The biggest change has come from the mix of foundry/logic to memory. And memory is increasing some. So yes, there are more layers, there’s more investment going on, but it’s still balanced by they’re going to ramp. We hope not just the R&D, but they’re going to be ramping in terms of across the board. That’s how we get to the model that we is based on process control intensity, inching up over time as processes just get more challenging.
Bren Higgins: Yes. The road map schedules have held pretty well together. What we’ve seen is customers adjusting some of the capacity plans. So as you look at 2024, you’re more likely to see, for example, more N3, you’re going to see N2 activity and we’ll start to see some of that soon. And – but most – the bulk of it will be more at N3. What you likely don’t see is you probably won’t see a lot of investment from our major customers and some of the more legacy parts of their businesses, where they pull back. But to Rick’s point, we’re seeing some of it. We’ll see that investment as they ramp. We’re seeing more investment today in production given the number of designs that are moving through the leading edge nodes and the different process flows is creating opportunities for or – and more challenges for our customers in terms of process control and defectivity challenges across different designs as they test design rules in different ways.
So I think we have our normal historic exposure to R&D and to ramp. But over the last few years, we’re seeing, particularly with the introduction of EUV and the progressing of scaling we’re seeing more adoption in what we call the HVM or the high-volume manufacturing phases.
Joe Moore: Thank you.
Operator: Thank you. Our next question will come from Blayne Curtis with Barclays. Your line is open.
Blayne Curtis: Thank you to let me take the question. I just wanted to follow back up on the comments you just made on foundry/logic. So it was flat. It seems like China is probably up within that mix, and then you said leading edge is weak. I’m just kind of curious how that changes for December. It seems like the outlook is fairly flat. So is that weakness in leading-edge stabilize, and then any perspective as to where leading edge goes next year?
Bren Higgins: So I feel like where we are today, I think your stabilization comment is the right one. I think we’ve derisked it and given that we tend to be more of a long lead-time provider. I think we’ve made a lot of the adjustments that we needed to make already in terms of how we’re planning for this year. And as we move into next year, I think if you just sort of aggregate leading-edge activity, we’ll see as customers start to provide a little bit more insight. But again, back to the stabilization comment, I don’t see it declining from here.
Blayne Curtis: Thanks. I just want to ask on service, in your letter, you talked about getting back to that 12% to 14%. I just want to know, is that assuming any utilization increases? Or is that just purely the tools coming off of their agreement?
Bren Higgins: Yes, it’s the latter. I think we’re expecting utilization to slowly improve, but the bulk of it will come from new tools coming into contract.
Blayne Curtis: Thanks.
Bren Higgins: I think to expect to start to see overall industry improvement into 2024, the first thing you’ll see is utilization start to improve. So we would expect that and then once you see that, and then eventually, utilization gets to a place where customers need new capacity and then those decisions happen.
Operator: Thank you. Our next question will come from Mehdi Hosseini with SIG. Your line is open.
Mehdi Hosseini: Yes. Thanks for taking my question. Just a quick follow-up. As you think about the R&D price, especially you highlighted gate all around at some point, we have to change the narrative to High-NA. And I want to just give an update, how do you see opportunities as it relates to High-NA specifically on the patterning? And I have a follow-up.
Rick Wallace: Well, I mean, what High-NA enables is the continuation of scaling, right? So that’s been good news for KLA. You notice process control intensity in general, but more specifically for KLA has gone up as EUV has started to be adopted because now you’re scaling – we’re not on what was traditionally Moore’s Law, but we’re seeing scale it. So High-NA means there’s going to be more scaling happening, and that’s going to be good and specifically good for KLA because it drives the highest performance requirements, which plays to our portfolio strength. So part of what our modeling is when we look – we don’t see a lot of High-NA happening in the time line that we laid out for 2026 for our Investor Day, it’s after that. But we’ll see early stages of it before 2026, and that will drive – continue to provide more opportunity for us to participate in higher process control intensity.
Mehdi Hosseini: Are you implying that Gen 5 could be – the use of Gen 5 could be extended to high-NA for…
Rick Wallace: Absolutely, absolutely. We’re still using Gen 4. We’re using Gen 4 now, because of the extensions that we made in the platform, not just in terms of wavelength, but adding more processing capability, the leveraging of AI, the use of both Gen 4 and Gen 5, actually Gen 4 will out ship Gen 5 this year, and we’ll continue to see that adoption. So, it really is talking about the critical layers and we have more extensions in mind mine in the – on the works that we’re doing right now for Gen 5 that will extend it well into the even in the high NA, which is going to come after that. So, we feel very good about our optical product portfolio.
Mehdi Hosseini: Okay. Okay. And then the second follow has to be on China, it seems like for KLA and the peer group, the China mix is getting closer to 50%. Could there be a scenario where opportunities for KLA would actually step up given the fact that many of these customers are new, and they have yield issue. And I understand China is mostly for trading edge, but with new entrants, new players with the higher entrances and improving yield by these new players, have a higher mix of China for KLA relative to the peer group?
Bren Higgins: Well, we have a lot of customers that are subscale, that are trying to develop process capability and demonstrate capability to customers, also invest for viability over time in terms of longer-term node progression. So, in early stages, upscale stages like that, you’re going to see a heavier investment in process control. Now, as they continue to put road maps, it might stay there because they never really had a huge meaningful amount of capacity at each node. But you do see higher levels of adoption early on as you’re trying to – because if you think about it, you might buy a few process tools here and there, but you need the whole suite of process control. And so that’s why we tend to see a little bit more activity there.
But I think given the desire to progress along road maps into progress nodes, you’re going to see, I think, a continued level of investment overall. But certainly, as you start to mature and if you’re running a limited number of designs, process control intensity will hire in production than it used to be, is still lower than it is in what we’ll call the ramp phase of a project.
