Great. Good afternoon, everybody. My name is John Vinh. I cover semis here at KeyBanc Capital Markets. We're pleased to have Lattice Semiconductor here with us. We have Lorenzo Flores, CFO. Welcome, Lorenzo. Great to see you. Thanks, John. Great to have you. Always a great event. Let's start talking about your data center growth. In the quarter you just reported, comms and compute grew over 80%, which would obviously suggest that AI data center is growing faster than that. Can you talk about how much of this growth is being driven by general purpose server versus AI? Yeah. I'm going to start with my best understanding, but I think I'll provide a little bit more color around the overall growth dynamic, which might actually be more helpful. At the beginning of the year, I think everybody in the industry was looking at industry expectations from about 15 million units for total servers. There was something about a 20% AI mix, and then it grew up to 16. Now, the latest past three weeks, we've seen the estimate for overall servers about 20 million units, and with a higher AI mix to them. I would say that in parallel, as we've gone through the year, we've continued to see our demand going up and up and up. At first we were a little bit surprised and went back and tried to understand it. What's really going on is sure, AI servers are growing. But also just the more standard foundational CPU-based server infrastructure that is enabling this agentic workload being broadly deployed. At the highest level, both phenomenon, growth in the AI servers and in the more general purpose servers, it's very beneficial to us. We've historically been really strong and had a pervasive footprint in standard servers. As you know, and we've talked about many times, we have a really great opportunity in AI servers as you build out the racks with many more attach points for us, up to 100 attach points or even greater than that in an AI server rack because of the complexity. When we looked at what's going on and honestly in real time, it's a challenge to parse those different drivers because our ship timing and the end customer ship timing don't necessarily align. What we saw is a really strong unit volume growth in our products going to servers. That's driven by our attach rate is growing. As we've talked about, historically, let's say five, four years ago, it was 1.x, then it went to 2.x. Now it's 3.x, and it's now even higher. We've been providing higher value FPGAs into those opportunities. We've gone from our MachXO2, MachXO3, MachXO3D, MachXO5, and some Nexus parts as we are now seeing an expansion in our ASP from three to four and so on. You have this multiplier effect for us, greater unit volume growth driven by higher attach rates as the complexity of the servers in both AI and general purpose applications grow and increasing ASP. So we've seen great growth. It's setting the foundation for future growth. Great. I know I didn't only answer your question, but I thought you'd get there anyway. No, that's super helpful. Do you have a sense of what your mix of general purpose versus AI is right now? What we've been looking at is our overall AI penetration, and some are on the data path, some are in the server domain. We don't do very much direct AI compute. That's grown into the high 20% of our overall business, which means it's a greater percent of the compute business. But again, we know where our design wins are. We know customers we're shipping to how they're actually using the devices not yet fully transparent to us. But in a way, it's like we'd like to know. More importantly, we want to ship all the units to them and drive the growth. Great. Maybe just to double-click on just your attach rates. I think when you're at 1x - 2x, a lot of your core functionality, my understanding, was driven by root of trust and maybe I/O expansion as well. As now you've gone to 3x, and now you're saying it's even higher than 3x, what are some of the functions that are driving the attach rates to such high levels right now? Yeah Beyond just root of trust? Great. I think actually back when it was a lower attach rate, it was primarily bridging an I/O expansion type of applications. What we see then, we saw the security capabilities, including root of trust layer on. Then now we're beginning to see an expansion into manageability applications like power and cooling. Then the basis for our acquisition of AMI builds on that growing functionality that we've seen in the FPGA business and the broader application set that's available to us and their capabilities, which really augment us. We used to say we're first on, last off in a machine, and they're first on, last off right with us. Then they've layered on top of their boot capability, this management capability, and it meshes really, really well with how we see our FPGA application set growing. Right. Yeah, that's how the attach rate grows, and then that's the growth path that AMI had intrinsically. Then for both of us together, that's this one plus one equals three phenomenon we're trying to drive with bringing more sophisticated solutions to the customers. Right now, we think it's very acutely needed in the data center. Great. I do want to touch on AMI in a little bit, but just a couple more follow-ups on data center. Your levers on data center is obviously market TAM, ASPs and attach. If you think about your data center business, I don't know, three to five years out. What's going to be a bigger driver of growth? Is it going to be ASPs or attach rates, do you think? Right. I will answer that in three parts. One is the FPGA part, one is AMI part, and I know you said you are going to get to that, but then I think the solutions together. For the FPGA part, if we were thinking only about silicon devices, unit volume growth will be significant and will continue to be the primary driver, and we will be able to have some expansion. What is going to be driving that is the number of attach points in the server for different functionalities. We had already been working on things, and we have talked in the past before, and I am sure some of you have heard this, things like leak detection, which 18 months ago nobody thought about, 12 months ago, people were just kind of wondering about it, now is a really critical application. So an opportunity did not exist before. We are really well positioned to drive that. The same thing with some of the more advanced management capabilities that companies run. So we would see continued attach rate growth on top of any industry unit volume growth. Right. Our ASPs will continue to progress. But I think the unit volume will be the bigger driver in the FPGA business. In the AMI business, their footprint will also expand, but also as their management capability is growing and is, t hey have the opportunity to again expand the value. So that might be more of a driver than footprint. But there is opportunities they have as they integrate open source capabilities with their inherent capabilities to also grow unit volume. So there is that. The third piece of it, and it is very important because it is core to the strategic value of the AMI acquisition, which is the solutions that we will integrate together. That is a unit volume play, but also a really high-value application we can bring to our end customer. So you would see much stronger revenue growth per unit from the solutions that we provide than from an FPGA or a manageability firmware kit by itself. Great. Just a follow-up on that. Is that greenfield or is that just a combination of the first two? Totally greenfield. Well, there's a nuance in the answer to that question, which is the problem statement for the hyperscalers is the same. How they would solve it by using say, an FPGA and some of AMI's firmware could be the same. But integrating it and adding more features and functionality, that's greenfield. It's an upscale in value that we think we can capture. So yeah. So the solution may look like the same, but I think the overall value would be much greater than the individual components. Yeah. It's helpful. Thank you. I got one more question for you on data center is when you think about rack scale, I think the teardowns have shown that you've got an outsized number of FPGAs in a rack scale rack. I guess the question is, how are you thinking about this going forward? Is this content sustainable? At Xilinx, I know this is a very different company. There was this phenomenon, base stations, where there was kind of an ASIC reversion. I know that you're more insulated from that for a number of reasons, but do you think it's sustainable? Is there an opportunity for you to grow that content with newer generations of racks? Right. I agree with your premise. It is different than. For those of you who haven't followed FPGAs for forever, the history has been FPGAs get to a certain volume and the customer says, "Okay, I'm going to design you out and put an ASIC in place." I'll tell you, even in my prior company, where that did not happen was where the pace of change was such that you really couldn't afford to do that. In our situation, given that we've deliberately taken this companionship strategy to the market, where we are not going head-to-head for core processing capabilities in the data center. But we are surrounding that with capability that makes real economic sense to offload. You don't want to spend two nanometer silicon, your silicon die budget on I/O expansion. Makes way more sense to do it in a 65 nanometer FPGA. By doing that, we in a way insulated ourselves, just as you said, from that phenomenon. Because you've also seen an acceleration in the design cycles for these racks, from two years to 18 months to 12 months, the need for flexibility is an inherent advantage for maintaining the FPGA footprint. Then you layer on top of it, there's the performance characteristics of our particular product, which is very low power. If you add hundreds of devices to a rack and they're watts in consumption per that's a problem. If you have milliwatts, it's actually pretty helpful to the overall solution, the power budget for the rack. So that helps us as well. The last piece of the sustainability does again come back to being able to integrate functionality that we can bring from AMI into the overall solution and put it in programmable. Then you do the updates and upgrades, taking advantage of the capability of the programmable FPGA. You're in really pretty good shape to provide a flexible solution for the customer. For longer periods of time. Excuse me. Okay. You always bring me to altitude. Yeah. We like altitude here. Yeah. It's a thing. Hey, so maybe pivoting to AMI, I think you talked about that you've already engaged with 100 different types of customers, hyperscalers, OEMs, ODMs. Other ecosystem partners, and it seems like feedback's pretty positive on the acquisition. Maybe what customer problems are generating the most amount of interest in the combined Lattice AMI solution- Yeah from these customers? It will be a very similar problem set that the hyperscalers and data center operators would be complaining about to everyone, which is how do I increase my uptime? That will refer back to inherent capabilities of manageability that we have where you can look at what's going on in your system and take action to do it. Security, a huge issue that we have to continuously advance our capabilities on and provide customers with solutions that are hardware and firmware-based, because threats keep expanding. Then there is power, cooling that you need to optimize how your rack consumes power, and can you keep it at the right temperature for optimal performance? If something happens, how do you prevent catastrophic failure? You talk about leak detection often. If you can respond instantaneously and shut off the source of the leak and have a managed power down, your system is in much better shape. Then back to the uptime. If you can bring up your rack all simultaneously versus a serial blade-by-blade boot, you are up and running much faster. So these are all capabilities we have been talking to the customers about in different forms. Then, finally in the end, because of the demand as such, the hyperscalers are looking at trying to get their systems built, and they are going to a mode of wanting to be able to use a heterogeneous source of supply. So different ODMs, different componentry, mixing old infrastructure with new infrastructure. How do you manage that all together effectively? You need the manageability at the firmware layer to do that and that is the very strong value proposition from AMI. That is great. Physical AI seems to be a kind of an emerging theme for you, but also many of your peers here at the conference. You seem pretty excited about it. What are the opportunities within physical AI where you are the most excited, and can you give us a sense of when we could see meaningful revenues from this category? Right. I think there is a picture we show sometimes, and some people get it and some people do not. But one of the really, really cool things about the way we approach the market is we play to our strengths that you were seeing manifest in the data center right now in the same way in the physical AI world, which is our great strength. It is like, well, we can bring in a bunch of different information. As I said earlier about bridging and aggregate data from multiple sensors, pre-process it. Send it to the core processor, get some action back, and then drive a reaction through a physical system. That is the same thing in a leak detection as it would be in a robot doing something. So what we see is the ability to apply this companionship approach there. We're very strong in sensor fusion and aggregating different forms of input, and particular strength in vision. Multiple cameras coming into one stream with some pre-processing and that's part of the Holoscan platform from NVIDIA that we're on. Then going out and taking action. We think we have a footprint in almost all of the robotics companies, and we are expanding that footprint analogous to what we did in the data center with continued access to different problems they're trying to solve. I will say that a lot of these companies are not as advanced as the more old line, traditional industrial automation companies in terms of their use of FPGAs. As you know from a long time, FPGA designers are not all that common in the industry. We're doing things to enable a more software-like programming environment for these guys using FPGAs, and we're accelerating our momentum there. Then what you see then is okay, when does this volume take off? We're not sure. We'll probably start seeing it in the end of next year, but we're going to have multiple sockets per instance. We are going then to be able to capitalize similarly as we are on the data center AI trend. I would say that as a backdrop to that, our core industrial business which is mostly traditional factory automation stuff is actually doing well. We're starting to see that recover. I think you guys might have seen the PMI across the globe is actually doing pretty well, except maybe in Japan. But we're not exposed to auto that much anymore. We're still seeing year-on-year, let's call it 20% growth in industrial from last year, and we think that will grow again next year. Not sequential mind you, but if you just look at similar periods last year, we're seeing consistent growth. That's great. On the industrial front, and maybe just more broadly, can you talk about what your lead times are doing right now? Yeah. And talk about what your visibility is looking right now. It's probably getting much better than it was a couple quarters ago. It is. And you want me to get into the supply side too? Yeah. Because our visibility is increasing. We have a phenomenon where our book has continued to grow in total dollar value and then expand out in time. We see that as very positive because we think we're getting a realistic view of when customers really want their parts. We always get this double booking question. I think it's a valid concern. We keep watching out for it. We try to manage that. But we have better visibility because our book is expanding, and we have really tightened down the channel inventory so we can see directly what end customers are really after. Having done that, then we turn to the supply side. Earlier in this year as we started to see this increased demand, our lead times did increase, in fact, some cases it got over a year. But we've been working with the supply side and managing that back down to the in the 40s. What we're working through right now is the pinch point is assembly test or packaging. Our fab side is we're pretty comfortable with, because we use older nodes not as tight as some of the newer nodes. Then we've invested a lot in, you'd see this in our CapEx, in tester capacity, for the final test and ship out to the customer. Yeah, we're managing through the assembly test pinch point right now. We have people like in the factories are people managing this directly, and we're expanding our supply chain because the way we think about it is this is not temporary. We had about a $500 million a year business last year. We'll end the year approximately double that in the FPGA side, and we expect to continue to grow next year. That means our supply chain has to be 2x + more robust, and that's what we're driving to. And your visibility? Our visibility is actually quite good, in terms of both the demand side and the supply side. We've been able to do a much more rigorous mapping than was necessary in the past. We've had to. I think [Ford] said that you have visibility till the end of 2027 at this point. Yeah, absolutely. We've booked out. I've got to be very careful with my words because when I say booked out to meaning we have bookings out through Q4 of next year. We are not booked out in the sense that we can't support more demand. We can support more demand. Got it. Yeah. Okay. It's just continue to support this growth trajectory. Sounds good. Well, thanks so much, Lorenzo. Anything else? That's all our time. Yeah. We're good. Thanks for having me. Thanks. Appreciate it. Thanks, everybody.