Cadence Design Systems, Inc. (CDNS) Earnings Call Transcript & Summary

All right. Welcome to the 49th Annual JPMorgan Technology, Media and Communications Conference. My name is Jackson Ader. I'm a software research analyst here at JPMorgan. I'm very pleased to have the CEO of Cadence, Lip-Bu Tan here with us this afternoon. We're going to -- Lip-Bu is just going to read something very quickly before we get into the fireside chat. And just a reminder to those in the audience, if you want to ask the question, just hit the button, type it to me, and we will get it -- we'll get it asked to Lip-Bu. But Lip-Bu, if you want to go ahead, I think, with the safe harbor statement, and then we'll do some introductions of the business.

Yes. Thanks, Jackson. Before I begin, I would like to mention that safe harbor statement. Today's discussion will contain forward-looking statement, and we'll make sure that -- make use of certain non-GAAP financial measures, and please see our most recent 10-K, 10-Q and website for the discussion of risk factors and our use of non-GAAP financial measures. Thank you.

All right. Perfect. So Lip-Bu, just a brief introduction of the business, kind of the markets that Cadence serves, end customers, that sort of thing would be great.

Sure. So I think -- thank you, Jackson. I think Cadence is a technical software company that provides software design tools, intellectual properties and hardware verification platform to semiconductor and system -- electronic system customer, that enable to develop silicon chips, PC boards and subsystem. We are executing towards our intelligent system design strategy, that -- the foundation is the EDA and IP portfolio. That is the -- providing that, I call it, the design excellence. And then using our computation of software expertise to expand into the system innovations and pervasive intelligence area. And then that triple our TAM market from $10 billion to $30 billion. We have embarked on that. Our customer base across vertical market, like mobile, cloud, data center, aero, automotive and then up and coming is the industrial medical-related area. And we have a strong culture of innovation. 38%, 40% of our revenue is in the R&D and the field support serving the customer. And we are laser-focused on winning and delighting our customer with a market-shaping customer and then help them to continue to drive innovations and then performing. And then the long story short, I think this working remote actually accelerate a lot of design activity. And so it's really exciting to see that. And then so far, we stick to our strategy, no change and actually accelerating our strategy.

Yes. Can we talk about that acceleration? I mean for the pandemic for a lot of companies, unfortunately, it was certainly a major headwind. But just curious what did the -- what did the macro environment or demand environment look like for Cadence in 2020 through the pandemic?

Yes. So I think clearly, the COVID is quite challenging for all of us. And I think we are kind of coming out to normalcy. And -- but there are still some area region having some challenges, like in India, even Taiwan recently. And so I think we just have to pay attention to it. I think the health and safety of our employees, our customers and partners are most important to us. So far, we are pretty much working from home and have been effective for us. But clearly, as I mentioned in the past, this 5 generation of waves, the 5G, AI, machine learning, industrial 4.0, hyperscale infrastructure change. And it really drive the sub-silicon renaissance. And so I think it's very exciting to see that happen. And in some way, and it really accelerating the digital transformation I mentioned earlier. And also, I think, clearly, when people are working from home, the infrastructure is accelerating and using cloud, using Zoom, and that really drive a lot of momentum and design activity. Actually, we see tremendous increase because of these 5 waves. And so we are excited. When they do design -- more design and more R&D actually is beneficial for Cadence and Synopsys. And then you will be -- they need more of our tool and IP and that's something that we are very excited.

So I'm curious about that. I mean, these design cycles take so long, months, years, in fact. And so I'm just curious, as the chips get more complex and the nodes get smaller, right, we're running out of numbers here. We're down to 2-nanometer, talking about or so. I mean what are you seeing in terms of turnaround times for -- from concepts to tape-out and ultimately manufacturing for these more complex nodes?

Yes. So I think this is a very interesting question that you asked. I think clearly, even with the supply constraint recently, we don't see any design slowdown. I mean there is tremendous increase in design activity. That is very positive for us. And then clearly, we have this tool. One is the Moore's Law and then driving now the process node. And right now, we're heavily on the 2-nanometer now in the development. 5 and 3 is already optimized for our customer and our foundry partners. So I think now this is one part that's going down there. And then the other part is the modern Moore. There's a lot of these new material packaging challenges and then the system knowhow, the RF integrations. And so those are kind of exciting for me. That catering is very well positioned to capture those opportunity. And so I think it really depends on what vertical market we are addressing. Some of the aerospace, defense, automotive, that is really beneficial for us. And meanwhile, the mobile, the cloud infrastructure, the matching down to the Moore's Law, and we are also very well-positioned to work with them and supporting them. And of course, our new system analysis and then some of the system company, right now today's 45% of our customer are system and service provider. They're not looking at the silicon. They have to look at total system. And then what are the electromagnetics field, the temperature, thermal, peak load. And then so those are areas that we are tripled down on it so that we can really meet their requirement and then providing the solution and tool for them to provision for some of these challenges in the design and then make sure that the first time pass is critical for them.

You mentioned a couple of industrial end markets, automotive, aerospace. Those are cyclical, right? And especially coming out of the pandemic, as we start to feel a little bit more normalization in terms of auto production and travel. Are you seeing a change in demand? Or are you seeing a change in customer conversations, maybe from some of these cyclical end markets?

Very much so. I think as I mentioned earlier, the design cycle sometimes is in the quarters and years ahead. And then you have to really design the product before you go into the productions. And then in the automotive case, in the past, the best practice is just in time, and then -- so that they can really drive the efficiency and cost management. But this time, the shortage of the supply is a wake-up call for many of them. And so right now, they're starting to really addressing the supply chain more effectively and more commit longer-term in terms of commitment to the component supply and from Tier 1 or semiconductor players. So our customer -- end customer, they have a better visibility of the requirement so that they can plan better. And in some ways, good, and that dialogue used to be, I don't spend too much time with the General Motor for motor -- we're focused more on the semiconductor player or the Tier 1. And now they are reaching out to us. And then in terms of the design, the methodology, and -- I think it's just exciting to me that they are paying a lot of more attentions and to the whole design innovations. And especially in the ADAS and also some of the sensors that they need to have and then some of the infotainment and the security, it becomes very important to them. And that's why we make investment on the greenfield software and that back into the system innovation besides the system analysis, also the embedded security and how can they make it more secure for the system player, whatever the chip that develop and the system that makes sure that it's really, really good and secure.

And how does the chip shortage or the fact that production -- automotive production has slowed so much? How does that actually impact Cadence financially? And in what areas?

Yes. So far, because we are not so much in the manufacturing, we are more in the design phase, so we don't see much impact to us. But clearly, we're monitoring very closely. They are like a subscript in some of the material that have a shortage that automotive had to shut down their factory because of shortage of components. And so those are more manufacturing related, but we pay a lot of attention to them and make sure that some of this customer, even though it's not our area of focus, we also -- because of my role, besides Cadence, I also have a venture investment. And then they reach out to me for help. And because of some of the supply components, I'm related to them. So in the way that I can help them to secure some of the component they need. And so those, I think, is a value adding. And then later on, hopefully, will be the -- beneficial for Cadence.

From a high level, could you just give investors an idea of what is the Cadence? What is -- what would you say is that the main competitive advantage that you have in core EDA over your competition?

Yes, a couple of area. And as you recall, we put almost 38%, 40% of our revenue into R&D and also in the field support. Delight the customer is my culture at Cadence. And then clearly, innovation is something that we really take pride of. Every year, we have 7 to 10 new products organically developed within the company, funding some of this R&D. And then so in terms of like one example is digital, and we doubled down on the place and route and then we doubled down on synthesis. And then using massive parallelism, AI and machine learning, and now cloud-native initiative we are working on. And then now we can push the whole full flow. And then we have 45 full-flow digital implementation from customer. Stay tuned, we continue to push that front. Likewise, in the verification, we're also pushing that from function verification to hardware emulation to prototyping. And -- so I think we are pushing the whole entire verification suite. And so that we can provide the full flow for our customer. And then likewise, because of customer requirement, they are really excited about our system analysis, the system design and analysis. Right now, we are growing at about 30% in terms of revenue growth. And then a lot of repeat orders from the customer. That is a very good indication that even though it's a new space, we're able to provide more than the incumbent player that can provide up to 10x performance, and they love it. They encourage us to do more. And that's why we bought 2 companies in the computational fluid dynamics. And then there we can really support our -- the aerospace, defense customer and even marine customer, electronic marine area. So I think, again, it just opened up a lot of new areas for us to expand. And so I think we continue to do that in the holistic way, not just an EDA tool. And then the other part is -- I mentioned, is working with the foundry. Packaging become more and more important. So not just the advanced node. As you know, the Moore's Law is slowing down a little bit. And then clearly, there's a packaging, 3D packaging chiplets and then not the silicon packaging, the system packaging. And we are very well positioned. We have all the pieces together. And then we work with our foundry partners to really drive the solution for the hyperscale, AI, machine learning company, and that's something that's really fast growing.

So you mentioned 45 full-flow customers. Is this -- has this been the way that customers have always preferred to buy, and now Cadence feels like they have the full digital flow to compete? Or is this a new phenomenon on the smaller -- on the more advanced nodes?

Yes. Some of our big customer, I think, the top priority for them is the best of tool, the best -- you had to be the best, either in the path and route or the best in synthesis, the best in the sign off. And so that is the first step. That's why we kind of start with the specific tool. We optimize for the performance using AI and machine learning, using massive parallelism and then now using the data analytics, something that we are investing to make sure that they really drive the performance. And then after that, you have other pieces, then you push for the whole platform, the full flow. The main reason customers like to have full flow, when you move down to 3-nanometer and 2-nanometer, unless you build a big team inside your company to optimize with different tools and then make sure that they talk to each other. And then able to drive the solution, the performance you want. And then pretty much, I think when you moved on to 3-nanometer, 2-nanometer, you had to pick 1 vendor to be primary. And because of sophistication, complexity of the design, and you need to pick one, and you only add on 1 or 2 that particular best-of-breed that you can use, then you have to really integrate them with the primary EDA vendor. And that's why it's become very critical and -- for us and for synopsis to own the full flow. And that is the game, and now they are going forward. And then besides the full flow from the EDA side, the other part is IP, the readiness of all the key STAR IP, plus the system analysis, plus the packaging. And that's why we are uniquely positioned. We have all 4. And then we just need to proliferate step-by-step to win the customer trust. And then my philosophy at Cadence is basically trying to be the trusted partner, long-term trusted partner. And we can really work collaboratively with the customer to optimize to their solution whatever the performance they're looking for.

How does the IP portfolio -- how are you able to keep pace with the advanced nodes on the design side with your IP portfolio?

Yes. IP is a new business for us. In 2010, through Denali, we bought that. And so we are kind of a latecomer, but now being late, actually better for us. So we can pick and choose what are the strategy we want to have. So we picked the STAR IP so that IP that are really valuable and that is essential for the customer, like the SerDes, like the Tensilica, those are really STAR IP. And then every process node is the new opportunity for us. And of course, all this IP, the outsourcing trend is very -- is going. And because it's very hard to keep track of the Gen 3, Gen 4, Gen 5 versions and you need a team to really do that. And then the other part is not like EDA. If you have some back is okay, you can work through it. IP, the quality is important. And so if you don't have the good IP, the customer screaming at you. And because they're affecting their production, affecting their revenue. So that is something where the take price, a culture that we need to have that first-time past culture even within Cadence. And we really need to drive that so the first time the customer call on us, the chip -- not the IPO going to be successful. Are we there yet? Not quite there. We have some challenges and like any other IP company, but I think we just work through it. That's something that I inspire my team to really drive that. And then if you can first time pass on the IP, that is a very profitable business, and it's very essential for the customer.

You've mentioned 3D simulation or the system analysis piece that Cadence has moved into over the last couple of years. Simple question, why? What prompted it?

Yes. I think a couple of things. And because of the growing design complexity and then also that they're now big of our -- big portion of our customer system, they really care about the -- this treaty and reliability is very important to them. And so the system analysis become critical for them. So they want to know about the EM, interface, the thermal in some of the peak area how do they provision carefully to design, make sure that you are not provisioned for that. And then that's why we thought that is really the customer requests us, why don't you look into that area? And then you guys have the computation software capability, and we have this particular need. And we have been -- you have been our trusted partner. Why don't you expand towards that? And then with that encouragement and also it's a very nice TAM market. The -- just like we bought CFD, computation of flow dynamics, is $2.4 billion TAM market. And that's exciting for me. And then we bought this 2 company, creating a lot of new customers that we don't have before. And then we can really help them to accelerate, proliferate even faster because they are kind of a small company. They don't have the sales and the financial master to do that. And we can really -- we pick the best technology and then like the meshing, that is very important, the system analysis like point wise. And we can really use our channel, use our capability financial to invest more to help them even more successful. So I think -- all in all, I think clearly, how to integrate all these pieces with our Allegro, our RF and then together with some of the system analysis, join me together and optimize, provide that solution to our customer and the customer love it. And so they encourage us to do more. And so I think -- I always listen to the customer, and they are always the best source of idea and then we can kind of work with them and identify the new product development that we can do. And our R&D team is extremely busy because they are so excited to support the customer and delight the customer.

All right. Let me push back for 1 second. Were there really many semiconductor companies, existing customers that were pushing you to get into computational fluid dynamics? Or was that just a -- is it an attractive market that you saw a nice asset that you could extend into?

Actually, did not portion this from our existing customer. They're asking for that. And in addition, through acquisition, we pick up some new customer that we never had before. And then they're even reaching out to us and then asking us, maybe you can also helping us to help design the silicon that we're also thinking about, the market is big enough. We want to build our own silicon and just like hyperscale players. And right now, each one of them, they are quietly building up their own silicon team because the workload have changed and more towards application-specific and domain-specific processing and then third party is because they basically cannot depend on the general purpose CPU or GPU. The timing, the cycle is not right for them. And they want to provide those solutions quicker at their own timing that they want to do. So they are quietly building up the custom silicon team. And then for them, they don't only look at the silicon, they had to look at the whole system power envelope and then the thermal requirement and then the interference. And so those kind of things that really play to our strength, and that's why they encourage us, "Hey, why don't you guy go and then do something and then helping us so that we can have fully integrated solution design flow that they can use. And then we love it, and we just thought that it will be great to be a trusted partner. We listen to them. And that I think we would want their heart. They said, "Well, you guys are listening, you guys are humble, and you guys are willing to learn our requirement and let's work together." So a lot of R&D, R&D interface and engagement, and that I like a lot. I mean, that's the best thing when your customer engage you in their next-generation of product they want to design and asking you for help.

And what do you think -- I mean, getting into this space into that -- into the -- whether it's electromagnetics or computational fluid dynamics, it's not necessarily greenfield, right? That's a saturated market with existing competitors. What do you think that Cadence can bring that maybe the existing competitors are maybe lacking at the moment? What do you think can be your differentiated angle?

Yes. I think a couple of differentiating angle. One, first of all, is our computational software. And EDA is much more complex. And so we can use that to our own advantage. And also, I think, clearly, we have a lot of know-how on the machine learning, deep learning, and data analytics. That's something that we are embarking across all our product line. And we can use that to our own advantage and massive parallelism. And then also, we are a new player, we are not the incumbent. So we can do it cloud native. So that we can embrace the cloud even more effectively. And then third party, now if you look at the background, Anirudh, my President; and myself, and the system simulation is our graduate school work. And so something that I'm doing that for nuclear engineering. So -- and then Anirudh do that from IBM on the system side. So something that we are very comfortable, and it's not boring to us. And now we can really apply and then the -- what we learn in the school and then apply that more. Sometimes being new, they have a more innovative mind rather than stuck to the same formula. You have a new way to do it better. And that's kind of -- I think our team is very excited. And then so we're going to do both organically develop, we put the team together. And then under Tom Beckley, our leader in this area. And then the other part is also continue to do acquisitions. And then so that we can tack in and then also integrate make it faster, just like what we did with the IP business.

What about go-to-market? How is the go-to-market motion? Any different? Or do you have any kind of different strategy in the go-to-market motion for the system analysis?

Yes. It's a very good question. Clearly, our EDA's channel and the sales is not good enough for the system analysis. So -- now system analysis tend to be more into the channel, the distributors. And then we have some from the PCB business. And then we're going to expand that. We're going to review that. And we also put together a dedicated small team just focused on system analysis. So in some way, I learned from this -- when you have a small startup like IP, you cannot use an existing measurement and existing team to do it [indiscernible]. And so we just have to have a dedicated small team to augment what we have. And we're going to be open-minded and then learn from the incumbents and then also learn from the new way of doing things. And so we have this culture of adopting a learning culture at Cadence. So we are very fast in learning, and we are humbled to learn it, and then we're going to apply that and then try a different way, how is the best way to really address the market. So we bring in all the experts to help us to drive that.

All right. We have just a few minutes left. So I want to make sure just to remind the audience, if you do have any questions, just submit them online. So a question around the recent refreshes, the hardware, both Palladium and Protium. So they sound like a significant upgrade it's -- these refreshes happen every, what, I don't know, 18 months, 24 months maybe. So I guess, on a scale of 0 to 10, how important is this refresh relative to maybe some of that you've had in the past?

Yes, it's a very good question. This is part of our verification for flow. Hardwares play a very important role because customers are facing mounting challenges in terms of the complexity and the system verifications and software bring up. And so I think this is very timely. And in fact, we just finished the Q1, is a standout quarter. The best quarter we ever have for the hardware emulations and also our Protium-prototyping platform. So this new product we launched called Dynamic Dual and Palladium Z2 and also Protium X2 is basically provides 2x more capacity, and that's very important to customers. And then secondly, we have 1.5x higher performance when compared to the previous versions. And in fact, the previous version, surprisingly, after 5 years, still selling like hot cake. We have the first Q1 record quarter. That tells you how good is the design. And then the Z2, we have, again, new custom ASIC and then the beautiful parties. These 2 prototyping for regressions -- software regression and prototyping verification to the massive Palladium before the production for verification. They're using the same common compiler. So from customer, the attach rate is very high because they want to do it best-performing in the prototyping development. They don't have to change and modify and then adjusting to the new hardware platform. They just automatically go to the hardware platform. That is very compelling for them. So I think all in all, we already launched a product. We're already taking orders. The momentum is very strong. We're very excited. I'm really proud of my team able to do it very effectively. And I think -- stay tuned, I think, there will be great momentum for us.

Can we touch on -- you just mentioned that there's a custom ASIC in your hardware and that your competitor Synopsys has commercial FPGAs. Just -- that's just frankly 2 different ways of skinning a cat. Why do you think that a custom chip makes more sense for customers?

Yes. I think that custom -- we have internal debate. We believe that when you have confidence about the products and then you can sell a lot. And then the custom ASIC is the way to go. And because it can really optimize. It's more complex to do that rather than FPGA and then also much closer. If you do the verification, you will do the design of the chip. And the hardware emulation is the most accurate in terms of do the verifications. FPGA is less accurate. So in some way, I think, clearly, we continue to want to do this custom silicon, and it's getting very complex, and it's a very big chip, and then it takes time to really build it. And then we put all the pieces together to make it successful. And stay tuned, we're going to have a series. This Z2 is a great success coming up. And then stay on, we're going to do Z3 and Z4, and then we already start developing.

Great. So if we think about the design verification suite, PCB and packaging, flowing through all the way to system analysis. But the area we really haven't touched on yet is that PCB and packaging portion here. How important is that lynchpin to be able to connect winning design flow and core EDA and also making sure that these new investments in simulation are indeed successful?

Yes. I think this is very good questions. If you can -- as you're all familiar, Moore's Law is slowing down, and that a lot of more different design, and packaging become critical. And you heard about chiplets, advanced packaging and 3D packaging and that becomes more and more critical. And then you can see that from TSMC, even the Intel IDM 2.0 had a highlight about the packaging requirement. So I think packaging is -- we are uniquely positioned because of the complexity of the package design, the packaging, the 2.5D, 3D packaging become critical. And then we have it, our 2 have been doing very well. And then how to mix the heterogene packaging like chiplets? And that's another skill set to do that. And then there's a lot of advanced packaging that we're still learning, and we are still investing. This will be the area that we have to focus on. We just announced the Sigrity X, and this is the next-generation signal and power integrity for PCB packaging, IC packaging. And then we have great support and endorsement from Samsung, MediaTek and Renesas and more coming and able to show 10x performance and capacity gains. So I think -- all in all, I think the packaging becomes more and more critical. And then that's something that we are very well positioned to capture as some of the design phase packaging become more and more important.

Okay. Great. Well, Lip-Bu, if John were here, I'd probably ask him about guidance, but that's all right. I'll save you from that. I really appreciate you taking the time in doing this. It was great. And we'll talk to you soon. Thanks again.

Thank you so much.