my question. Can you hear me? Yes, yeah. Okay, great. So, Anil, you talked about some of your AI products, Cerebras, CMAI, and Electro, all that getting traction. Help us understand how is that driving your ACV uplift right now? Are you seeing more impacts on the you know which segment like digital design verification any of that and in the same context also could you talk about i know you talk about age uh you know uh design going into more age devices as another um area of growth when do you think that's going to when you're going to see that infection point yeah those are two good good questions so

In terms of AI product itself, like we have mentioned in the past, we have these five major platforms, which now at this point, we are engaged with all our major customers on. Of course, a few years ago, we started with digital implementation, which is Cadence Cerebris. I think at this point, Cadence Cerebris is adopted with all our major customers. And then verification with Vericium and then Allegro X. you know, Optimality, you know, Virtuoso Studio. And these are all platforms which are, you know, something you have to add to your base config, right? So we are seeing customers adopt more and more of these solutions. Now, in terms of your question on edge, you know, I always believe that I said so many times in the past that know ai will have multiple phases to it of course the first phase was more infrastructure more on the on the data center side but i believe one of the biggest phases will be what we have called for a long time is physical ai which will be more in the physical world you know more edge devices more you know cars and robots and drones And then also sciences AI, you know, with things like life sciences and material sciences. So it will go through at least these three big horizons. So the good thing with the physical AI, you know, which is more on the edge or more, if there are more power constraint, you know, applications, we already see a lot of design activity starting on that. Now, some of the products may come out, you know, a few years later, but just like on the data center side, you know, we saw products in 22 or 23, but the design activity was there a few years before that. And especially in physical AI, they are by nature edge and also more mixed signal in which cadence is very well positioned. So that's why I was saying, you know, even though there is weakness right now in the autos in terms of semi-revenue, but in terms of investment in R&D, there's a lot of investment in R&D with EVs and self-driving. Same thing with drones and robots, which is independent of humanoid robots. Industrial application and aero and defense applications will be huge for those things. So I think that will drive a different kind of AI and require even more domain-specific silicon for AI on those devices.

And your final question comes from the line of Joshua Tilton with Wolf Research. Your line is open.

Hey, guys. Thanks for sneaking me in at the end here. I'm just going to keep it to two quick clarifications. My first one is on China. I know there's been a lot of questions, but I guess, John, I'm just trying to understand, has anything changed since last quarter that's either increasing or decreasing your conviction that China won't have two down years in a row? And then on the hardware side, if I remember correctly, historically, you know, you talked to how you tend to only guide for like six months of visibility, specifically for hardware. Are you taking a similar approach this year as you set the guidance for 2025?

Thanks, Josh. Taking the second question first, yes, exactly the same approach. Hardware is a pipeline business. We typically have a good visibility into the pipeline about two quarters out. We finished with a very, very strong hardware bookings quarter in Q4 and have plenty of backlog to deliver against in the first half of the year. that demand the pipeline still looks strong, but visibility into the pipeline this time of the year is more difficult. We'll have more visibility into hardware in the second half by the time we get to the summer. So exactly the same approach that you've seen us do in previous years. And then your question on China, can you repeat it again?

Has anything changed since last quarter that's either increasing or decreasing your conviction that China won't have two down years in a row?

I was just worried about my ability to predict what China revenue would be. I was very pleased that all through last year, Q1 was the low quarter. Q2 was higher than Q1. Q3 was higher than Q2. Q4 ended up higher than Q3 for China revenue. But it's tough to have visibility into the pipeline for Q1 because quarters are only getting assigned now and the salespeople tend to keep their cards very close to their chest this time of the year. So I thought in the absence of any other information, the prudent thing would be to de-risk the guide and assume flat for China. And then if our experience is better than that, we'll take the guide up through the year.

And I will now turn the call back to Anirudh Devgn for closing remarks.

Thank you all for joining us this afternoon. It's an exciting time for Cadence as we enter 2025 with product leadership and strong business momentum. Our continued execution of the intelligent system design strategy, customer-first mindset, and our high-performance culture are driving growth and profitability. Fortune and Great Place to Work named Cadence as one of the world's best workplaces in 2024, ranking it number nine. And on behalf of our employees and our board of directors, we thank our customers, partners, and investors for their continued trust and confidence in Cadence.

And ladies and gentlemen, thank you for participating in today's Cadence fourth quarter and fiscal year 2024 earnings conference call. This concludes today's call, and you may now disconnect.