Let me go ahead and get started. I'd like to welcome everyone this afternoon. It's my distinct pleasure to welcome to the Fireside Chat, the management team of KLA Corporation. To my immediate left is Rick Wallace, who's the CEO and President of KLA. He's joined on stage with Bren Higgins, who's Executive VP and Chief Financial Officer, also Head of Global Operations. Didn't have that on Bloomberg, or I would've put that in my notes. Yeah. Of course, in the audience in attendance is also Kevin Kessel, who is VP, and heads up Investor Relations at KLA. We've got about a 30-minute Fireside Chat in this session. If you have any questions, there is a mic in the middle of the room. You do have to go to the mic. We're not passing around this year out of COVID protocol. Don't be shy, and I'll try to reach out to you. With that, Rick, Bren, really appreciate you guys joining us this year. It's great to be back in person after a one-year hiatus on the virtual side of things. Rick, my first question is always kind of an open-ended sort of question to give you the opportunity to talk a little bit about the core IP, but more importantly, the value proposition you think that KLA represents for potential owners in the room. Okay, great. That's the question? That is the question. Okay. Core IP, value proposition. Basically, if you look over the last several years, the entire industry has done extremely well. Right. You've done better than the overall industry. I guess the short question is, what do you have in store for act two? Yeah, great. Well, John, thanks for having us. It's great to be in person and to see everybody. It's remarkably refreshing to have the quick interactions, right? I think that the last couple of years, since our Analyst Day, so for those of you who don't know, we hosted an Analyst Day in 2019, in September of 2019. At that time, what we talked about were some of the drivers that we thought were going to continue to drive the semiconductor industry, and we identified some of the obvious ones at that time. We didn't identify all of them as it came to pass, but we identified 5G, high-performance computing, automotive, and of course, we're looking at some of the trends at that time in virtualization, although it wasn't quite what happened. I think through those cycles, since that time, those drivers have actually been stronger with one hiccup in automotive. Right? Our thesis was, as the semiconductor industry continues to expand, the part that KLA plays will become increasingly important, and we'll be able to outgrow the industry through a combination of increasing our footprint, driving our service business, driving the new acquisition that we did at that time, Orbotech. Also what we do, inspecting and measuring of wafers and semiconductor process control, was going to be more and more important as the design rules started scaling resumed. What we didn't count on, we talked about the digitalization of everything at that time. What we didn't count on is the 10-year acceleration that we got in that through COVID. I think what happened is, with the exception of automotive slowing down and then now starting very quickly, we're in a very different period of the industry where semiconductors, ubiquitous, multiple industries, and high growth rates. KLA is very well positioned to support that with the technology that we have, the operating system that we identified at that time, and the fact that we're involved in every major technology innovation that's happening in our customers. The resumption of scaling is maybe one of the biggest drivers for success, and that is, for many years, the industry had really slowed in terms of growth in foundry and logic because the same lithography technology was being used. When EUV started happening, which was about that time was the beginning, that drove more opportunity, and KLA, being very well positioned at that time, very profitable, generating a lot of returns, returning that to our shareholders through buybacks and dividends, we also felt we were going to be in a position to outgrow the industry. The thing we got wrong is the industry growth, right? Nobody had anticipated this kind of growth, and I'm quite sure if we had said it at that time, no one would have believed us in terms of the way it's grown. We continued to execute quite well. We went through the COVID. As challenging as it was, I think we demonstrated our resiliency as a company by executing. Our market share continues to grow, and that's largely based on the investments that we made and the historical view of what we've done. Our footprint outside, of course, semi and into packaging also is proving to be important, and we think that's another leg of growth. We think we're in early stages of the pre-COVID 2019 Investor Day. We'd still identified a number of things we thought would happen by 2023, and we still think those things are on track at a much higher WFE level, which I think is supported by the end markets. Our operating model, our gross margins are all best in class in terms of the overall industry, and they've scaled as we expected in the last couple of years. 2022, we don't have a lot of visibility, but I have to say, this is the most visibility I think we can remember having. 2022 is going to be another very strong year. We don't know exactly. We'll talk more about that when we finish this year. I think we're well positioned to continue to grow and continue to drive operating value, operating model, and then drive returns back to our investors. Well, Rick, to hit on the point on 2022, I respect the fact that you don't think you have a lot of visibility, but you guys were more vocal than a lot of your peers talking about the shape of next year, almost to the extent that it sounded like there are real KLA specific drivers that are really helping the business above and beyond sort of the rising tide of WFE. Help me understand why you guys seem to be a little bit more sanguine than some of your peers discussing first half, and your ability to grow first half over second half. Well, it helps to have, what, $5.5 billion backlog. Right. Or more than that. Clearly, the backlog position and the sales funnel visibility is something that's given us confidence around the first half. What we said, for those who didn't hear it at earnings, was that we saw the first half growing in the high single digits versus the second half of 2021. That's a total company statement, but obviously, that's being fueled by the WFE part of the business. What has happened this year is absolutely that the WFE number's been a supply number more than a demand number. Demand has been very strong. Everybody's well aware of the different supply chain constraints that we're all dealing with to try to ramp up to support our customers. We are adding capacity against that backdrop, capacity being people, infrastructure, and in our supply chain. We feel pretty comfortable as we look at the first half of the year relative to that. Based on the input we're getting from customers, based on the funnel, I don't see it slowing down anytime soon. We're not going to put a point position out there in terms of what 2022 looks like until we finish this year, but it looks like another year of very solid growth for the industry. A lot of big projects coming online, right? In addition to the build-out, there are many that are going to come on. Even in the second half of 2022, there are some that are slated to. We don't have that visibility, but the drivers are definitely there. Well, Rick, I guess the point I was getting at is, if you look bottoms up in your product portfolio, are there things in the NAND market that you're becoming more exposed to? Are there things at a large microprocessor company, as they try to revamp their IDM strategy? Right that you're becoming more exposed to, that is sort of an incremental kicker to just WFE being strong next year? Right. Start with inspection, because I think that's an area that many people are familiar with KLA, and we'll talk about optical, laser scanning, and what we call BBP- Also, e-beam. What we've seen in the last couple of years, the last time we were in person and talked about it at the analyst day, is we felt that EUV was an inflection in driving increased need for our equipment, because some of the unique problems around EUV. Also, it's a double factor because scaling has resumed, and now on top of that, you have kind of a renewed set of competition around the leading edge. You could argue that before there wasn't much competition. There was a leader, and there weren't as many designs. In 2019, we talked about the proliferation in the number of designs of advanced nodes. That's only increased. What we've seen is we have very broad design, we have a very strong indicator that the next design node is going to be a big one, and that's going to drive a lot, and we're seeing increased regionalization, which is another factor that's driving more demand. All those factors kind of bubble up to our challenge right now, I think our biggest challenge as a company, say in the optical, which has had the best year by far ever, is going to be continuing to We've done a good job of supply chain- It's going to be meeting the demand that has continued throughout this year to go up. Metrology also is seeing a lot of drivers. Packaging and the whole EPC business has been good, but it's kind of been what we forecast in 2019. It hasn't been as influenced by these drivers that we're seeing in WFE. It has a little bit, but it's still on plan. I think we're in great shape from all those drivers, and the competitiveness of our products has never really been stronger. We're seeing that in the way that we're winning at designs and we're winning in advanced nodes, and if we look at the number of places our tools are being utilized, we feel very good about 5, and eventually it'll be 3. I think we'll be well-positioned for that, too. Well, Rick, I know this has been a point of frustration for you, because there's this view with investors that the existential thread is you guys dominate the space that you're in. You did that in an optical world, where we're now moving to more exotic sort of ways of trying to find defects. The risk is only to the downside relative to your share position. Can you talk a little bit, you and I have had this conversation, about the value of incumbency, the value of the software stack that you've been able to develop through just sort of years and years of running more wafers on inspection than any other company? Yes, John, I thought you might ask that question. I have some thoughts for you. One, if you go back in time, and I think about the history of KLA, one of the first things we did to grow the company was figure out how to go from being an analytical set of tools used in engineering development to being in something that scaled with production. The whole inline monitor movement was phase one of our growth. The next thing that happened was we merged with Tencor and became a portfolio company. The great thing about that for me, because I was a product manager trying to sell one product, is suddenly I could sell a portfolio and not tell the customer what the answer was, but present them alternatives. The other factor is thinking about when we scaled the company and when I took over as CEO and the rest of the management team, we went from being kind of a founder-led to next-generation-led, and we changed the way we managed the business. We looked at every business and we thought, "We've got to drive profitability in every one of our businesses. They have to have gross margin performance." We looked at that as the measure of success in a product. What is your gross margin? Are we investing enough in R&D, and what's the return? It's a long way of saying we've evolved into this operating model, which invests heavily. We out-invest all our competitors. We feel like we're very focused on solving the customer problems. Because I thought this would come up, I took a look at the 5 nm process flow. 5 nm process flow has about 1,000 steps, depending on how you define steps. The current plans for inspection and measurement. Only a couple hundred of those steps will be monitored. 65 roughly is the average of number of steps that'll be done with BBP, and that'll be split between Gen 4 and Gen 5. 190 of those steps out of the 1,000 will be with laser scanning, and one to two will be with e-beam. The reason for that is because what customers are balancing, they're trading off is what can they see and how much of the wafer can they sample. The challenge with technology is it's always that trade-off, which is why laser scanning has the most, because frankly, it has the most coverage. It's used on the layers that are considered the least critical. Now, 10 years ago, those would've been by far the most critical. We keep investing in every one of these technologies, but the ratio has actually stayed remarkably the same. The thing e-beam is used for is that old engineering work and now more in metrology, which is why we've invested pretty heavily in e-beam. I actually think the biggest production use of e-beam technology, there's a review which has scaled with inspection, but the biggest inspection opportunity is going to be in reticle, and that's the product KLA has. Yep. The reason we did that is we think it's the best cost-effective solution for the first few years of EUV for reticle. We will need to upgrade that for High-NA EUV, but that's 2024, 2025-ish. For now, we think it's because of the area that you have to inspect and the speed. You're allowed a lot more time in a mask shop than you are in a wafer fab. That's been kind of the split. Rather than be frustrated, I thought I'd share some of that data. The chance that that number of layers on e-beam is going to dramatically change, it could go one to two to three to four, but it's not going to go to 50, and it's not going to go to 100. Those percentages are why that ratio remains the same. If you look back 10 years, those were the same ratios. The growth for KLA would be if we could expand the 65 layers and the 190 layers, so those 250 layers to 300 layers. That'd be a lot of growth. Our customers are constantly battling the economics. The same thing that keeps e-beam at a small number keeps those numbers gapped and managed by our customers. Does that make sense? No, it makes perfect sense to me. Again, my sort of quick back of the envelope math is you've got margins that are 10 points higher than any one of your peers. You must be doing something that's valuable for your customer, and you must be the only ones that are able to do it. Well, the portfolio helps a lot. Back to this point of because we don't tell them that they should use. Like I said, years ago, I was there trying to tell them they should use our system for everything. I think that there's this balancing act, and I think we are constantly evaluating the technology that we have, how valuable is it for customers? Do we win the early battles? We do share by share, layer by layer, and that's really, this business is an in the trenches business. We focus on that. It's partly why Bren supports the idea that we have 700 applications engineer in the field. Because we're helping customers use that capability so that they can manage these ramps. The good thing is, I think 5 nm, it's heavily served by optical. Years ago, people might've said that was going to be e-beam. We already have the data. It's not. 3 nm, we'll keep making improvements to the five. Underneath that question is the importance of algorithms. Right? Early on, KLA was a heavily software algorithm company, and I got told by our auditors when I first became CEO, "Don't talk so much about software because you'll get reclassed." Now it's okay. I'm told it's okay. Most of our engineers are software algorithms engineers. In a sense, what we sell to customers in process control, what we really sell is information. Yeah. That's what we sell. We sell information, and we have optical, and we have mechanical systems to gather that. The real value, I used to joke with customers, there's no value ultimately in finding a defect, only if you could do something about it. There's definitely not any value if you can't find the defect. Right? The algorithms allow us to extend these platforms. One other thing that's interesting, Gen 4, the prior generation of optical inspection, will outsell Gen 5 this year. It's 12 years old. Yep. It'll outsell Gen 5 this year. The demise of optical is nowhere to be seen, and we keep making these algorithmic improvements to not only Gen 5, but they port to the work we're doing on Gen 4 and to the laser scanning. Every one of these tools becomes more relevant, and the more we can squeeze out of them, which is what AI and ML does, and why we have a centralized group that's been working on that for years, is we just create more value for our customers, and they can use those tools longer. Well, Bren, I might have misheard you at the last analyst day, but I think in a sidebar conversation, we talked about the fact that CPUs and GPUs are not an insignificant portion of your bill of materials in a lot of your tools. Right. Which goes right to the point of kind of the algorithms and the software aspects. It allows us, to Rick's point, to extend the platform over long periods of time. If you think about if you have to go develop a new platform, that's a $200 million effort in R&D in multiple years. If you can use these capabilities to extend the life of those platforms, that the incremental R&D is more modest, the rate of returns on the product development is great allows us to drive the model that we have, also redeploy some of those investment dollars to other parts of the portfolio. When Rick brought up supply earlier, it's been an issue that's plagued everybody this year. You guys have done a great job. I'm just kind of curious, relative to one of your peers who recently reported, it seems like supply issues got worse in the quarter. Any update that you can give to the investor community about kind of how you're thinking about the supply side of your equation right now? Well, our business is different, right? Yeah. We're a lower volume business. It's a higher mix business. There are some unique attributes to that mix in terms of some of the subsystems we have and the lead time of those tools, or those parts. In those situations, we have unique suppliers. We have suppliers that generally work with us, so we don't compete against our competitors for that supply. The inherent or intrinsic lead time is pretty long. We have these capacity agreements in place over time that the trigger points are very clear about how much capacity do we get when. We plan our ability to deliver based off of those systems, and we do a lot to try to pull that in terms of long lead time material that we buy from our suppliers or try to pull in some of those components. In general, that sort of sets the upper limits of what we're able to build, and then we go work all the issues in between. Some of those issues are the more fungible part issues or some of the semi issues that have been well chronicled out there. Usually, we can manage that within the time fence that's established by the long lead time materials. We do place long-term commitments with our suppliers. We don't cancel them. We carry more inventory. We made a decision long ago that to optimize the business that we have, to be able to have the differentiation we have, we need to have key custom hardware that drives the service business as well. We're going to carry the inventory, we're going to manage those suppliers in a way that we have very tight relationships. We have to nurture those relationships, and we optimize around delivering that capabilities, and we accept higher levels of inventory, making capital commitments, things like that to ensure we have supply. I think there's some unique dynamics for us in terms of how we're managing through it. We still have challenges every day. I wake up with a new problem, and we're escalating at very high levels, to try to get attention. Thankfully, for our business, some of the demand is a pretty small%, so we can work with our suppliers through that. I think in general, look, if I could get more parts, I could absolutely ship more. I think we're managing through it pretty well so far. Haven't seen any real changes in all of that. It's been fairly consistent. It's challenging. It's been challenging for the last six months or so, expect that it will continue to be. We're making investments to ensure that we can continue to add capacity through next year to support the growth profile that I mentioned. John, going back to early 2020, the pandemic, we were in lockdown. If you recall, a number of companies suspended guidance. We did not. Yep. This was not a surprise that we were going to be able to navigate it. At that time, we felt confident that we had enough visibility, even though suddenly we couldn't travel, we couldn't do those things. We did widen our guidance, but we didn't suspend it, and we hit it, and we exceeded it, and we've done that through. Actually, none of that is a surprise to us. We were a little bit more surprised with some of the challenges that others face, but I think it is, there are differences in these businesses. I think that we'll continue to manage, and as Bren said, we're not done. I mean, the shortages that are out there, hats off to our procurement people, because there's a lot of work going on every day to manage it. We're also good customers, and I think that matters a lot. Yeah. No, that makes a lot of sense. Rick, I know it's a little uncomfortable to talk about customer-specific questions, but you're in an industry where there's not that many customers. The one that I wanted to explore, either directly or somehow indirectly, you look at Intel's IDM 2.0, and correct me if I'm wrong, historically, they have not been as prolific a user of inspection tools as some of your other customers. As they move to this IDM 2.0, they try to become a foundry. They try to do five nodes in four years. All of that generically sounds relatively good for someone in your seat. Am I thinking about that the right way or the wrong way? Again, not to talk about a specific customer, but in general, if you're trying to accelerate your technology and you're trying to catch a leader who's ahead of you, if you don't invest in many things, but process control being one of them, you're just not going to have the visibility into making the improvements and iterating on the process. Intel has spoken publicly about their need to increase their level of investment overall. They obviously didn't stand out with process control, but we report what areas we have logic and foundry and memory, so it's not hard to understand that dynamic. I think that for every company we've seen that wants to accelerate advanced technology node, it is a benefit to KLA. We saw it happen when Samsung ramped up their efforts in foundry. Obviously, TSMC has been leading the pack, we are all for multiple players at the leading edge. That would be a great thing for us because not only would it drive that one segment, but it would continue the overall evolution and the continuation. I think with EUV, we're back to scaling, and it's kind of, we'll see who could do what. I suspect that the history of Moore's Law, 18 months for doubling the number of transistors, that went away a long time ago. We're no longer at a no progress, right, which was kind of where we were. We're at a, for KLA, kind of the Goldilocks perfect rate of innovation at the leading edge on Moore's Law, because it's long enough, there's still progress, we still need new capability, but there's huge amortization of all the engineering that's going on. I think actually for our customers, you look at the profitability stack of the whole industry, it's not like one part is subsidizing the other. Everybody's making money right now, and partly because that's a more manageable, I think, cadence for innovation. Rick, you've gone out of your year this way to kind of remind us that you're outgrowing WFE, which is a very impressive thing to think about given that a lot of WFE this year is coming from some more trailing edge type technologies. Right In foundry logic. Before we start talking about Orbotech and the acquisition and how that might get you more leverage to some of that more mission-critical, not bleeding edge stuff, help me understand within the core sort of inspection business, how this sort of mission-critical growth in CapEx, you're levered to it, or is it just that you're really levered to that bleeding edge stuff? Obviously, bleeding edge is what drive most CapEx anyway. Yeah. I think that's pretty clear. To your point, a couple things have happened. One is, if you're in the trailing edge right now, you're short of capacity. Yield is like found money. There have been a lot of improvements, upgrades to just managing yield to squeeze more out, because it's really hard to add capacity quickly if you're in an existing fab. It's very hard to bring it in at all. You have that dynamic. The other one you have, not insignificant, we talked about Analyst Day, is automotive. Yep. Automotive is driving a higher level of reliability. Some of the Gen 4 business that we're selling is actually a variant, it's not a great word, I guess. Another model of Gen being used for automotive. That has a long upgrade path. We've seen an upgrade in terms of process control in those customers. Frankly, we've gone and sold more to them, is the other thing, is we've made more effort to sell. I think a lot of it is that, and there is sustainability in there because once they start recognizing the economic return on more yield on a fully ramped and appreciated fab, it's good business for them and it's good for us. John, I think there's two other things, right? The first is that we've got a number of businesses in the company that scale like process businesses do. As customers are adding capacity, film metrology being one area, unpatterned inspection being another, reticle inspection, where these are scaling with wafer starts, those businesses have seen faster than market growth. The biggest thing, though, is that because of what we were talking about earlier with the introduction of EUV enabling scaling, enabling a lot stronger end market diversification, particularly at the leading edge, is that those design starts are consuming a lot of that capacity. I can't remember the last time I've heard our customers talk about adding capacity at multiple nodes. Yes, most of it's at the leading edge. What that does show is that they're not taking down capacity. Capacity, right at the previous node. The purchases for the next node, they're not trying to reuse any of that capacity. Harder to do in a more complex environment with the introduction of EUV, they're not able to do that as much, that's been a driver for our business as well. Rick, I'd argue one of the high-class problems you've had is M&A. You've had a high bar for M&A because you've got such a great core business. Orbotech was a very different M&A than the company has historically done. As you mentioned earlier, KLA merged with Tencor. That was sort of very complementary markets. Orbotech's a little bit more of an adjacent sort of add-on. What did you see in that asset that attracted you to it? I guess more importantly, since you've had it, how has it performed? Right. Great question. We start with are they exposed to markets that we think have fundamental growth? It's very hard to do anything if you're not experiencing growth. They were. It was the front end of 5G, front end of automotive, and we thought that all the work on packaging was going to continue. In fact, we'd been in the packaging space, we bought ICOS years ago as a hedge on More than Moore. We actually had a footprint there, and this was more of that. The other thing they had was leading technologies in the businesses, market share positions, and enough synergy in technology that we thought we could add more to them. The thing that surprised us, I think, was that the upside of the channel has been greater than we thought. Because we thought these are all different customers, but packaging is actually evolving to some of the same customers. Those customers, they might have admired Orbotech, but they weren't going to make big bets on them. Where they'll have KLA at the table because they know that if we say something, we're going to be held accountable for it, and they're always trying to work through that. Actually, there's more upside to come because of that. The other thing is the operating model and the profitability of Orbotech was greater than when KLA and Tencor merged. Even though it felt like KLA is a much more profitable business. it wasn't always. We think a lot of the work that we can do, that we did already in KLA, we can do with Orbotech, and we've already, a couple of years in, seen some of that happen. We think the profitability. They've got great talent, I think they've got exposure to good markets, and there's new technology that we're bringing. I would say that it has been successful from that standpoint. It's been frustrating to not be able to travel and visit them, but I think that they might be okay with that. I do think we've made a lot of progress and it also sets the basis for the potential for some bolt-ons there eventually, not right away, but I think there are other opportunities for continued growth. always, always, it has to beat buying back our own stock, just for all the investors. That's the metric we use, is what's the better use of capital? Is it buying back our own stock or buying an adjacency, and we're only going to buy it if it's better than buying back our own stock. Our EPC business, we talk a lot about the front end semi EPC business and it growing mid-40s against a market backdrop of 40% plus or minus this year. The EPC systems business is growing in the mid-20s. Yeah. We are seeing nice growth out of that business and, to Rick's point about driving a stronger operating performance, we are growing the operating margins or the incremental operating margin greater than the company targets of 40%-50%. Right. It's I think the thesis around what we thought, and that's despite a flat panel market that hasn't been great. Yep. I think that the thesis we had is very intact. I think the technology's increasing and becoming more complex over time, and that plays well to KLA's strengths. I was sitting in a product development review recently, and I was listening to our CTO talk about designing optics for a next generation product in the Orbotech business. Leveraging our own internal capability to design the necessary system to be able to deal with some of the density requirements. Those are the kinds of things we thought we'd get out of this, and so far so good. Synergy's probably double, about. We were close to $100 million. Once we're done with all the facilities and full systems integration in terms of our target, and our plan was $50 million. I think we're making the kind of progress we thought. Great. With that, I think we've come to the end of this session. I want to thank everybody in the room, and especially for Rick and Bren for spending time with us this afternoon. Great conversation. Thank you. Thanks, John. Thanks for having us.