Thank you for coming live. Great. Welcome back, everybody. I'm Joe Moore, Morgan Stanley Semiconductor Research. Very happy to have with us today Bren Higgins, EVP and CFO of KLA Corp. I always want to say KLA-Tencor Corp. I'm sorry.
Don't say it, Joe. Don't say it.
I'm a little old school, but KLA Corp. So, Bren, thank you so much for being here.
Yeah, thank you for having me.
We'll start with some overview comments, and then we'll jump right in.
Yeah. So, yeah, thanks for having us. We're happy to be here. And, you know, 2025, as we go into 2025, we're feeling pretty optimistic as a company. I mean, certainly a re-acceleration of investment at the leading edge across both logic and foundry and in memory is good for our business. The last couple of years have been more trailing edge focus. I'm sure we'll talk a little bit about that. But certainly ensuring that the company, from a priority point of view, is ready to support our customers at the leading edge, and that's shipping our most advanced systems, our ability to ship and support and service, to be able to support what is expected to be a pretty strong ramp, the 2-nanometer node activity around high bandwidth memory is all just a really strong part of the market. It's our strongest market. We're excited about that.
We have to continue to collaborate and deliver very advanced capability for our customers as we look at the technology roadmap over time. We're excited about some of the innovation that's coming, and we've got to ensure that our flywheel is about collaborating closely with customers and identifying unique problems where we can differentiate, making the investments we need to make to deliver that capability, and then being able to scale it, sharing the value of it, and then repeat that cycle over time, and there's lots of opportunities for us, we think, here moving forward with the scaling roadmap, what's happening with architecture changes, power distribution, and, of course, what's happening in advanced packaging, which I'm sure we'll talk about.
And then finally, the continuous efforts in driving productivity across the company, leveraging automation, leveraging AI, both in terms of how we run the company from an administrative and execution point of view, how we write code and leverage that capability in our product development cycles, basically to ensure that we've got the operating model that we can continue to deliver on what we expect to be very strong revenue growth, on incremental operating margin leverage on that revenue growth of our target model of 40%-50%. So those are the targets or the corporate priorities as we look at where we are this year. 2025, as I said, driven by leading edge, strong investment there. China is correcting, digesting a bit after a number of years at high levels of investment, also some export control effect that happened at the end of last year. It's affecting the numbers.
So, I think leading edge investment, China sort of offsetting each other. Our legacy customers, it's not getting better, but not really getting worse as we look at this year. And the NAND business growing a little bit off of a low level. And then in DRAM with the focus on HBM, we're seeing some investment there. So, I would expect WFE to grow in that part of the market. So, we talked about WFE levels of somewhere in the high 90s in 2024, and we think it's up mid-single digits or so here as we move into 2025 with the breakdown across segments consistent with what I said. Obviously, advanced packaging is also an accelerant to growth here. We did over $500 million in the company in advanced packaging business in 2024, and we think that number is at least $800 million into 2025.
So you couple that with our growing service business and the construct in terms of growth in 2025 looks pretty encouraging versus last year. So I'll stop with that, and then I'm sure we'll dig into some of these things I talked about with your questions.
Yeah, no, that was a great overview. I mean, I think we were a little positively surprised on the WFE forecast given that you did have 2024 closed stronger probably than most people thought. So we ended up in that $99 billion range, and you did have these incremental headwinds in China. So maybe just kind of talk about what that looks like. It is, as you said, a pretty big inflection. We go from kind of geopolitically oriented investing in legacy nodes. Now we're investing quite a bit in areas like Gate-All-Around and advanced packaging. It doesn't seem like we're investing that much in like 3-, 5-, 7-nanometer. It's really much more kind of at that very cutting edge. So that seems like that should play to KLA's advantages over the course of this year.
Yeah, I think you summed it up pretty well. And at the end of Q4, given the timing of some new restrictions, it certainly created a little drama in the last quarter of the year. But the team's executed pretty well as we managed around it. Obviously, we need to make sure that we can comply with all the rules and new regulations. And then at the same time, we want to do everything we can within that construct to support our customers. Big focus right now on N2 certainly is a big driver this year. There's some interesting dynamics that are happening at the N3 node and N5 node as it relates to product mix changing relative to the way the tool or the fab was set up from a tool point of view.
What I mean by that is, as you start to see more high-performance chips moving through some of these nodes, most of them are at the N5 node coming to N3, the characteristics of the die are very different, right? They're larger die. Defect density is more of a challenge. Certainly, the value is very different. And so where you might have set that fab up with an expectation of a certain mix and then all of a sudden that mix changes, it drives the need for some different capability. Single-die reticles drive a challenge. Large die have a single-die reticle. How you inspect those reticles creates a need for different capabilities. So we are seeing what we call backporting, where at the 3 and 5-nanometer node, customers are making additional investment to support what is a changing mix. But yes, you're right.
Most of what's driving our business this year is a strong focus on N2, and our customer has been very positive at the level of interest and design start at equivalent sort of maturity levels versus N3. And so that's a great driver. And I think our relevancy, our share of WFE opportunity so far, what we've seen and what we expect to see over the next two to three years as this node ramps to maturity is higher than what we saw at N3. Architecture change being the biggest driver of that. The last time KLA share of WFE approached 8% was when we switched from planar to FinFET devices. And now as we move to Gate-All-Around, we're seeing a similar effect where new defect mechanisms, new processing steps are all driving higher levels of KLA relevancy.
We think from a share point of view, we're in a pretty good position across the portfolio.
That's a great point, though, on N3 driving these massive die sizes and suddenly defect density matters a lot more because yield is geometric with die size.
And then drives the need for higher capability, where at one point you're like, "Okay, we know what we're monitoring for at a certain process, a certain design that's pretty predictable in terms of the design rules and how that translates to yield maturity." And then all of a sudden now you have to do more debugging in a new chip, a new design creates different challenges and the need for higher capability to do some of that, particularly when these dies are so much more valuable.
So that's a great point. Thank you. Maybe if we look at some of the markets, starting with DRAM, I think that's the biggest difference. You guys have, I think, double-digit growth in DRAM WFE this year. We're down, and we're down not because of any kind of market issue, but just the numbers got really high last year. So maybe put into context what's driving the optimism there around that market.
Yeah, different mix of customers. You did have some China business, low-end DRAM that drove some of the investments last year, not a factor this year. Most of the investment is focused on advanced DRAM supporting high bandwidth memory. So that's where we're seeing the investment. We're seeing it from a couple of customers at a higher level versus the third one, but at the same time, a lot of focus on this part of the market and a lot of innovation that's coming in terms of die stack plans and so on. So it's an interesting market. And then because of EUV and DRAM, we've seen a step up in process control intensity in advanced DRAM, which is a driver, but an HBM device creates a whole different set of issues as it relates to redundancy, larger die. You have larger die because you have to drill TSVs.
You have microcontrollers in the die itself. You have a more advanced logic-based die. And so for all these reasons, and of course, you need the reliability of each die in the stack to be performing consistent with specification. So all that is driving process control investment. And then you package it, and then the packaging part is separate from that, but certainly is also an inflection as you package that with the logic device.
Yeah, maybe if we could double-click on that, the HBM intensity for process control, you mentioned the TSV and the layers and the base die. Within the DRAM die itself, though, there's also a difference in process control intensity?
Yeah, it is for those reasons I talked about. Redundancy and repair is a big factor in DRAM where you can actually then do repair as chips come out. And so you're willing to take a little bit more risk. Now with a loss of redundancy because the die is bigger and the need for that space for the chip, that is created, you have the margin for error is tighter. Yeah, and so customers are having to inspect more there.
It seems like the pace of technology, I mean, going from HBM3 to HBM3E to HBM4 on the scale that we are. It seems like an important driver as well.
Yeah, I mean, a lot of that in the package. I mean, you're still going to have all these dynamics that are affecting the actual DRAM chip itself. But then as you package an HBM and you start looking at what's going to happen with the pitch shrinks as it relates to the bumps as an example, you start to move into hybrid bonding techniques potentially down the road because of some of the height restrictions. And then the more you stack, you have to make sure each one is functioning consistent with specification. So there's a driver of just ensuring that the chips function before you start the expensive packaging process.
Interesting. Thank you. KLA versus DRAM. Your DRAM number, by our numbers at least, outgrew the industry 2022, 2023, undergrew a little 2024. It seems like with what you're characterizing, you should have pretty strong growth. I guess granted other process control has a lot of growth around HBM too, but sorry, process chambers and things have a lot of growth around HBM also. But can you just talk to the prospects for you to outperform that DRAM WFE the next couple of years?
I think a lot of the things I talked about are going to increase the share or the intensity level, process control intensity in DRAM with the introduction of EUV, as process control as a percent of WFE went from 8% to 9%, up closer to 9% to 10%. If you look at some of these device issues that we just talked about, it pushes that number up another 100-200 basis points, we think, over time. We feel pretty good about the investment profile from the customers that are investing. The engagement level is high. The business that they're supporting and the strategies are pretty well aligned. We feel pretty good about what's in the plan for this year.
Okay. Maybe shifting to NAND, you mentioned getting a little bit better from a low level. Kind of interesting that companies are lowering utilization and increasing spending at the same time, but we've actually dug into it. There's a fair amount of inability to do QLC on legacy wafers when we haven't invested in two years. So it seems like there's quite a bit of technology-oriented investment that has to happen even in a tougher economic scenario. So can you talk about, you said you're seeing some of that, and then can you talk about the prospects for process control intensity in NAND to go up as well?
Yeah, you know NAND's not the strongest market for KLA in general, and it's growing off of a fairly low base. You're seeing layer counts increase. The upgrades are providing a certain amount of incremental supply for the market. Certainly, a greenfield opportunity is a bigger one. As we move from planar into NAND, it created other challenges around metrology. As you grow the structures vertically, there's a lot of different measurements that are now required. Wafer flatness is an issue. Overlay becomes an issue as you do double stacking and so on. So there's lots of interesting things that are happening there, but it's really the investment we're seeing right now is off of a very low level. I think NAND WFE this year is what high single digits maybe. Peak NAND WFE back in 2018 or so was probably high double digits.
So you're looking at modest increases off of a very low base. It's a driver, not the strongest driver from a process control point of view, but as we look at the overall WFE forecast, are we getting another billion, two billion out of the NAND market? Yeah, I think we will.
Great. And then back to foundry logic, the N2 node you talked about being strong for you. You've talked about picking up share of spend. I feel like it's a little tricky for us because everybody's kind of saying they're increasing share of spend in N2, but obviously there's a lot of with new transistor structure, there should be a lot of need for process control intensity. So can you just talk about your confidence in that? And then does it matter if we have one, two, or three customers that are doing kind of Gate-All-Around backside power?
Yeah, no, it's a good question. The fact that I'm talking about it this early in a node ramp certainly gives you some confidence, or at least gives me confidence that I'm comfortable with what we're seeing in terms of expectations around our share of WFE. Some of that obviously coming from intensity. So what I've said just for folks who don't know, as I said, look, if you compare an equivalent number of wafer starts at N3 versus our expectations for N2 over the first three years or so, we think that KLA's share of WFE is about 100 basis points higher than it was at N2 than it was at N3. So there's a couple of things that are driving. First of all, N3 is a node that's full, right?
And so the ability for the customer to migrate and use some capacity at N2, which historically is always a reuse factor that happens, is fairly limited. You have an architecture change. So you have the same number of EUV layers, which are still very challenging scaling layers, but then you also introduce the architecture change, which is more process steps. You're depositing more layers. The performance of the transistor is you get multiple sheets there, and typically the performance is going to be the performance of the weakest sheet. So there's a lot of focus on metrology to make sure that you get voltage contrast and transistor performance. And then you end up with different defect mechanisms. So you have all the usual defect challenges, but now you're seeing new defect types in this transistor structure.
We introduced a new version of our flagship Gen4 product, optical pattern inspection product, that we had to upgrade enough with so many bells and whistles. We had to make it non-upgradable. There's no way we could do it in the field. But because of the wavelength in the system and the ability to identify these defects that are recessed, it lends itself to the capabilities of the architecture of that system. So a whole new defect use case opportunity for one of our existing platforms in an environment where things have been moving in a positive direction from the scaling roadmap over the last couple of nodes. So for all those reasons, we feel very good about the position.
We're also leveraging some electron beam technology. KLA's all in on e-beam across multiple markets where we can use the inspectors with AI to look for low contrast areas to point the optical inspectors, so having the portfolio, we can increase the relevancy of optical by leveraging electron beam capability, so there's a lot of opportunities and a shortage of challenges here, and I think, A, I talked about reuse, number of designs, and then technology change, all very, very good for KLA. I think, look, at the leading edge here, question about providers. Certainly our primary customer is driving the N2 node. I mean, from a KLA point of view, we're very collaborative with that customer, and I think that the customer is. I think that's one of the things that has changed over time.
I think the collaboration versus transactional sort of supply chain versus a more strategic supply chain relationship is a factor in allowing us to be there to help solve long-term problems, to make the investments that are going to be required to solve the roadmap. The customer's under a lot of pressure, and I think as a result of that, I think there's a willingness to engage at a different level, and it's increasing our relevance as a percent of the market, so I don't want to talk about individual customers beyond that, but I feel pretty good about what we have here from a sort of a leading edge roadmap that's driving the business here moving forward.
But feel free to not answer this if it's uncomfortable, but it seems to me that the projection is that one customer's driving the lion's share of it and the other two maybe upside if they come back and start investing. Is that kind of the rate? I think we had setbacks with both of those, with customer two and number three last year that were very public, so.
Yeah, I would say that we have our plan is pretty de-risked in terms of what we expect from our customers, in terms of that fits everything I've said. Yeah, and so incremental investment I think would be upside to the.
Great, and then looking at N2 from your main customer's perspective, really interesting that compared to their last couple of nodes where one or two big customers were the pipe cleaners for that, there's a whole bunch of people that are using N2 relatively quickly, so there should be more design starts, more diversity of product that runs through there. Seems like that should also be good for you guys from a process control standpoint.
Yeah, the number of designs is, when you're debugging a process as you're ramping it, the number of designs, and debugging around one is a lot easier than multiple ones because designs test the design rules in different ways. So you end up ramping at a less mature state generally. And so the number of designs is absolutely a driver. And because of that, as I said before, you've got the previous node is consumed with a lot of designs that are flowing through. It also, I think, portends that as you move into the current node, you have the drivers I mentioned, but then also you're likely to see it be a widely adopted node, which drives a sort of a continuation of that behavior. As you move to the next node, you have to start over again, and the reuse factor continues to be lighter as well.
But not only these designs, but the nature of what we think is coming in terms of the types of designs back to large die defect density, very valuable die. Like anything else, if it's very expensive, you're going to be very careful with it. And so we're seeing an increase in sampling versus a traditional mobile die just given the value. So I think those designs are going to actually then contribute. It's always been good for KLA. I think this is one of our messages. We talked a lot about design start environments and reuse and a scaling roadmap and architecture changes driving our business. But as you start to shift the revenue composition, that there are additional drivers in there that allows us, I think, to continue to increase the relevancy of the company moving forward.
That's helpful. Thank you. Can you talk a little bit about Rapidus? It seems like there's pilot lines ramping up there, pretty high potential spending plans. Can you talk about your relationship with them? That's Rapidus being the Japanese sponsored project.
I think it's safe to say that whenever we have a customer that's investing, we make sure that we've got the resources in place to support that customer. So you would expect to see some investment happen there with some limited number of wafer starts as that project spins up. We started to ship into that project back in Q4. We tend to be a little bit earlier than most because all the process tools that come in have to be qualified. So it's a factor in our revenue plans for 2025. We'll see over time how that scales, and obviously it's going to be impacted by the ability to progress the roadmap and to be credible with customers and be able to prove a roadmap over time. But we're encouraged by the activity. We're staying very close to the customer.
We feel very good about the share position we've had, and we're making the investment in application service, putting parts in place to be able to support that fab as it comes up.
Thank you. You mentioned China restrictions. Can you talk to the restrictions that you've put in place around the government decisions next year, last year, I should say? And then moving forward, I'm not going to ask you to prognosticate exactly what happens, but how do you navigate the potential uncertainty of new rules, new tariffs, things like that?
We have to be nimble. And we certainly have more compliance folks in the company than we did five years ago to really understand. And of course, as you might expect, given the complexity of what we do, there's a lot of nuance to the regulations. New regulations that dropped in December, we talked about a $500 million impact to 2025 in terms of lost market access. A portion of that is service, about 30% of the number, plus or minus 100 or so. Now, estimates don't really change that much. We've seen some marginal or incremental strengthening on the leading edge, but also some incremental growth in China. So estimates more or less came down or didn't really come down. They stayed relatively flat. So look, we have an active engagement with the government to try to provide some perspective on our position.
Our view has always been that we will support the government directives. We want a level playing field. We think that as it relates to how U.S. suppliers are treated relative to some of the foreign competitors. And we do, as I said earlier, we do everything we can to ensure that we're following the law. And at the same time, within that construct, we try to do everything we can to support our customers.
Helpful. Thank you. Your RPO declined last quarter quite a bit, and the RPO for beyond 12 months declined. Can you just talk to what's going on there?
About half of it was related to business that we had booked. RPO is effectively your backlog. Business that we had booked that was now restricted. We just took it out. Now, could we get licenses for some of that business as we start to engage with the government? There is perhaps more clarity in these new rules where there might have been some ambiguity before. Does that lend itself to licenses? Maybe. We'll see. The licensing process has been somewhat onerous and lengthy from a time point of view. It could be a mitigating factor, but I didn't want to sort of factor it in because it's a little hard to handicap that. We'll see in that area what happens. What was the rest of the question? I'm sorry, Joe.
Yeah, just in the longer-term RPO.
Oh, oh, RPO. Yeah. So that was a factor there. Then there was a piece of it that looked that RPO had built up over time related to some of the supply constraints. So we had a lot of business out there in 2021 and 2022, a lot of greenfield projects, customers getting in the queue, new projects that were coming online down the road. And so orders would want to get, customers would want to get in the queue for deliveries when the facilities were complete. We also had certain products that were supply constrained, and so they wanted to get in the queue for that. Over time, as we've shipped to those projects, I mean, backlog is backlog comes in, backlog goes out, right?
But we've seen some of those projects we've shipped to. We've been able to manage through some of these supply chain issues as we've gotten new capacity online. And as a result, the backlogs went from very elevated levels, over a year in terms of months of backlog, and now they're somewhere around nine to 12 months. Historically, in the company, it was around six months. I don't think it ever really gets back to six months, but it is normalizing a bit in terms of customer behavior generally consistent with the past. I think customers are more sensitive to ensure that they get the tools they want when they want them. And so I think there's a little bit more discipline in ensuring that they're getting in the queue, we're getting the orders. If you're a new customer, we typically take deposits also.
I think you're going to see some normalization effect in that. I don't think it ever gets back to where it was, but I think we like to live generally between 9-12 months in terms of the mathematical backlog, and I think it's going to normalize to that.
Is your lead time commensurate with that, or have you seen lead times? Have you been able to bring lead times in as these supply chain issues have been?
More or less. I mean, look, we try to run our factories to ensure we have the ability to deliver to all the different demand scenarios we can contemplate. So we don't lose business because we can't deliver. So there's always some juggling effect that can happen. We have some customers that are large customers that don't necessarily give us a lot of lead time. We stay very close to those customers, so we've got a distant view of forecast. But at the end of the day, we do the movement of tool slots across products to ensure that we can deliver. But generally, if you're talking about where you're quoting, you're quoting in that timeframe.
Okay. And then can you give us an update on the services business? You mentioned there was a China impact there, but otherwise pretty consistent growth.
Yeah, the China impact is you lose a fab access. So we had about 15% growth last year in the service business. Our service business is a little unique in that most of the revenue is contract-based. We've had multiple growth drivers for service: A, the shipments that we've had over the last few years as WFE has continued to grow, the useful or the average lifetime of tools increasing, and the average, the ASP that we're getting on service, both on our advanced systems, but just in general on the tool base. If you go back over time, I would have, the average life used to be in the low to mid-teens. Now it's closer to 20% of ASPs, probably closer up about a point overall.
If you would have asked me when I became CFO, I would have said we capture 50% of a tool revenue in a service stream over time. Today, we're capturing over 100% of a tool ASP in terms of service over time. So there's been a lot of good drivers for our service business. So as we move into this year, when you lose access to a fab that was going to contribute revenue, you have the headwind of that. So I think service is likely grows this year somewhere in the high single digit, maybe low double digit, depending on how things play out. So you have that effect.
But over the long run, those drivers I've talked about, and if you're planned generally, as we think about the long-term of the industry, that you have a growth rate of semi-revenue, you have slightly rising capital intensity, that it's a fairly efficient model. And so capacity is going to get added to meet the demand. And so ultimately, we would expect that if there's capacity that's restricted that can't meet its ability to supply some of the long-term demand, that that capacity will have to get added elsewhere. And so that's as we think about the long-term 12%-14% service growth model, we take a fairly efficient view.
Obviously, in the short run, there's a headwind, but over time, that's why we still continue to feel comfortable with that, plus some of the drivers I mentioned earlier, opportunities and packaging that are evolving from a service point of view, that the 12%-14% model is one that we're comfortable with.
Great. Let me open it up to the audience for questions. I have more questions if we don't have any. One in the back.
I have a softball question, which is, so we hear a lot about how pretty much everything is good for more process control spending. You go to new nodes, you add new innovations, you want more EUVs, good for process control. Is there anything that's in the roadmap that is neutral to process control spending, or it's just sort of not affecting the outcome for you guys, or even a little negative?
You know, one of the, you know, I've talked a lot about some of the drivers in the market and where those percentages are. Obviously, the mix of WFE is always a factor because the nature of Logic Foundry process control intensity is very different than memory, although memory is improving, like I talked about earlier. So as the mix shifts, it will be a factor in terms of our ability to maintain a relative growth trajectory. Now, our general view is that the growth of WFE and Logic Foundry will be faster over time than memory. But you can end up in certain periods of time where memory tends to be more of a cyclical driver, where episodic periods of higher levels of memory investment. So in those periods, it puts a little bit of pressure on the relative performance model.
So I would say that for the most part, the mix factor is a factor. But I think in the areas, as I've said earlier, in the areas which have been bigger headwinds, I think we're starting to see some changes in that.
Can you speak about the opportunity in hybrid bonding, potentially that could be incremental TAM for you guys, and what kind of tools or process steps do you think you can leverage your front-end expertise?
Yeah, I think hybrid bonding creates one of the next inflections potentially in high-bandwidth memory. I still think it's further out in terms of adoption, but it allows, it creates more process. It enables tighter feature sizes and a very fundamentally different inspection requirement. And so we think it plays to the strength of the capability we have in the portfolio. Today, we're addressing most of the memory opportunities with our lowest-end systems. So as the requirements change, as lines and spaces shrink, it creates the opportunity or the need for more capable systems, and we have the product roadmap for that. So it's one of those inflections, and particularly in environments which are growing, you need inflections to create a new opportunity for selection. And so I think as we get there, it creates an opportunity for KLA, but I think we're still a ways off from that.
I have a question on e-beam, given the momentum this year. Is KLA's strategy still focused more on bundling, where it's kind of complementary to sort of the optical portfolio, versus say, maybe you're moving more towards defect inspection, that your competitor is relatively dominant in that space?
Yeah, so in the electron beam market, there's multiple applications. There's metrology, there's review, and inspection. And what we've done historically, it's been fairly complementary markets. Electron beam systems are incredibly sensitive, but very, very slow. So being able to use electron beam systems to try to point from an inspection point of view, to point your inspectors, as I mentioned earlier, is and has been a reason for being in the market. But there are opportunities for that market where it's growing, where customers want to use that capability for very small defects, or in situations where they're looking for subsurface defects, which we call voltage contrast, where we're coming into the market with single beam capabilities, multi-beam capabilities. There's also in-die metrology that's also part of the e-beam market.
And then there's review, reviewing defects that you're finding with your inspectors, and being able to do that review and classification process very, very quickly. So there's a lot of opportunity across e-beam. If you look at the broader dollars that are being spent in inspection, so pattern inspection, probably 80% of it gets spent with optical solutions. But given the growth that we expect, there will be opportunities to see the e-beam parts of the market grow. We think it's fundamental to the portfolio. It adds some complementary features to other products. But on a standalone basis, there's an increasing level of opportunity. And we need to be in all the markets, and so we've invested a lot over the years to ensure that we have platforms that we think are competitive. And so we're excited about what's to come.
We'll have more to say about it at our investor day in June.
Any other questions?
Maybe if you could talk to the question of competition in optical. You know, it comes up from time to time. Some of your competitors talk about growth in optical from small levels. Anything to note there? Obviously, KLA remains quite dominant in these markets.
I think the competitive environment has been fairly stable. Look, it's an attractive market. There's a lot of good growth factors. Our profitability levels are attractive, so it's always going to attract competition. We think that we're unique in that we can offer a portfolio solution to customers that we don't have point products, and our customers can choose. We're somewhat agnostic about this, but depending on where they are in a node maturity or node ramp, technical solutions that meet sort of technical requirements, and then also very quickly pivot to more capacity-centric systems, and there's a network effect across the systems in terms of how we can leverage them, how a customer can leverage the data that comes out of them, so I think it makes it a very strong competitive mode. Haven't seen any real changes in that part of the market.
I'm assuming you're talking mostly about pattern inspection competition. So I would expect we'll see 2024 market share report out here in a couple of months from Gartner as a third party that reviews it. I would expect market shares to be very stable.
We were talking in the hall about the Entity List customers that cannot buy U.S. tools, moving to Chinese tools where applicable. Is there anything they can do in the inspection realm?
There are some local competitors in China. I'd say that where we have lost, where we've lost access to customers competitively, it's been to non-U.S. suppliers, and we have some Korean competitors, some European competitors, Japanese competitors that have been able to service that business where we haven't had access to it, so we do run into some local indigenous competition, but given the complexity of what we do, I think the nature of the portfolio, the level of support we can provide, generally that competition has had very minimal impact, but they're out there. There's, you know, I would say $200 million of investment that's happening in places that we, that's being captured by those competitors.
Great. Well, that brings us to the end of our time. Thank you so much. Appreciate it.
Thank you for having me.