This conference call transcript was computer generated and almost certianly contains errors. This transcript is provided for information purposes only.EarningsCall, LLC makes no representation about the accuracy of the aforementioned transcript, and you are cautioned not to place undue reliance on the information provided by the transcript.
spk01: Good afternoon. My name is Brianna and I will be your conference operator today. At this time, I would like to welcome everyone to the Cadence second quarter 2024 earnings conference call. All lines have been placed on mute to prevent any background noise. After the speaker's remarks, there will be a question and answer session. If you would like to ask a question during this time, simply press star, then the number one on your telephone keypad. Thank you. I will now turn the call over to Richard Gu, Vice President of Investor Relations for Cadence. Please go ahead.
spk07: Thank you, operator. I'd like to welcome everyone to our second quarter of 2024 Earnings Conference Call. I'm joined today by Anur Devgn, President and Chief Executive Officer, and John Waugh, 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 forward-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, we'll present certain non-GAAP measures which should not be considered in isolation from or as a substitute for GAAP results. Reconciliation of GAAP to non-GAAP measures are included in today's earnings release. For the Q&A session today, we will ask that you observe a limit of one question and one follow-up. Now, I'll turn the call over to Anirudh.
spk09: Thank you, Richard. Good afternoon, everyone, and thank you for joining us today. Cadence delivered strong financial results for the second quarter of 2024 with broad-based momentum across our product portfolio. Bookings was stronger than expected, leading to a healthy backlog and underscoring the robust demand for our innovative technologies. We exceeded our outlook on all key metrics and are updating our revenue guidance for the year to over 13% year-over-year growth. John will provide more details on both our Q2 results and updated outlook for the year. Generational trends such as hyperscale computing, 5G, and autonomous driving, all underpinned by the AI super cycle, are driving strong design activity across multiple verticals, particularly in data center and automotive. Along with increasing chip complexity and system companies building their own silicon, these trends are creating tremendous tailwinds for our differentiated solutions. We are steadfastly executing to our intelligent system design strategy, extending our leadership in core EDA while steadily expanding our footprint in the new system design and analysis area. Customers are ramping up their R&D spend in AI-driven automation. Our Cadence.ai portfolio offering unparalleled quality of results and productivity benefits continues to gain momentum with orders more than tripling over the last year. Our solutions are enabling the massive AI infrastructure build out across the semi and system space. Additionally, we continue embedding AI in our EDA, SDA, and digital biology solutions. In Q2, our long-term development partner, NVIDIA, broadly deployed Palladium Z3 to deliver to its next-generation AI product roadmap, further solidifying Cadence's leadership in the industry. A marquee hyperscaler meaningfully expanded its partnership with Cadence in Q2 through a broad proliferation of our Cadence.ai EDA, SDA, and hardware portfolio. The growing Foundry ecosystem is driving increased design activity and creating significant opportunities for our industry-leading products. And in Q2, we expanded our collaboration with several leading Foundry partners. We announced that Cadence.ai digital and analog tools were optimized for Samsung's advanced node SF2 gate all-around process. driving enhanced quality of results and accelerating node migration. We extended our long-standing collaboration with TSMC through a very comprehensive and innovative technology advancement, ranging from 3D IC to design IP and photonics and providing optimized digital and analog full flows for TSMC's latest N2 process technologies. Our Integrity 3D-IC platform is the industry's leading unified design, analysis, and sign-off platform for multi-chiplet architectures. Integrity has been certified for all of TSMC's latest 3D fabric offerings and now has enabled several new features like hierarchical 3D-IC design. We also announced that Integrity has been enabled for all of Samsung Foundry's multi-die integration offerings, accelerating the design and assembly of stacked chiplets. Additionally, we released a complete Intel Foundry EMIB Advanced Packaging Reference Flow that is optimized to work seamlessly with Intel 18A technology. We are also collaborating with multiple foundries to optimize our industry-leading IP cores for AI, HPC, mobile, and automotive applications for the advanced process technology so as to ensure seamless integration into customer designs. We saw strong momentum in our IP business with a delivering 25% year-over-year growth in Q2. As we executed to our profitable and scalable growth strategy, AI use cases HPC and heterogeneous integration were the primary drivers fueling the demand for our HBM, PCIE, GDDR, 112 Gigs, 30s, and UCIE products. We expanded our system IP portfolio with the addition of Cadence Janus network on a chip solution that manages high-speed communications effectively with minimum latency. enabling customers to achieve their PPA targets faster and with lower risk. Emulation and prototyping have become mission-critical elements of chip design and software bringer flows. Following the launch of our market-leading Z3 and X3 platforms, there is robust demand for these best-in-class systems, particularly by AI, hyperscale, and automotive companies. and we continue to ramp up our production capacity accordingly. Verisium, our AI-driven verification platform, continued seeing rapid customer adoption with several market-shaping customers, including Qualcomm, successfully using Verisium SimAI for coverage maximization and achieving up to a 20x reduction in verification workload time. Our system design and analysis business continued its strong momentum in Q2, delivering 20% year-over-year revenue growth. As chiplet-based architectures gained traction, our industry-leading integrated 3D IC platform had increased adoption and expansion from large deployments at 5G, hyperscale, memory, and consumer customers. Our AI-enabled Allegro X design platform is being rapidly adopted and driving competitive displacement as multiple aerospace and defense hyperscalers and EV customers take advantage of the platform's productivity and next-generation capabilities. Allegro X's in-design analysis capabilities are also driving a pull-through of our multi-physics analysis solutions. In Q2, a leading EV auto company forged a strategic partnership with Cadence, making a significant investment across the breadth of our multiphysics portfolio. With the close of Beta CAE in Q2, we now offer a comprehensive multiphysics platform covering electromagnetics, electrothermal, CFD, and structural analysis solutions. Our digital IC and custom businesses delivered another solid quarter. Proliferation of our digital full flow at the most advanced nodes continued, with close to 40 full flow wins over the last 12 months, especially at hyperscalers. With over 400 tape outs, customers are increasingly relying on Cadence Cerebras, the leading AI tool in the industry. as it continues to deliver amazing PPA and productivity benefits. For example, Cadence Cerebras has been delivering up to a 10% PPA gain for a global marquee systems company and is now deployed as part of the default flow for their latest designs at the most advanced nodes. Samsung Foundry leveraged Cadence Cerebras in both DTCO and implementation to achieve more than a 10% leakage power reduction on their SF2 gate all-around platform. SocioNext utilized Certus closure and Tempus sign-off to reduce timing closure time by 73% and double productivity while reducing memory cost by 90%. Our AI-driven Virtuoso Studio is the leading automated solution for analog and RF designs. And its new AI features allow much more efficient migration from one process node to another. Virtuoso Studio added 35 new logos in Q2, led by top hyperscalers, aerospace and defense, and automotive customers. In summary, I'm pleased with our Q2 results and the continuing momentum of our business. The AI-driven automation era offers massive opportunities, and the co-optimization of our comprehensive EDA and SDA portfolio with accelerated computing and AI orchestration uniquely positions us to provide disruptive solutions to multiple markets. Now I will turn it over to John to provide more details on the Q2 results and our updated 2024 outlook.
spk08: Thanks, Anirudh, and good afternoon, everyone. I'm pleased to report that Cadence delivered strong results for the second quarter of 2024, finishing the first half with backlog of approximately $6 billion. Also, we expanded our multi-physics platform in Q2 by completing the acquisition of Beta CAE. Here are some of the financial highlights from the second quarter, starting with P&L. Total revenue was $1,061,000,000. GAAP operating margin was 27.7%, and non-GAAP operating margin was 40.1%. And GAAP EPS was $0.84, with non-GAAP EPS $1.28. Next, turning to the balance sheet and cash flow, Cash balance at quarter end was $1,059,000,000, while the principal value of debt outstanding was $1,350,000,000. Operating cash flow was $156,000,000. DSOs were 49 days, and we used $125,000,000 to repurchase Cadence shares in Q2. Before I provide our updated outlook, I'd like to share some assumptions that are embedded. Our updated outlook includes Beta CAE, and it contains the usual assumption that export control regulations that exist today remain substantially similar for the remainder of the year. Our updated outlook for 2024 is revenue in the range of $4.6 to $4.66 billion. Gap operating margin in the range of 29.7 to 31.3%. Non-gap operating margin in the range of 41.7 to 43.3%. Gap EPS in the range of $3.82 to $4.02. Non-gap EPS in the range of $5.77 to $5.97. operating cash flow in the range of $1 to $1.2 billion, and we expect to use approximately 50% of our annual free cash flow to repurchase Caden shares. With that in mind, for Q3, we expect revenue in the range of $1 billion and $165 million to $1 billion and $195 million, gap operating margin in the range of 27.7 to 29.3%, non-GAAP operating margin in the range of 40.7 to 42.3%, GAAP EPS in the range of 83 to 93 cents, and non-GAAP EPS in the range of $1.39 to $1.49. And as usual, We've 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.
spk14: In conclusion, I'm pleased with our strong Q2 results.
spk08: We exceeded our outlook on all key financial metrics. A good finish for the first half and ongoing demand for our solutions sets us up for strong growth in the second half of 2024. As always, I'd like to close by thanking our customers, partners, and our employees for their continued support. And with that, operator, we will now take questions.
spk01: Thank you. We will open the line for questions. At this time, I would like to remind everyone who wants to ask a question to please press star, then the number one on your telephone keypad now. We ask that you please limit yourself to one question and one follow-up. We will pause for a moment to compile the Q&A roster. Your first question comes from Charles Shi with Needham and Company. Please go ahead.
spk03: Hey, good afternoon. Thanks for taking my questions. Anirudh and John, maybe the first question, I do want to ask a fairly big question, a big picture one. So you did pick up your outlook for the year, but some of that really comes from beta CAE. The broader question is the global semiconductor sales, it's on track to grow a lot faster, let's say, compared with you and your peer synopsis. But this seems to me kind of like a reversal of the trend of the last three years when you actually did outgrow the semiconductors. But with so much AI being a big driver for semiconductors, We do wonder whether it's either through pricing or through some other measures, Cadence can actually gain a little bit bigger piece of the pie from overall semiconductor, especially from AI. I don't know if you can provide some thoughts today. I'm not necessarily asking how to change the trend in terms of the value capture, but any thoughts would be great. Thanks.
spk09: Yeah, hi, Charles. Thanks for the question. I mean, first of all, I'd like to say that, you know, overall we are pleased with how we are performing. You know, if you step back, because you asked a longer-term question, right? If you step back, you know, we will deliver. We expect, you know, more than 13% revenue growth and, you know, about 42.5% operating margin. So I think that's a best-in-class combination of both revenue growth and operating margin. And then if you look at our CAGR over the last three years, which is one of our kind of favorite metrics, that's also performing pretty well in terms of growth and margin expansion. And you mentioned semi-cycle. I mean, it's encouraging to see that there is going to be growth this year, which it was not there last year. But as you well know, Charles, we are tied to the R&D spend more than the revenue of our customers. And of course, if the revenue goes up, you know, they're more likely to spend on R&D. But in general, you know, our customers, you know, both system and semi-companies continue to spend on R&D, and these are, you know, long-term projects. So we'll see how that goes, you know, as the semiconductor revenue improves. But this is not an instantaneous effect on R&D spend. You know, there's always some, you know, lag sometimes. But we are encouraged to see the improvement in semi-spending overall, you know, semiconductor revenues. So I would like to say, and you can see in our backlog also, we maintain a pretty healthy backlog. So overall, I think things are performing well, and this AI is broadening out. I mean, you know this well. AI is broadening out beyond data center, which we are glad to have great partnerships, to automotive, to more edge consumer devices like phones and PCs. So overall, I feel pretty good about the industry and, of course, our position in it. as the essential provider of design software.
spk03: Got it. Maybe a quick follow-up on China. It looks like China revenue is still pretty light in the second quarter. So I recall you were thinking maybe China contribution is probably going to be slightly less than the mid-teens or less than 15%. even if, let me assume, the China revenue gets to like a 14%-ish, it still implies a little bit of a second half re-acceleration of China revenue growth. Is that still the case, or you think maybe compared with three months ago, China actually may get a little bit weaker as you previously thought? Thanks.
spk08: Thanks for the question, Charles. And that's I mean, regional revenue is notoriously hard to predict. I will say that at the midpoint of our current revenue guide, we only need China to get to 13% of overall revenue to be able to hit that midpoint. I mean, when you look at performance in Q2 in the first half, we had a very strong bookings first half, very pleased with customers' response to our new hardware systems. The IP and SDNA businesses continue to grow strongly. um core business and continue to scale really well and you know we're focused on profitable revenue growth i know in your first question you indicated that we hadn't raised the outlook but we did raise non-gap eps by six cents uh we're very pleased with the improvement in profitability um and when you look at the the current guide we're actually on track for 50 incremental margin excluding the impact of beta ce now beta c is in our guide um but it's in our guide at what we previously communicated in the press release 40 million dollars of of revenue and about 12 cents dilution to non-GAAP EPS. There is an impact to Opcash as a result of beta CE as well. But overall, very, very pleased. We thought it was prudent to assume lower China revenue for this year at the midpoint of our guide, but that's it. So we only need 13% to get to the midpoint of guidance.
spk03: Thanks, Anirudh and Gian, for the additional color. I appreciate it. Thanks.
spk01: Our next question comes from Gianmarco Conti with Deutsche Bank. Please go ahead.
spk15: Yeah. Hi there. Thank you so much for taking my questions. So on my first one, could you talk a little bit about the implied Q4 ramp-up to 29% growth at the midpoint of guidance? And what has given you the confidence in reaching the targets? Is it mostly hardware visibility coming through, or are there an unusually high number of Q4 renewals that you're waiting for? Any color here would be great. Thank you.
spk08: Yes, Gianmarco. I mean, there's no real change from what we said last quarter. I mean, it's effectively the shape of the revenue curve for the year. We're expecting upfront revenue, a lot more upfront revenue in the second half. It's just the timing of shipments, really. Upfront revenue typically comes from IP, hardware, and to a lesser extent, some software on the SDNA side. With the hardware, it takes time to build the systems. We have higher revenue in Q4 versus Q3 as a result. But also from IP, we recognize revenue in IP based on the timing of deliveries. We're confident in that guide. It's just the shape of Q3 and Q4 is what we have in the guide now.
spk15: Okay, great. So my follow-up will be on hardware. If you could talk a little bit about how much visibility you have actually in H2. Are you booking and delivering in the same quarter, hence why we're not seeing a major uplift in backlog growth? Or is there a different amount to it? I'm trying to understand if you're booking, manufacturing, and delivering all in the same quarter for hardware, essentially. Thank you.
spk08: Thanks for the question. Yes, in some cases on the newer systems, there is a timeline, a lead time to building the systems. We have more bookings than our ability to actually fulfill those bookings, but we do have some inventory of the older systems. We're able to deliver those in the quarter. Yes, I mean, there's always a mix. We did have a challenge in the past with with getting inventory and building the inventory as fast as we could for the amount. I think we've dealt with a lot of that. You'll also see in the Opcache guide, we're planning to purchase a significant amount of raw materials for building inventory in Q3. That's the biggest portion of the change in Opcache guide.
spk09: Also, just to add on the overall hardware, cycle, as you remember, we launched new systems in April, you know, just a couple of months ago, a few months ago now. And the response to them has been phenomenal. Actually, we, you know, these palladium, especially both palladium and protium, but these systems can design chips, like I mentioned last time, with capacity of 1 trillion transistors. And the current biggest chip is like 200 billion transistors. Most of them are You know, a hundred billion or less. So we are five to 10 X higher capacity than what is needed. You know, so that should suit the industry well for, you know, next several years. And I'm also what pretty pleased about is that we delivered, uh, production deployments of, of our system, new systems to some very major customers. So we highlighted, uh, you know, Nvidia, our development partner, this significant deployment of Z3. also one of the leading mobile system mobile companies in the world and one of the leading hyperscalers so it's across multiple markets that we delivered our latest systems which are performing exceedingly well so that sets up very well for the future and also competitively and we have a significant lead given the nature of our systems you know it's a combination of proteome is based on fpga And then Palladium is based on our own chip at advanced DSMC process. And, you know, Cadence is the only solution that does that and provides a unique value. So overall, I think hardware business is performing well. And as you know, these are multi-year upgrade cycles. So, you know, this is not all in 24. So we'll see how things go in 25 and 26.
spk14: Got it. Thank you.
spk01: Our next question comes from Vivek Arya with Bank of America Securities. Please go ahead.
spk11: Thanks for taking my question. So on an absolute basis, in fiscal 24, organic sales growth rate is robust. But in terms of revisions, it has stalled, right? Essentially, no real movement since what you suggested at the start of the year. So I'm curious, Anirudh, how has the year transpired versus what you thought? And how do you think about bookings and backlog trends into the second half? Should we expect that backlog stays around the $6 billion? Will it start to Just I'm trying to understand that should we be thinking about sales accelerating from here or, you know, this being kind of the sustainable growth rate for the company?
spk09: Yeah, I have a very good question. So, in general, what I would like to say is, you know, like we mentioned last two times, you know, the shape of the curve this year is unique to, you know, Cadence, you know, given multiple factors, uh you know this is not what we expected last two years so this time it's more back and loaded for the reasons we mentioned before so so the guide is a little different and we are also you know given that shape of the curve more prudent in our guide like we were in q2 and then we rather overachieve and and deliver that and give the team flexibility to to you know uh do the right business for the long term so i think that's the difference this year versus last few years is Given the shape of the curve, we have more prudence in our revenue guide, like John mentioned. And John can comment on the backlog expectations.
spk08: Yeah, I mean, we don't guide bookings, but we were very pleased with the strong bookings in the first half. And I get the question, Vivek. I mean, essentially, you know, we're seeing strong demand for our hardware systems. We're seeing strength across all our businesses. And I guess your question is that when you add in beta CE, you're not really taking the revenue guide up. I think... Essentially, if your question is what would we like to see improve, I think it's the China revenue percentage. It was 12% in Q1, 12% in Q2. It improved in Q2 over Q1, and we think it'll continue to improve through the year. But right now, our guide only needs to get to 13% China to hit the midpoint of that guidance.
spk11: And from my follow-up, you mentioned beta CAE, quite a drag to EPS. I think you mentioned 12 cents. dilution and almost, I think, what is like a $300 million hit to operating cash flows. Can you describe that acquisition a little more? And when does it start to become accretive to your financials? Thank you.
spk08: Yes, Vivek. On the $300 million drop in operating cash, just to clarify that, that about 40% of that 300 million drop is due to M&A. I mean, in things like beta CAE, some of the purchase price, the geography of where the cash impact goes, some of that payment flows through Opcash. The bigger portion of the impact on operating cash is our plan to purchase a lot of inventory raw materials for the hardware demand that we're seeing. We're pre-purchasing a lot of inventory. So you'll see our inventory spike in Q3 with all of the raw materials we're purchasing. We want to make sure that we have all the raw materials necessary to ramp up the build out of our hardware systems. And then in relation to BDC, I mean, it's very recent acquisition. It's no different to what we have in the press release. In fact, on the press release, we said we were expecting 40 million revenue at the midpoint. That's embedded now in the guide. We're expecting 12 cents dilution. on non-GAAP EPS. That's also in the guide now. And we expect it to be, I mean, operationally, it'll be accretive next year, that there are some interest costs associated with the debt, but we think it'll be accretive next year.
spk09: Also, a couple of things to clarify. So one thing, this purchase of inventory for the hardware systems, I mean, that will be used over multiple years. It's not just for 24. So I think it's a one-time investment that pays for several years, and that's a prudent decision to make to get the right kind of parts for the future. And then on beta, it completes our system analysis portfolio to add structural analysis. It also strengthens our position in automotive. Of course, data center is a big part vertical with all the AI super cycle. But I think one of the other exciting verticals is automotive with all this electrification and also AI getting added in self-driving or driver assistance. So we see a lot of design activity in automotive. Also automotive is also moving through chiplets and 3D IC. So I think automotive has all the three tenets of our ISD strategy. It has silicon content that is increasing. And more and more system design, of course, is needed for the design of automotive and AI for all the data and computational software. So for that reason, you know, Beta CAE completes our portfolio in automotive and positions as well in the future. And this is not just with the semiconductor companies doing automotive, but also the system companies now. You know, OEMs doing more and more chip design, doing more of our system solutions. And I also want to highlight and congratulate McLaren. There was a big news this weekend. McLaren got one and two in Hungarian F1. And we have been working with them for the last few months and years. And it's good to see them do well as we deploy. So I think the automotive solution that we are driving is a combination of silicon system and then AI. And we are seeing the results of that through organic and inorganic expansion.
spk14: Thank you.
spk01: Our next question comes from Joshua Tilton with Wolf Research. Please go ahead.
spk14: Hey, guys. Can you hear me? I'm clear, Josh.
spk13: Great. The first one is just kind of more of a clarification. I know there's been a lot of questions around the mix in upfront versus recurring. I guess what I'm just trying to understand is, And I could be wrong with my math here, but it feels like it was the upfront component was still a little light in 2Q. And now we're a little bit more second half weighted, more 4Q weighted because you need time to develop inventory. Am I thinking about that the right way?
spk08: That's fair. Josh, I would do the inverse on you in terms of bookings were stronger than we expected in Q2 and we got some uplift and recurring revenue. Took a bit of pressure off on the upfront side. And we are, I mean, we're taking orders. We've got strong demand for the hardware and we're building those hardware systems as quickly as we can, particularly the newer hardware orders. IP is doing really well. and system design analysis is doing really well. And what we've reflected in the guide is our expectation of how much of that revenue will fall in Q3 and Q4. We took the opportunity, we really de-risked the guide for the year by reducing our expectations for China. Upfront, we still expect to be in a range of 80% to 85%, but I think we might get slightly more recurring revenue as a result of the strong bookings in the first half.
spk13: that makes super clear. And then I guess just my followup to that is, I guess it's another visibility question, but how much of what's baked into the guide from an upfront perspective, do you feel like you have like, you know, good inventory levels to meet that guidance or does the guidance that you put out today still require you to build and develop inventory between now and shipping those boxes?
spk08: Yes, but it's, it's, um, We definitely need to build hardware and you'll see the impact on our inventory in Q3 with the amount of raw materials we're purchasing. But as Andrew says, that's a one-time thing that we're doing to try and get raw materials and to build those systems as quickly as we can. But a lot of the upfront revenue in the second half comes from the strength in our IP business. And we have those orders in backlog and it's just a case of executing against those. We also have some SD&A or system design analysis up from revenue that's scheduled to occur in Q3 and Q4. Again, most of that is from orders in the system. On the hardware side, it's kind of mid to high single digits is what we were expecting the SVG group to deliver to be able to hit the midpoint of that guidance.
spk13: Super, super helpful. And then just a quick follow up is really awesome to see the recurring revenue growing sequentially this quarter. Is there any way you can maybe help us on what the expected recurring versus upfront mix is supposed to be in 3Q? And then I'll leave the floor.
spk08: Oh, I don't have that to hand, but let me come back to that.
spk14: Let's see if I can dig it out here. Thanks, guys. Congrats.
spk01: Our next question comes from Ruben Roy with Stifel. Please go ahead.
spk05: Yes, thank you. John, just a very quick question, and then I guess a follow-up, and then I'll ask a real question. But on the inventory purchases, am I right in assuming that that's mostly for the Z3X3?
spk06: And has anything changed in terms of when you're thinking about general availability of those hardware products?
spk08: Yes, that's correct. But the vast majority of the purchases are to get raw materials to help build those new systems.
spk06: And then in terms of the type?
spk09: Yeah, just to clarify, I mean, also, I mean, we have two systems, right? So palladiums, we design ourselves and we manufacture the TSMC and protium. We also design ourselves, but the silicon itself, is primarily from AMD with Xilinx FPGAs. So a lot of this purchase is for X3s and FPGAs, and that should serve us for multiple years. On Z3, like we said, we're already shipping them, and they're already deployed in production this quarter. So I think Z3 is slightly different than X3 in terms of the mix of the silicon content, just to clarify that.
spk06: Okay. I apologize. I thought they were going to, you know, sort of certain customers are not generally available, but thank you for that. And then the real question is just around some of your top customers have been accelerating the rhythm of bringing, you know, sort of their very complex shifts to market. NVIDIA and AMD certainly have accelerated their roadmaps to, you know, sort of a one-year rhythm. Are you seeing any changes in sort of the way, you know, your business is impacted or affected, you know, kind of by the acceleration of their product roadmaps yet?
spk09: Yes, I would like to, I think we're seeing more and more design activity, you know, like you said, the rhythm or the cadence of the products. And also different kind of chips. You know, it's not just the big data center chips, but even within them, you know, there is more and more customization. Of course, the hyperscale is doing their own silicon. And then now, you know, we talked about our partnership with, for example, Qualcomm. And they are doing, you know, consumer or edge, you know, laptop AI devices. So the amount of AI is also spreading to other verticals, not just obvious the big one on data center. And data center design is accelerating, you know. And I think when we look at it, we still see that the data center part of AI still should accelerate at least the visibility we have for next couple of years. So we'll see how that goes. And therefore, and the other thing is automotive. Automotive takes normally a little longer, but it's already seeing design activity and the deployment maybe a few years down, maybe after data center. And then consumer and PCs already starting with phones and laptops. overall we do see accelerating deployment of ai into the whole semiconductor ecosystem and we are very proud of our position in it you know whether it's 3d ic you know whether it is data center chips whether it's our own ai products you know we are we are winning almost all kind of engagements on all on our kind of cadence.ai portfolio so overall we do see more and more design and deployment of AI infrastructure and our own AI product.
spk08: And if I could just come back to Josh's question, sorry, if I could just come back to Josh's question on the revenue mix for Q3, that for recurring revenue, we expect, sorry, 80% to 85% of revenue to be recurring for the year, and Q3 includes the middle of that range, and then the balance is...
spk14: So you can do the math and work out what the upfront piece is.
spk01: Our next question comes from Jay Vlieschauer with Griffin Securities. Please go ahead.
spk12: Thank you. Good evening. Anirudh, question about the evolution of the product portfolio for EDA generally and perhaps for SDMA specifically. what I'd like to ask about is how you're thinking about packaging the products. Over the last year, you've introduced a couple of products with the term studio in the name, and I'm wondering if you're thinking about more and more bundling or packaging of that kind via that nomenclature for the EDA products, and then specifically for SDNA, now that you do have multiple codes, how are you thinking about packaging or integrating across the various simulation codes that you've assembled now in acquisition, then I'll ask my follow-up.
spk09: Hi, Jay. Good question. So, I mean, as you know, in EDA, when we go to lower nodes, you know, there is more integrated solutions which are required, you know, with digital or analog or verification. And And that is further accelerated by use of AI. So, you know, like Cerebras, for example, in digital will integrate, you know, not just place and route, but also synthesis and sign-off. So I think that trend is definitely there. And same thing with, you know, Verisium, our leading AI product for verification, also integrates the four major verification platforms we have. So it is more and more platform-driven approach. And we can do that now with SDNA now that we have a complete portfolio. And we mentioned like a leading EV company, like OEM, one of the leading, the most advanced EV companies deployed our entire portfolio. So as we have a bigger portfolio in SDNA, it does let us do what we have always done in EDA, focus on solutions, not just on individual products. and integrate solutions with our own kind of native integration, whether it's analog, digital, verification, and now with SDNA.
spk12: All right. As follow-up, I know it's still quite early in the propagation of AI and ML by you and your peers to the customers, but are you beginning to see any commonality or convergence towards a relatively small number of use cases that customers are mostly employing the tools for? And then relatedly, are you also seeing AI ML adoption having any meaningful effect on your services revenue?
spk09: Yes, Jay. So what I would like to say is that the number of use cases I see is increasing at this point. I mean, of course, one of the You know, one of the biggest use case that we started with was digital implementation. Since it is so kind of heavy kind of design process, so automating the digital implementation process was a huge benefit. And we talked even this quarter, Cerebras being deployed at one of the leading system companies for the default flow, also used by Samsung. Also, you see verification being used by Qualcomm. So I think what is happening in that Cerebrus or the implementation use case, two things. One is that it is going not just for design, but also for DTCO, you know, design technology co-optimization, and also for higher level in the design process like floor planning and 3D IC exploration. So it's all not just for implementation of the design, but also for architecture and exploration. And the other thing is there's more workflow automation. As customers get used to Cerebras, they're using it not just towards the end of the design process, they're using it right from the beginning throughout the design process. So it allows us to do more workflow automation, and Cerebras has also evolved to allow much more of an entire workflow rather than a specific implementation use case. And then same thing is happening in terms of, you know, more and more use cases. For example, packaging, you know, Allegro X is doing pretty well. And, you know, recently one of the leading customers in 3D IC used this capability to automate, for example, routing for automating, you know, placement, which was not there before in PCB and package design. So overall, I do think it's maturing of the workflows. And then with these LLMs and Gen AI, we have several workflows for taking spec to RTL, and we highlighted some of them last quarter. So I actually do see finally that we are always kind of building out the AI infrastructure, these big companies designing chips. But I do see now there's a turning point in deployment of AI for the design process. with the initial workflow being Cerebrus and digital implementation now to expanding to, you know, LLM-based, you know, spec, you know, expanding to DTCO, expanding to 3DIC, of course, expanding to analog, packaging, verification. And we do have the most comprehensive AI portfolio in terms of all five major kind of product lines. So actually, it's a pretty encouraging view compared to a year ago.
spk14: Thank you, Edward. Thank you, John.
spk01: Our next question comes from Harlan Sir with JPMorgan. Please go ahead.
spk04: Hi. Good afternoon. Thanks for taking my question. Is the bookings profile for the full year still expected to be 40% first half, 60% second half? Because if it is, then that would imply book to bill greater than one for the full year, total backlog. you know, up about 9% this year to about $6.5 billion. But I guess how much of that backlog is due to the beta CAE acquisition? What I'm just trying to figure out is X beta CAE, if core cadence orders and backlog are expected to be up this year, which would continue the strong sort of six to seven year trend of increasing orders and backlog for the team.
spk08: Yeah, it's hard. And again, like I say, we're not guiding bookings. But we were very, very pleased with a strong first half for bookings. The beta CEE contribution to backlog is very, very small. It's immaterial because beta CEE, their revenue is upfront. So that's the upfront piece of the business rather than the recurring revenue. But yeah, we're very pleased. I mean, you can typically expect us to always be driving for a book to bill of greater than one, but we don't guide bookings though.
spk04: Okay, perfect. Thank you. Anirudh? There's a pretty interesting dynamic with your memory customers, right? They're big customers of your custom product family virtuoso, but they are moving more and more to advanced digital design, right? The HCM control logic chip, for example, is moving to leading edge technologies and advanced chip design. Similarly, with some of your NAND customers, they're moving towards more of this sort of bonded CMOS periphery to array The periphery chip, again, is also moving towards advanced visual design as well. So are you starting to see more adoption of your advanced digital implementation and verification products by your memory customers? And then does the leadership in memory via virtual also sort of give you an advantage as they bring on more advanced logic design capabilities?
spk09: Yes, Arun, that's a great observation. And we are fortunate to have very deep and long-standing partnership with all the major memory companies. You know, at least there are three big ones, and then, you know, maybe two at the next level. But overall, we are, you know, given our strength, like you mentioned, in Virtuoso, which is the platform for choice for all memory implementation. And yes, there is a lot more digital and implementation design happening at memory companies, you know, primarily driven by HBM and other trends. And actually, you know, Also, there is some kind of, I mean, they were always doing digital, but it's a lot more now. And there is trend of even integrating, you know, TSMC technologies with kind of memory. And given our strong partnership with TSMC, that also helps us with the memory companies. And as you know, they are also doing a lot more 3DIC, you know, all the three big memory companies. And these memory layers are going from, you know, they're actually one of the most advanced 3D IC with the memory layers going from 8 to 12. And that also, you know, plays to our strengths. And even in my prepared remarks, I mentioned, for example, our partnership with Samsung and 3D IC. And then so is true with the other kind of two major players in memory. So we are pleased in our position in memory and the emerging trends of, you know, HBM and 3DIC integration. And we'll see, you know, how that progresses. But I think memory is often overlooked in, you know, I don't need to tell you, but just in general, memory is often overlooked in the big AI super cycle. It's not the big chips, logic chips, but memories play a very essential role And we are very well positioned, both with the leaders like NVIDIA on the logic side, and then we highlighted Samsung and the other big memory players in general.
spk14: Thank you for the insights.
spk01: Our next question comes from Jason Salino with KeyBank Capital Markets. Please go ahead.
spk02: Great. Thanks for taking my questions. So lots of questions so far on the hardware timing, but I think John, on prior calls, you said that the hardware delivery times typically are like eight to 10 weeks. That's what it is for like a normal cycle. But are you saying the lead times for the Z3 X3 are longer than this because the demand is much better than what you're seeing?
spk08: Demand is strong. Demand is strong. What I was trying to point out was that we do have inventory of the older systems that we can deliver right away. um the newer systems we're having to build them as quickly as we can because the orders are coming in faster than we can build them um so the lead times is a bit of a moving target in that respect we we are planning to purchase a lot of raw materials and and build as quickly as we can in q3 so you'll see a significant uptick in our in our inventory balance at the end of q3 okay now that's helpful and then just a clarification i think i think you were saying like the sda group to hit your guide needed to do like
spk02: made the high single-digit growth? I'm not familiar with SDGs. Is that the functional verification kind of guide for the year? And then if you're parsing out, does that imply what do we need to see on IP, since that's the other upfront component to hit the guide? Thanks.
spk08: Yeah, that's fair. Yeah, IP is having a really, really strong year, as is system design analysis. They'll be our two fastest growers for the year. And then again, there's some upfront revenue from them. It's more weighted towards Q4 versus Q3. So we have the shape of the curve is really driving our guide. And I would categorize it as prudent. Anirudh, would you add anything?
spk09: No, that's right, John. And sorry for the acronym. SVG is System Verification Group. So when we say SVG, that means verification. So I think that's what John was implying. Verification should grow, but right now we are not assuming massive growth in verification for this year. But it should grow compared to last year. And then, like you mentioned, some of the upfront revenue is also IP. If you remember, we highlighted in Q1 a new partnership with Intel. You know, and they are, you know, we are deploying our IP portfolio for the Intel process. So it takes time to do that and deliver to that. And some of it is in Q3 and Q4. And also this time we talked about our expansion, you know, partnership with Intel on both on EMIB, you know, their packaging and 3DIC platform and 18A. Just to clarify.
spk14: Yeah, thanks for the lingo acronym help. Very helpful.
spk01: Our next question comes from Lee Simpson with Morgan Stanley. Please go ahead.
spk16: Great. Thanks for fitting me in and well done and a good quarter. Just wanted to get some clarification. I don't know if I heard correctly, but I think I heard you say that a mobile OEM had taken the Z3 platform. And if that's the case, do you have a sense for what the emulation work might be? Would it be for chips on device or would it be for chips both on device and perhaps in, let's say, a network situation? Thanks.
spk09: Well, good question. You know, we don't comment on individual customer or specific customer use cases, but in general, you know, these hardware systems, as you know, are used both for chip design and for system software bring-up. So both use cases are there, and, you know, we are the leading platform given Palladium and Proteum. And this is true for all, even in the AI use case, you know, even in the data center AI use case, a lot of it is for software development. Actually, Palladium is the platform for choice, even our AI chip customers to give a model to their customers, you know, because even before they have a chip, they can give a Palladium model to see how it performs. So it's both for chip design and also for system design and system software. And that's true for multiple major verticals. data center, mobile, automotive, these things. Yes, thank you.
spk16: Yeah, thanks. Just on those multiple verticals, if we look at the incidence of 3D ICs coming through, I get the sense that that's starting to hit the tape now in automotive. You have mentioned EV companies as a collaboration of late, but you have mentioned a number of chip makers also. I wonder if it's possible at this point, or even if it's relevant, to just maybe talk about the split between the customers. Are we talking system customers, i.e. OEMs and Tier 1s in the majority right now, or is it still major semis, chip makers for the automotive work? Thanks.
spk09: Great question. So first, like you correctly pointed out, I think the 3D IC is – a lot more prevalent in automotive than, let's say, 6 to 12 months ago. And of course, the original genesis is HPC and data center AI, but now all these chiplets and 3DIC platforms are moving to automotive. And I think over time, we'll move to other verticals like consumer. They already moved to laptop, for example. Several of the laptop chips are 3D IC, but I think that will be the progression. It will gradually go to all verticals, but definitely active and automotive. Because as you know, with the chip-led architecture, the customer doesn't have to redesign all the chips. And also they can use some standard chips and have some specific chips which are more value-added for them. And that's particularly true in automotive as each OEM wants to differentiate versus the other OEMs. So this trend is not just for semi-companies, to your question. It's also there in OEMs. And I think actually even more, I do think that the 3D IC trend makes even more sense for end OEMs because then they're able to customize and differentiate versus the other. So we are seeing that, and we are seeing that in other geographies as well, you know. Because, as you know, China is pretty strong in EVs, and then U.S., and then Japan, there's a lot of activity. So overall, I think automotive, there is more activity on 3DIC, including both Semi and OEMs.
spk16: Great. Thanks, Anirudh.
spk14: Great cover.
spk01: Our next question comes from Clark Jeffries with Piper Sandler. Please go ahead.
spk17: Hello. Thank you for taking the question. My first question is, you know, Anirudh, how do you expect the delivery of these third-generation systems to translate to additional software consumption in the recurring revenue portfolio? You know, these products are happening with verification acceleration software bring up, but, you know, how do you see that additional consumption panning out after the delivery of a new ZRX system? And then I have one follow-up.
spk09: Yeah, great question. I mean, like what John was saying, when we say SVG, you know, system verification group, so hardware is part of that group. You know, even though hardware is a significant business, we organize it as part of verification. And one of the big reasons for that is apart from in verification, apart from the hardware systems, we have a lot of other verification products, you know, which are doing pretty well, like Jasper for formal verification, Exelium for logic simulation. And the customer is looking for an integrated solution on verification. To the earlier question that Jay had, you know, about, you know, what is happening in SDA, you know, in EDA we always have believed for the last several years that it's going to be an integrated solution in verification. So the stronger our hardware products get, we do expect it should help our software verification products, you know, things like Excellium and Jasper and Verisium. Because some of the hardware capacity is also used for what is called SimXL, simulation acceleration, in which they use Palladium to accelerate logic simulation. So there is a natural tie-in between verification software products and verification hardware products. Now, exactly how it pans out, we just have to see. But the strength in hardware should help us in our overall portfolio strength.
spk17: Perfect. And then one follow-up for John. I think just to kind of finally put a cap on the whole discussion around timing, I guess then is it fair to say that that sort of $600 million of upfront revenue in the second half, maybe a majority of that is coming from IP and SC&A and not necessarily the Gen 3 systems and that maybe the interpretation is that there's going to be more of a demand curve in the beginning of 2025 rather than this $600 million being strongly driven by third-gen planium and protium. Is that a fair takeaway?
spk08: I think that is, Clark. Yeah, that's exactly right. I mean, we always knew that it would take time to build the hardware system, so we originally included that in our guide in the first place.
spk14: All right, perfect. Thank you for taking the question.
spk01: Our final question comes from Joe Vrewink with Baird. Please go ahead.
spk00: Great. Thanks for fitting me in. I did want to follow up and stay with verification and just this raw material investment. So the thing I'm trying to reconcile is I would imagine you entered this year expecting the new platform, strong demand, needing to scale production and therefore invest in inventory. So I guess I'm wondering what changed in the quarter that warranted this updated assumption for cash flow and raw material purchase? And really at the heart of the question, did something change about your hardware demand expectation, not necessarily for 2024, but maybe out into 2025? And we just happen to be getting that news now because of the need to update your cash from ops forecast associated with the inventory input.
spk08: Yeah, that's a great question, Joe. I mean, the leaders of that business, I spent plenty of time with them this last quarter because they were monitoring what the demand was like for the new systems. Demand is quite strong. And then the key thing to highlight here is it's a one-time multi-year purchase of inventory raw materials. They feel very, very confident in the longevity of these systems and the longevity of that demand. And they wanted to pre-purchase multi-years of inventory in Q3. Now, that was news to us, if you like. So we thought that one-time thing, we wanted to include it all now in Q3 and not impact. That doesn't impact next year. I mean, it'll be favorable for next year's operating cash.
spk00: OK, great. Then lastly, you mentioned at the start, orders for Cadence.ai tripled year over year. I don't think that's possible without also getting a lift in the base business, both across EDA and SDNA. I guess on an ATV run rate basis, what has the AI lineup meant for Cadence overall? And does this create a step up in value where as you start pulling these contracts from backlog and coming quarters, it'll become more noticeable in revenue and we'll kind of see the AI contribution more than we have to this point?
spk09: Yeah, good question. I mean, AI is adding, you know, like I was mentioning before, I mean, it's almost become like table stakes now. So all our new contracts include our Cadence.ai portfolio as customers get more and more kind of used to using them. And once you start using the AI portfolio, it's difficult to go back to not using them. So without getting into like what happens exactly to future revenues and bookings, you know, we are always cautious about that. But in general, there is uptick. with more customers deploying AI and whenever we have new contracts, you know, we are including them as it makes sense in that.
spk14: Okay. Thank you very much.
spk01: I will now turn the call back to Anirudh Devgaon for closing remarks.
spk09: Thank you all for joining us this afternoon. It's an exciting time for Cadence with strong business momentum and growing opportunities with semiconductor and system customers. With a world-class employee base, we continue delivering to our innovative roadmap and working hard to delight our customers and partners. On behalf of our board of directors, we thank our customers, partners, and investors for their continued trust and confidence in Cadence.
spk01: Thank you for participating in today's Cadence second quarter 2024 earnings conference call. This concludes today's call. You may now disconnect.
Disclaimer