KLA Corporation ($KLAC)

Good morning, and welcome to the second day of JPMorgan's 54th Annual Technology Media and Communications Conference. My name is Harlan Sur. I'm the semiconductor and semiconductor capital equipment analyst for the firm. Very pleased to have Bren Higgins, Executive Vice President and Chief Financial Officer of KLA here with us today. I've asked Bren to start us off with KLA's view of the wafer equipment, process control, advanced packaging, a clean environment this year and next year, KLA's growth outlook within that, and then we'll go ahead and kick off the Q&A. So Bren, thank you very much for joining us this morning.

Harlan, thank you for having me. Good to be here. So maybe I'll just level set a little bit from some of the themes coming out of earnings and our recent Investor Day that we had on March 12. Certainly, it feels like that was a long time ago given some of the momentum that we continue to experience in the industry, even though it was only several weeks ago. But certainly, 2026 -- and that's really across all segments. If you look at what's happening in the leading edge, there's nice broad-based participation at the leading edge. We're really encouraged by that. Process control intensity is higher there. And given the nature of what's happening with high-performance computing, it does drive some unique benefits, we think, to process control and the value of it. On the memory side, of course, we've seen the pricing environment, and I'm feeling a little bit of that as a headwind in our business, as it relates to my own computing requirements, but additional spending that's happening there and then the advanced packaging feels that it's accelerating. And if you go back to maybe January time frame, we thought the packaging market might have been somewhere in the sort of plus 20% overall. Now I think it's in excess of 30%. And in our business, we see that growing even faster in semi process control. We talked about an environment that gets us to $1 billion which translates into a high 50% kind of growth rate. So really all 3 segments. Service business underneath all that continues to move forward. And -- we have a portfolio approach to this market that I think favors us. I'm sure we'll talk a little bit about that where we're able to meet our customers' technical needs, but also their economic needs, which vary as they move through some maturity and learning cycles on a new node. So we're pretty excited about that. 7.5x our nearest competitor, about 360 basis points over the last month since 2021 or so. So I'm pretty encouraged by what we're seeing in the market, the momentum there in the products. At Investor Day, we outlined a 2030 target model of $26 billion in revenue for the company and $84 in earnings. That was $26 plus or minus $2.5 billion at the top line and earnings of $84, plus or minus $8. And a path to that with growing KLA wafer equipment intensity, a continuation of what we've seen over the last 5 to 7 years, which we're excited about. But if you look at the conditions of high-performance compute is it's a greater percentage of the total. We feel very good about those dynamics driving more relevancy and process control. Our service business, we also upticked that in terms of our long-term growth rate of 13% to 15%. So you end up with a 15-ish percent growth rate for the company against an industry backdrop of 12%. So we think we'll outperform the market baseline -- for our equipment business for systems business by 4% to 5% here over the next several years. And -- the financial model continues to hum and we outlined an environment of opportunity, both in terms of incremental gross margin improvement spending profile that's being driven by some dynamics in the industry that allows us to continue to drive scale across our business and also the leveraging of AI capabilities in terms of how we drive our overall productivity. So I'd like to say 1 of the best business models in the industry gets better on that trajectory between now and 2030. So maybe I'll just stop there and we'll go into your questions.

Absolutely. No. I appreciate the opening commentary. And why don't we continue from where you left off, which is the Analyst Day, 15% top line growth CAGR. We'll also talk about the strong share gains and WFE and process control growth outperformance in a minute. But at the Analyst Day, I think part of the -- what was clear at the Analyst Day was that part of the share gain and absolute market leadership is your R&D scale, an aggressive cadence of new product introductions, right? Between 2021 to 2025, you introduced more platforms and programs versus the same period 10 years ago. And average platform or program spend is 70% higher, right? This year, team is set to introduce platforms across all 13 of your product segments, that's the most in 10 years, right? So give us some examples of the platforms. The teams will be bringing to the market across process control across advanced packaging, and even software and analytics.

Yes. No, it's interesting. And 2 things are true. The first is we're spending a lot more on development, right? And if you look at our programs, we have -- we've said before, we have over 10 major programs over $100 million in the company. So it takes time to develop the capabilities that we bring to the market. But at the same time, we have a pace of innovation that's happening in the industry that's allowing us to leverage and optimize the existing platforms that we have. And historically, if you were on a much faster technology cadence, you'd have to spend more to keep up and you'd have to transition faster. But we have a pace in the industry, it's not exactly the 2-year historical Moore's Law cadence, but a pace of new innovation that drives the design environment, but also allows us to optimize the capabilities we have. Also, innovation isn't just scaling, scaling is a factor, smaller feature sizes, but also transistor architectures and other things like that, that are driving new capabilities that drive those designs. So we're able to drive, I think, scale on what we do, and we're really able to optimize the platforms we have. And so that's I think part of it, and I think it's pretty important part to our view of R&D scale, R&D as a percent of revenue will likely trends down, part of it being growth, but part of it, this dynamic you talked about. Product introduction is an important aspect of KLA. It's something that Rick and I -- in terms of a significant architecture change versus just new capability. But the other thing, particularly in this environment and the environment that we've lived in over the last several years, is it allows us to take a look at cost versus price. And as we work with our customers and we deliver a new capability that introduces a cost of ownership improvement over an established baseline that -- how much of that improvement do we take and how much of that improvement in the form of higher prices that we transfer to customers is something that we spend some time on. When you're in an environment where you're seeing cost pressure, that introduction allows you to then make those changes, really hard to go to a customer and say, I know we've already negotiated on this product, but now I'm going to go raise the price on that, right? It just doesn't work and in a somewhat consolidated industry. It's very hard to do. So it matters a lot for us to introduce this new capability for those reasons, but pricing but also competitive dynamics. And it's worked for KLA for a long time and it's something that we think is an important part of this the R&D model, but also the execution to be able to deliver it and ramp it and execute against that new capability and those commitments we're making with our customers.

And if we step back for a moment and look at the current industry dynamics, and it's a topic that we've been writing on for the last several years, which is the whole emergence by our semiconductor companies and their customers to focus on bringing more custom silicon ASIC, we call them ASIC solutions to the market. We have seen a resurgence in advanced ASIC XPU chip designs by all of the cloud and hyperscalers and the AI frontier model developers, more than 100 designs and design starts at the 3-nanometer, 2-nanometer node, these are some of the most advanced, most complex chip designs in the world. And given the large chip sizes of these custom XPUs, different circuit topographies. Your foundry customers, I think, have to custom develop special inspection, metrology recipes for all of these new ASICs, right? How significant of a driver is this dynamic to overall sort of foundry process control intensity?

Yes. You covered a lot there. I think, first of all, the design environment is always good because it introduces change, right? And change is hard for our customers to manage it's also introducing change of very high-value parts. And so the predictability of the yield, but also the performance specifications are very important. And then the you want to be able to deliver to a very tight market window, obviously, demand is very strong. And you don't want to start too many wafers. You don't want to start too few. So there's a fab efficiency element that comes for trying to manage all of those dynamics and process control plays a pretty significant factor in that. The other thing is it occupies the nodes. It drives up the commitment to those designs. And so as the next node starts to ramp, the previous node is generally very well utilized. And as a result of that, the purchasing for the next node tends to not be able to reuse any of the capacity or very little of the capacity from the previous node. So that tends to be a factor to customers are always trying to optimize their capital efficiency. So if they -- they have some utilization rates come down on previous now they try to use some of that capacity. A little harder when you've got a road map that's moving was easier when the road map was fairly stagnant back in the last decade. But that's certainly a factor. Now if you look at these particular designs, as I said, they have some of the challenges of production, right? Defects is a challenge, right, where yield, if you have the same number of particles falling a defect -- if you have a lot of chips or create defects, you have a lot of chips on a wafer versus a few, it has a significant impact on yield. And so you have that issue, but you also have the value of it. So you're willing as a customer, in some ways, it's kind of like insurance, to spend more on insurance, if you're ensuring something that is higher value. And so if you're producing something, you're going to spend a lot -- that has a lot of value. You're going to spend a lot more time ensuring that, that device not only yields but also functions as designed. So large die benefits KLA in a unique way in a way that I don't think really benefits process companies all that much, the size of the chip doesn't matter. The capacity matters, so big die that drives more capacity matters. But the dynamics of large die and how it affects process control and large valuable die is a unique benefit for the company. So designs uniquely challenged us. I don't think really the fact that you're running a lot of designs, a high-mix environment doesn't really change a process and opportunity when you're exposing the wafer, the but matters a lot to process control. So these dynamics are, I think, unique. And part of the reason why we see our relevance over the next 5 years, continuing to increase.

Yes, makes a lot of sense. Let's talk about the WFE spending environment as you mentioned in your prepared remarks, your WFE outlook for '26 started off the year at sort of low-to-mid $130 billion that was ratcheted up to $135 million, $140 billion at Investor Day and now to over $140 billion on the recent earnings call. That's 3 upward revisions in roughly 6 months, and you even mentioned in your opening remarks, since earnings, you've seen further sort of momentum. When you decompose what's driving each of these upward revisions over the past 6 months, how would you separate the contribution of the incremental demand? In other words, is it new fab project visibility being solidified. Is it customer pull forward to secure capacity slots ahead of an oversubscribed environment or is a lot of it just existing capacity footprints, where customers just want to squeeze more yields and output in a supply-constrained environment.

I think it's more the former of some of the things you said. I think, it's more clarity around what's coming in the second half. And as you get, I think, firming of schedules, certainly, our customers feel demand pressure. They're building below the demand levels that they have. And so I think as plans are solidifying, we're seeing the pressure to not only the clarity around some of these schedules, but also pressures to kind of pull in. And customers are trying to cadence all the tools that they're getting and to be able to ramp the facility. And so you always have those dynamics. But the momentum is clearly strengthening, right? I mean, I think '26 at some point, we'd probably bump up against a limit of what can be done because of fab space. And then that's 1 of the reasons why '27 conversations and visibility is so high. But I think it's really a number -- it is growing, not stagnating at this point. It's very strong.

And we'll get into the mix in terms of -- that will be the next question first half versus second half. But again, you mentioned sort of just even since earnings, you've seen a bit more momentum continued momentum in the business and it sounds like an incrementally sort of better outlook as you pointed out for the second half of the year. When we think about sort of end market sort of foundry logic versus memory is more of the more near-term momentum that you've seen biased 1 way or the other?

I'd say it's across the board. I mean 1 of the challenges we had early on is that it was a pretty quick turn in terms of customer demand starting to accelerate. If I go back to just the fourth quarter of calendar 2025, we weren't anticipating to see the kind of growth that we have in 2026 and talking about the November time frame or so, we saw that start to change -- and for our equipment, right. Our lead times, intrinsic lead times tend to be pretty long and generally around optical components because we're effectively building these high-performance microscopes, -- and so we have to go procure that supply. And it generally takes longer. Now we do a lot of things to try to hedge it and to reduce our lead time to customers. But there's only so much of that you can do. And so when you have that quick turn, it does take some time for our business to change. So the sequential performance in the first half of the year has been below, I think, what our customers ideally would want. But we see that starting to improve into the second half, and we've been pretty public about that. So I think from a momentum point of view, as we move in the second half, I think the second half is, we'll call it, sort of mid- to high teens, maybe 20% over the first half, somewhere in that ballpark, high teens to 20%. And so we think that we'll see some of that catching up happen. And if you go back and look at the last time in 2020, 2021 time frame when the last sort of significant ramp in the industry during the COVID time frame, you saw a similar dynamic play out. Now we think our supply chain is pretty resilient, because it tends to be pretty captive to KLA. We sometimes compete within our own product lines for capacity. But our supply relationships, particularly around critical components with long lead times, we tend to have relationships to those customers with the supply and just us in our space. And so as we make those commitments, and we're investing in them, we feel pretty good about their ability to deliver. As you start to move through an environment where you start to feel the strain and all those hedges and things like that you have built into your system start to -- the longer it goes start to feel pressure. So I think in the long run, we feel very good about our ability to sustain what looks like a pretty strong 12 to 24 months at a minimum. And we'll start to see some of that momentum pick up in a more meaningful way as we move into the second half of the year.

And on KLA's mix, high 60% foundry/logic, low 30% memory mix you anticipated for this year. Help us understand sort of first half versus second half foundry versus memory mix and capacity versus technology migration mix within that view?

Yes. So I think if I look at it and can move around a little bit from a revenue recognition point of view, in the second half has a little bit more memory. I think there's some timing around certain projects opening in the second half of the year that's influencing some of that. If you look at it overall for the year, though, memory is growing faster, but logic is still I think you're seeing this growth in leading edge of the logic. But if you look at other parts of logic, particularly like in China and legacy markets are slower. But the inflection is really in the industry is happening in DRAM. So I think in the second half, you'll see probably DRAM will be a little bit higher than logic. But you're seeing strong momentum across all of it. So I think that's probably the biggest issue in it. I mean, certainly, Memory is a slightly higher percent this year, and you have a lot of technology migration that's happening in memory. And so particularly around conventional memory, so that does put a little bit of pressure on process control intensity in that environment, offset by what's happening in high bandwidth memory, which is a different animal, if you will, in terms of how customers are investing in given some of the unique characteristics of high-bandwidth memory versus conventional memory. But for conventional memory in an upgrade environment, -- the opportunities for us are more limited. Now when you move into a more greenfield environment as you're retooling and equipping a new fab of new capacity, then I start -- that starts to change. And we're seeing with the progression of advanced design rules in conventional memory, but also the growth rates of high-bandwidth memory that creates a higher intensity environment as we move into next year.

Now on the April earnings call, speaking about next year, you articulated a view that the 2027 WFE growth rate should be higher than the growth rate expectations for 2026. In my mind, that's a fairly meaningful statement given the team typically has been a bit more measured on sort of forward period sort of color sort of giving it to us historically, right? So what specifically gives you the confidence to commit to 2027 growth faster than 2026? Is it specific customer sought commitments? Is it the visibility on the bookings and backlog, greenfield construction program visibility, the breadth of customer engagements or just the combination of all of the above.

Yes. So new fabs in logic, fabs and memory, new fabs and DRAM, but also in flash. So greenfield is part of it and particularly given the nature of the conversations around aligning those schedules with when these facilities come online. I think that's the biggest factor for us. Our customers certainly feel the pressure of the demand environment. And I think even after all that the capacity that's planned in 2017 is still shipping short of where their demand signals are. So that's probably the biggest factor overall. I think there's a -- I have a higher confidence level in the sustainability of some of the broad-based investment that's happening and also at the leading edge. As we move into next year. So I'm encouraged by that as well. You're right. It's unusual. We tend to have, I think, longer lead times. So I'd say around certain product types, perhaps more visibility -- but that engagement with customers is pretty high. And so that gives us the confidence that we'll see that. And our customers are translating their views of growth that are consistent with that.

You talked about the market share gain -- another year of share gains?

Well, certainly, in the electron beam world, particularly on the e-beam inspection, we've seen nice share improvement, and we've seen good customer adoption around single beam solutions and multi-beam solutions we're able to leverage what we call co-intelligence, where we're able to leverage some of that capability to increase the relevancy of our optical inspectors, particularly in production, where you can use -- these are some of our most advanced systems from an AI point of view, and you can use AI to identify hotspots or care areas, some of the terminology we use inside the company that you can point your inspectors more effectively, look for certain defect signatures and so on. So not only does it drive the point product competitive dynamic, but it also allows us to drive more relevancy across portfolio. Reticle inspection is an area of opportunity, and we saw some gain there as well. That market gets served both in the fab, where you have requalification of reticles that are in use. -- reticles are have defect challenges where you -- because you're scanning across the reticle printing devices that if you have a defect, you print on every one. And so it's important that either whether it's in fab that there's a contamination dynamic or the mass shop when you're writing the reticle or processes control in tendency is very high. There's a number of platforms that serve different parts of that market. And so when you look at that overall, we saw an increase in overall share, driven mostly by the real quality market in terms of where the growth came from. Packaging was another area of share gain. We moved in the #1 position in packaging. One of the things that's happening in packaging is -- that market is moving to need for higher end, front end like solutions. If you look at the integration of packages, particularly for high-performance compute. And so we've been able to gain traction, leveraging our front-end tools. We did a lot of development around the macro inspection products that we're now from handling and environmental point of view that we've now been able to leverage across the other rest of the portfolio. And so that engagement, particularly on the logic side, and share improvement has been very high. Memory, I think there's improving momentum. Some of the memory requirements are not as advanced from a technology point of view. And so I think that there's opportunities in memory. We're seeing it improve, but it will take longer. But certainly, the logic side is starting to reflect. And as we move into this year, we're starting to see a higher mix of more advanced systems that as you start to shrink the design rules and you move into like hybrid bonding type techniques that, that drives the need for more advanced capability. So I think that there's continued momentum there. And even on the base product that you compete in the market, the tool we won with is down rev to the current tool we're competing with today. So we think that the road map is very strong, both in terms of the products that are serving the capacity part of the market, but also those that are optimizing the sampling strategy, leveraging the front-end capability that we've had for some time. Finally, optical inspection continues to be a huge driver in the industry, both Gen 4 and Gen 5, Gen 5 on more advanced small defect issue and challenges and then using Gen 4 for gate all around because of the broadband nature of the architecture allows us to do a lot of a berry defect type analysis, which is a challenge when you start to move into more of a 3D structure environment, and we're seeing that in memory, but we're also seeing it in...

So you've got high bandwidth memory on the logic and foundries year-over-year. This year, you actually broke out the process control systems portion of the advanced packaging growing to $1 billion, like you said, high 50s, 57% year-over-year is the target versus your prior view of 30% growth. So what drove the upside in the growth outlook for this year? And where is KLA -- so we get this question a lot, where is KLA over-indexed? Are you overindexed to HPM, are you over-indexed to the 2.5D to 3D transition has that happened starting now and for the next kind of few years?

So the thing in packaging is that the market has, as I said, has moved in both in terms some of the processing requirements, but also the lines and space shrink, the density of the design rules as they shrink. It's driven the need for more front-end solutions. And if you look at the base macro inspection product that was used in packaging. That market generally segmented around front end was KLA and then we have some competitors that address more of the back-end opportunities. And so as that market has moved towards front end, naturally, the front-end product. So we've seen, particularly in the logic area has been where most of the share gain has happened -- you mentioned some of the new kind of evolving opportunities, they are also coming with disti. And so that will create further opportunities because the sensitivity road map is increasing, and so we'll leverage those other tools. What's happening in the memory is traditional bump inspection type opportunities, which is a less advanced or less complex application. You also have part of the market that's been running really at really high utilization. So there's the opportunity to compete. It's also a challenge to that making a change in a high demand environment is not something our customers like to do. So we've seen some momentum there. And I think that, that continues as you start to see the bump dimension shrink, driving more eventually moving to other bonding techniques to stack those die. I never talked about packaging back in like '21 or '22, we were like 50 basis points of market, 100 basis points. And now if you look at our share of the overall market, we're in 2025, I think it was 6%, what, over 2026, given the growth rates that we were talking about earlier that probably gets us up into the above 7%. So our relevancy in that part of the market has changed in a pretty meaningful way. And I think it -- you have -- it's 1 of the factors as we think about the long term. There's a lot of all these, I think, really good drivers in core WFE, both in memory but in erlogic and memory, but also you're able to augment this growth with what's happening in the packaging environment. The sampling rates are very high. The last thing a customer wants to do is go through all the integration that goes into integrating 1 of these very valuable devices and it doesn't function or doesn't work and you have to do rework or you have a performance spec question. So sampling rates, I think, continue to be high, and I don't see that changing all that much moving forward. And you'll start to see, as I said, the introduction to drive more optimized sampling strategy at more advanced systems. This year, we're going to see even BBP tools adopted in packaging lines. Well, right, to just give you a sense of some of the complexity challenges our customer is facing, certainly, you're going to see laser scanning tools adopted in a more meaningful way. So it's 1 where the portfolio is hard to compete against established tools, established relationships with customers, established performance over time.

Your services business grew 15% last year with an exit run rate of about 18% year-over-year. You carry that momentum into this year, right? 16% year-over-year growth in the March quarter. As you pointed out at Analyst Day, the team's target is a forward CAGR of sort of 13% to 15% growth, but -- looking at this year, very, very high customer utilizations. KLA is offering more advanced services, lots of focus on driving as much output per fab as possible through yield improvement, right? So coming into this year, driving above that 13% to 15% CAGR. I mean, does that momentum continue? How should we think about services growth profile for this year?

So I think this year will be in line with the target. It was slightly above the target last year, and I think it was impacted by some billable activity that posed at the end of '25. But cause the number to spike a bit. But in line with -- we think this 13% to 15% that we were able to -- that we upticked and that's despite some of the export control pressure that we have where we lost fabs -- is a testament to the -- what's happening within service. Our service business is very unique. There's no systems business in it. 80% of the revenue stream is contract. Contract renewal rates are very, very high. But in excess of 95%. The lifetime of tools because they're serving a lot of these markets for long periods of time. We're up in the 20-year plus now lifetime. If you would ask me that question back when I was first became CFO of KLA, I would have said 10 years, you would have said how much revenue over the lifetime of a tool -- in service stream compared the ASP, we said 40% to 50%. Back then, today, it's in excess of 100%. So we're driving more service revenue over the lifetime of the tool than the ASP value of the tool. What's happening in the market today is driving higher -- put them into our system. That creates an extra element of growth over time. So for all these reasons, we think that not only does the growth rate move at 13% to 15%, but we also think that the incremental gross margins move above 55%. [Technical Difficulties - Please refer to the preliminary transcript that will be posted shortly.]