Good afternoon. This is CJ Muse with Evercore ISI, semiconductor equipment analyst. Welcome to day two of our Evercore ISI Inaugural TMT Conference. Very pleased to be hosting KLA Corporation. We have Bren Higgins, EVP and Chief Financial Officer, and Ahmad Khan, President, Semiconductor Process Control. We are going to do just a fireside chat. Hopefully, lots of good Q&A. First off, just want to say thank you both for being here, and look forward to talking with you.
Thank you for having us, CJ Good to be here.
Excellent. I guess I figured I'd start off with the current environment. In your last earnings call, you talked about lead times, about a month or two greater than normal. Curious, when you think you'll get that back to a more normalized level, and is there an upper limit as to what you could actually ship this year, given this extension of lead times?
I think one of the things that's been interesting about this ramp and the momentum we've seen, if you just go back to where we were in November and what we thought this year might look like, we were talking about a mid-single-digit type growth year. We went to earnings in January, and we were talking about mid-teens. As we've moved through the last earnings cycle, now we're in the mid-20s. Clearly there is a bias to the upside here, and we've seen continued momentum in the industry. The lead times have been longer. They've extended out.
We do have inherent lead times on a lot of our products that can be as long as one year. We do a lot to plan and work with those suppliers to hedge our long lead time materials to put us in a position that we can deliver. Generally, all products, on a weighted average basis, right around six months or so from the time that we start planning for a shipment to the time that we actually ship. It's a little bit longer, as you said. We talked about this at earnings. We talked about an environment this year that we would see sequential growth through the year. We didn't quantify what that would look like. We thought we'd see that as we move through. Obviously, that means the second half is stronger. We were adding capacity.
When I use capacity, I talk about it in the broadest sense. We were adding people, we're adding clean room space in all of our major factories. We're working with our suppliers. We're really proud of our suppliers for a lot of the work that they've done. They're investing. We're investing in them. For all of that, we would expect that we'll be able to deliver to that sequential growth, as I mentioned. We would see some of that capacity to continue to come online. I expect we'll be able to work that down. We'll be closer from, let's say, eight months down closer to six months as we get to the end of the year.
I think we're pretty well-positioned, certainly the momentum has been building in the industry, I would expect that we'll be able to continue to support that momentum moving forward. I'm not going to quantify the second half. I have a lot more to say about it when we get to earnings report out on June, at the end of July. I feel pretty good about how we're positioned and how the company's positioned from a relative growth perspective, in terms of WFE growth.
Excellent. At the same time, you're hearing multiple-year CapEx projections from leading manufacturers. I'm just curious, as it begs the question, are we seeing a structural change potentially in the market where your, whether it's backlog or soft backlog, becomes much more concrete and your visibility might now start extending much farther as companies want to get in line? Or is this just systematic of WFE tracking better and therefore it's just taking some time for the equipment industry to adjust? How are you thinking about it?
Well, I think it's tracking better both in terms of size, but also the pace in which we got here, and so I think that's been part of the challenge. Certainly, the end market dynamics that are driving our customers to talk about long-term investment plans, and I think it's a function of, A., the strength in the markets, but also it's the rising capital intensity. For all those things, I think it's good in terms of how we plan. You always have customer readiness and timing issues that you work against. We have taken some, I'll call it risks, in terms of how we're managing with our suppliers, how we're positioning the company to ensure we have dedicated capacity. Are we putting orders out for longer periods of time?
Doing those kinds of things, I feel pretty comfortable with the goal of being flexible and being able to respond outweighs the risks of having some extra parts, particularly given the strength of our service business and the way demand has stratified to where there's a number of markets that require trailing edge tools, so tools are living longer. We think it's a reasonable trade, and we want to make sure that we're positioned for that upside. We're also making investments in, as I said earlier, in facilities to make sure that we have the space to support a higher through-cycle growth rate going forward than what we've seen in the past. We're excited about the opportunities.
I don't think it necessarily means it changes the visibility, although I think we're working much more closely with customers, and given the nature of what we sell, we've already had a pretty good engagement already. Typically, customers hold their slots with us because the demand for these products is so significant that if you get out of position, it's hard to get back in, and somebody else will take that slot. I think that that part of the discussion has already been pretty good, and I just see a continuation in that. I don't know, Ahmad, do you have thoughts? I think the other thing I would say is that there's quite a few inflections in the marketplace that are happening on the device architecture side, and that is really pushing customers to make more investments in R&D, and we benefit from that. EUV implementation is quite important.
It's increasing, and KLA participates quite well both on the mask side and also on the optical inspection side. EUV and DRAM is getting introduced. That, of course, requires infrastructure spend for our customers, and that puts us in a very good position. I think all the dynamics that Bren mentioned are the right ones. In addition, I would say is that the device infrastructure and scaling that is happening is putting quite a lot of investments with our customers where we participate.
Excellent. That's a great segue to your process control business. Just as a starting point, you've guided the business to grow 30+%, at least in April, that's what you thought the business would grow in 2021. Curious, if there were to be upside, where would you expect it to come from customer base profile as well as from product base? Not saying that this is going to happen, but if it were to, where do you think the upside could come from?
Yeah. I'll start and let Ahmad jump in later. I would say that first of all, the biggest strength that we're seeing in our business is in our pattern inspection business. Pattern inspection has inflected with the introduction of EUV with the Pixel Logic foundry. We had a strong growth year last year. We expect pattern inspection to outperform the market again this year. I think that's an area of investment driven by the dynamics I talked about. Reticle inspection is another area where we expect that to grow faster than the overall market. The reticle qualification, both in terms of mask shop but also fab requalification in an EUV environment, is becoming increasingly challenging, we're seeing opportunities there.
As you know, CJ, we sell a number of products to support that from the high-end mask shop requirements of pattern inspection all the way through requal, which is done with reticle tools for requalification, but also doing print check with high-end optical pattern inspection, wafer inspection tools, where you're printing on the wafer to validate the fidelity of the reticle. Those are two areas. I think the thin film business behaves a lot like a capacity business for KLA, and so certainly, as you've seen this ramp in WFE, you've seen a lot of strength in thin film. I would expect that to grow faster than the market as well. Yeah, I think it's a pretty good summary. I think with device scaling continuing and implementation of EUV, smaller defects are becoming more and more important.
For that reason, our Gen4 and Gen5 products, optical pattern inspection products, are seeing pretty significant growth rate. We're glad that we were feeling this pressure from the market to build and ship more and more systems, and our collaboration with customers is very good. We were able to see this demand coming, and for that reason, we are prepared. Optical wafer inspection is doing quite well with the implementation of EUV. As more and more EUV layers get implemented, our mask infrastructure is very, very critical with optical mask inspection, which does all the pattern qualifications in the mask shops, and followed by we're introducing new products now with very high-resolution capability for mask shop qualifications. In 2019, I'd spoken about Gen5 EUV extension for print check capability for mask qualifications in the wafer fab.
That's been a critical product for our customers because there's no other really good solutions that are production-worthy today for mask qualifications, and we're seeing growth there. Lastly, I would say metrology has seen quite a bit of growth with number of layers increasing, both in the films area, the Optical CD area, the overlay area. To take a segue, as you see, our customers continue to scale, and they're going to bring new innovative designs for N3, N2, and beyond. For that, there's a lot of investment being made in research and development with Buried Power Rail, with Gate-All-Around, EUV implementation and DRAM. KLA does really well in these areas is because each time you change the transistor architecture, the metrology and inspection modes completely change.
For that reason, we have to develop new capability, work closely with our customers to implement that capability for them in R&D and eventually in production. I think those are the three main areas where we see a pretty large growth.
Excellent. Kind of a good segue to the next question. I guess, Ahmad, I was hoping you could kind of speak to the most exciting tech inflections that you see ahead. You alluded to Gate-All-Around with High-NA EUV ahead. Eventually, still many years out, but eventually 3D structures in DRAM, higher NAND stacks. Where are you most bullish from a process control perspective in terms of intensity and growth? You can define the timeframe, whether it's 12 months or five years. What are you most excited about?
I'm actually excited about all of them, the reason is because each time you change the transistor architecture or you do advanced scaling, KLA benefits from it. From a business perspective, but also intellectually, it's very challenging to come up with these new solutions. One of our strengths really is that we have a v ery wide portfolio in KLA to offer to our customers. Our customers rely on us to create these solutions. They come to us and explain all of their metrology inspection challenges in the early part of the engagement. With EUV, we have seen that with Gen4 and Gen5 growth.
We have seen that in reticle inspection. There's quite a lot of growth that's happening there, and also film control. Now with Gate-All-Around coming in, that changes the front-end transistor architecture quite a bit. For that reason, there's a lot of growth that's going to show up in metrology. Film control is going to be very critical in metrology for the front-end transistor architecture, and we're developing new systems for that.
At the same time, while we saw a huge growth in Gen5 for EUV, we believe that with Gate-All-Around and 3D DRAM, Gen4 will see growth rate. Gen4 wavelength is very in line with those types of challenges. That's for that reason we're developing a new version of Gen4 that's going to come out in the market in the next two years that is going to really focus towards Gate-All-Around. 3D DRAM is a little bit further away. It's about three to five years away. For sure, KLA will participate in that. I think the first inflection in DRAM is going to be the implementation of EUV. Those are some of the inflections that are happening.
Buried power rail is another inflection, and for that, we are developing a number of different tools for overlay control, for wafer-to-wafer bonding, and those implementations, and I think I'm very excited about those inflections. In the end, we have developed and released a new product that utilizes X-ray for metrology control, for shape and overlay control, and that product is now implemented at several customers as customers go to CMOS under the array and also go to higher stacks. We're going to be enabling them with these technologies. Again, what excites me is that we have a portfolio approach that enables us to really do investments in new product development and ensures that our customers can count on us with each of these inflections. I'm excited about the fact that the industry is continuing to scale, which brings more challenges for KLA, and that we can help serve them.
Excellent. Perhaps we could drill down into each of those. I guess maybe starting with Gate-All-Around. I believe you guys have been talking about both new optical layer metrology tools to support that market. I think you just highlighted the Gen4 update. Is that the new optical tool we're thinking about? I guess, could you elaborate on what is so special with nanosheets and why a new tool is optimal?
Again, our thinking always is we have a very good simulation capability which enables us to work closely with our customers to determine what structures they're going to be printing on the wafers, and we can simulate which wavelengths and which light bulbs we would need to develop well ahead of the need for the customer. We do that two to three years ahead of the implementation. In this case, the wavelength that is required is very in line with the Gen4 architecture. We need to add many other additional capabilities in Gen4 to be able to do Gate-All-Around. We started this development a few years back, and as Gate-All-Around gets introduced, we will have this new version of Gen4 coming out in the marketplace to deal with these challenges.
At the same time, on the metrology front, there's a lot of sensitive shapes that our customers want to control with very high degrees of freedom, and we're working on new optical technologies and heating technologies to do that control, and those products will be coming out in line with Gate-All-Around timing.
CJ, one thing that's really interesting about the state of the industry today is the extendibility that we're getting off of multiple platforms. Historically, you transition to the next generation, and since it was mostly about linear scaling, you would need the incremental resolution and sensitivity that the next generation would provide. You still need that, particularly with the introduction of EUV, because we are continuing to shrink and scale from a litho perspective. These other technologies that customers are pursuing are enabling us to continue to leverage other platforms to solve some of these problems, that the problem isn't necessarily just about finding smaller and smaller defects. It's great for our gross margin profile. It's great for our R&D intensity that we can extend these platforms.
We still have to invest in them, but that extendibility allows us. If we didn't have that, we'd have to continue to develop next generation. If we knew our newer platforms, we'd probably have to spend more on it. We can take very mature platforms, extend them this way, solve some of these customer problems, and do it with, because of that maturity, it has good design stability and a good cost-optimized platform. It's really one of the unique things about this environment, both in terms of leading edge, the things we're talking about, but even as you take it back into the trailing edge with some of the automotive markets, industrial markets, IoT markets, that we're able to serve a lot of those markets with older generation tools that we're making modifications to, but extending platforms that the incremental margin is really nice.
I think that's a change for the better that we're seeing going forward, I think is a contributor to the higher levels of profitability that we've seen overall.
It's pretty helpful. Again, all around, I guess maybe if we could move to EUV, and we'd love to hear your thoughts around Gen4, Gen5 applicability there, as well as any update in terms of timing for your actinic inspection tool.
Yeah. Maybe I'll start and then you can continue. One of the things about EUV is that, as you know, that the wavelength of the system is far shorter than the previous generation scanner, 193 nanometer scanner. At 13.5 nm, not only you print smaller lines and spaces, but at the same time, the defects that print are also smaller. For that reason, we have to innovate not only in wavelength on the inspector side, but also on algorithms and techniques that will be able to catch these defects. We have seen quite a lot of growth in our optical inspection portfolio as EUV has been implemented year-on-year. These growth rates are well above the WFE growth rates for these businesses.
Clearly because due to scaling and due to EUV, the sensitivity requirements have increased quite a bit, and our customers are counting on us with Gen4 and Gen5 platforms to scale with it. We also have introduced our eBeam portfolio. We were not present in eBeam inspection for several years, and we introduced a new product called eSL10, and this product works very closely with our Gen5 product, Navy, and then also with Gen4, to ensure that we are able to focus on areas that customers really care about, but at the same time provide full wafer coverage for process control. That's doing quite well. There's quite a lot of good reception from our customer base for that. With that, I think that's mainly the EUV introduction and how our optical business is growing.
On your second question related to actinic inspection, I think I'll step back and say today for N5, most of the reticles, I would say 85% or greater, are going through KLA's optical inspection system, a 193 system in the mask shop that qualifies most of the reticles. We believe that that continues in N3, with our 193 system, where most of the reticles will go through that system. It's a production-worthy system, dynamic database capability, which customers really count on to catch the most sensitive pattern defects, which will eventually cause yield-limiting problems in the wafer fab. We are introducing our next generation eBeam inspection system. It's a multi-column reticle inspector. We've shipped our first one to our customers. That system is focused on pattern inspection. For reticle qualification in the wafer fab, we have introduced Gen5 with EUV extension.
I spoke about that at the 2019 Investor Day Conference, since then, we have implemented that tool system in all of the EUV customers, it's doing quite well for requal capability. I think that the final question related to our actinic inspection system, we believe that introducing the product at the right time is most critical in ensuring that the cost of ownership and the capability matches our customer needs. It's more or less in line with High-NA EUV timing, when High-NA EUV gets introduced. We believe that is the right timing for an actinic inspection system that would meet the customer requirements and also be the right cost of ownership solution. Until then, we have our Teron 640e system for pattern qualification, 8SX system for pattern qualification, and our Gen5 and EUV 670e XP2 system for requal. We believe that that's what the market needs today.
Yeah, CJ, from a competitive point of view, I think one of the things that we offer our customers is a portfolio of solutions that allows them to balance their technical needs with their economic requirements. As Ahmad said, whether you're using the optical system, we just shipped our first eBeam multi-column tool for EUV high resolution requirements for EUV reticles. The combination of Gen5 and requalification systems for KLA allow them to optimize their use cases in production, both from a technical point of view, but also from an economic one. I think there are many examples of that across multiple markets. I think that's one of the unique competitive advantages of KLA, and I think we're seeing that and expect that that play out in the market.
You don't get rewarded for being early, we think we address the market requirements with what Ahmad has talked about and what I spent some time on here. We have a roadmap that we're confident in terms of delivery of future capability. I think one of the things around optical inspection with EUV, the EUV has driven a number of new designs at the leading edge. That's driven because of the Moore's Law attributes that have come from incremental scaling, has created a much broader design environment. Our customers are managing multiple product process flows and multiple products that they have to deliver in market windows that are tighter and that deliver volume product. You're seeing an inflection for inspection and optical inspection capabilities, the workhorse of production. We're able to address that market with multiple products.
It's also because of the technology drivers of EUV. I think the demand for those designs is causing the use of that capacity, but also that the ability to migrate that capacity to reuse it and try to optimize for those investments at the next generation nodes is harder for customers because the demand is there, but also you now have these technical challenges. I think that's been an inflection. Rarely do we see investment at multiple nodes like we're seeing now. I think that's also a change in the industry driven by the introduction of the seven EUV scan.
Got you. Very helpful. Well, you beat me to the punch. I was going to next get on the competitive landscape. I guess to follow on to your question around portfolio effect, I was hoping you could perhaps attack the benefit of a portfolio effect from your market leadership perspective and the breadth of products that you have. How does that enable you to stay in the pole position in all of process control?
I think the first part I would say is customer intimacy, right? We have very good customer relationships, where our customers feel that they can share all of their challenges with us. Because we're not a point product company and we have a very wide portfolio, we are able to engage with the customers on all of their metrology inspection challenges first. We start with very early R&D engagements where customers give us their designs, where we do simulations to understand what would be the right product needed at the right time, followed by near R&D engagements where customers have the designs on the masks and on the wafers, where we bring in our early engagement systems, and then eventually we work with them with HVM.
The idea of a portfolio approach is that a customer can give you all of their problems, you can help first solve them, and then you have a portfolio that enables you to really improve the cost of ownership. If you think about it, we have eBeam inspection system that has the highest resolution in the marketplace to find the most smallest and killer defects. That system enables our Gen4 and Gen5 system to do full wafer coverage of defects across the wafer, across all the lots, and across all the designs. We have reticle inspection that takes customers' designs into consideration and does data database inspections, which are very critical to find the smallest defects that will cause reliability issues or failures in customer designs.
We have laser scanning systems, which are able to offload layers from our high-end optical systems over to a cost of ownership capability that our customers would need as they do in-line monitoring. We have larger size defect inspection systems that do line monitoring for non-critical layers. That's kind of like an example of I've gone from reticle inspection all the way to very high-resolution systems to optical systems that do full wafer coverage and then to do cost of ownership layers and then more cost of ownership layers and helping solve customer problems and yet providing very good cost. That's the portfolio approach that we try to use. KLA at the same time has a very good investment thesis, reinvestment, and this enables us to get into new segments that we weren't participating in before. eSL10 is that example.
KLA was not present in eBeam inspection, but because we had such a large optical portfolio, our customers asked us to be present in eBeam inspection so they can give us all their problems. Another example is we participated in Optical CD for metrology applications, but as our customers went to 3D, customers said, "This is very hard to do optically by itself. Can you please invest in eBeam inspection and X-ray metrology so that we can see deep into the structure and determine what is the shape of the product and everything else?" The third example I can give you is MMA. We did an acquisition many years back for flatness for wafer manufacturing called ADE. We were able to take that technology and now introduce the same tool for in-line wafer shape control for 3D NAND and DRAM.
Only KLA is able to do that because, again, we are not a point product company. We have a full portfolio. We can see the technology and then reassess the technology in other segments. These are some of the examples of having a very wide portfolio that enables us to serve our customers well.
Very helpful. I guess, hoping you can kind of address the increased potential competition on the eBeam side. If I take a step back and think about the monopolies that I see in the world of semi equipment, I think of lithography and ASML, I think high aspect ratio etch for Lam, and I think of optical inspection for you guys. With defects getting smaller and smaller and requiring eBeam, and your market share clearly dominant in optical, less so today in eBeam, what does it take for KLA to really drive market share on the eBeam side? What can you tell the audience to assuage fears of rising competition on the eBeam side from either AMAT or ASML or others?
Why don't I start, Ahmad?
Yeah. You can fill in after. They've always worked in a complementary way. You always have this challenge with these technologies of balancing resolution and throughput or speed. One challenge with eBeam is eBeam can be used for defect discovery use cases, but it is very slow. When we talk about slow, we're talking like thousands of times slower than optical solutions. Because of that, it really doesn't solve a customer's production problem or challenge that, look, I've got to be able to inspect the full wafer, I've got to be able to do it in production times, and so on. They've always been complementary technologies, and if you think about the pattern inspection market, 80% of it generally is solved by optical solutions and about 20% from eBeam. That's been the general ratio. It moves around a little bit.
It can be as high as 85, as low as 80, but generally in that ballpark. Where we are with that is, how do we increase the relevancy of the optical inspection products? It isn't one where we're looking to detect contrast. We're not necessarily looking to resolve defects. We have to find signal to noise. We've introduced eBeam capability not to have another product in a market that is a nice market, but not a huge market relative to the optical market. Can we use, through the use of machine learning, can we use the AI tools to train our inspectors to be more relevant in a production environment? That's how we address that market. I feel very good about the extensions of optical capability. I think as we introduce more and more eBeam capability, it will help augment that.
To the portfolio approach of how we're going to address customer problems, we can use that to drive optical. Optical's the workhorse. It's what customers want from a production point of view. If we can make it more relevant over time, then we can create higher levels of density for it.
Yeah, I think that's excellent coverage. I would say the same thing, which is there's no substitute for optical inspection. Gen4, Gen5 laser scanning systems, we are investing in Gen6. We will continue to invest in optical and optical resolution and techniques that will find smaller and smaller defects. I don't think that changes, and we'll continue to do that. I think the optical relevance will continue to stay there. On eBeam front, what we are trying to do is really, as Bren said, add portfolio products that'll assist our optical portfolio overall. That is with eBeam inspection. I talked about eSL10. We're introducing a review product. We are introducing some products at the metrology front to help our optical metrology systems.
I think we will continue to do that because we have a portfolio approach where optical essentially takes care of the workhorse, and then we get assistance from eBeam. I spoke about that in 2019 at the Investor Day, where we said we had a vision of eSL10 assisting Gen5, and that's what it's doing now. There are other areas, like for example, in reticle inspection, we are introducing a multi-column eBeam inspection, but very high-resolution and high throughput capability for reticles. One of the reasons why you could do that is because the way reticles are made and structured and how many designs you have, it is conducive to having a multi-column system that could meet throughput requirements. In wafer inspection, it's just not possible to do.
Again, portfolio approach, optical is the path, and then we introduce point products to assist optical to help our customers overall.
Very helpful. Well, I think we've got 30 seconds left. One last question for you. I think investors would really agree that you're probably going to achieve your 2023 target model at least a year early. With that as the backdrop, where should investors be focused next for KLA?
Well, it's interesting, maybe it's two years early, in terms of you just look where consensus is, and of course, we have to achieve that this year, but certainly in position to do so. That's really been driven by the strength of the overall market. There is a number of products that Ahmad talked about that we talked about at Investor Day that we're going to contribute incremental opportunities for share and intensity. KLA share of WFE to improve as part of our 2023 plan. Those products are coming to market. We're seeing the market, but aren't contributors of steady state revenue yet. I think that's still to come and why we're confident that in our ability to drive KLA share of WFE up 75 to 100 basis points over the next few years. We feel very good about that.
The leverage in the business, we've been talking a lot about competition. I think the best indicator of the true differentiation of KLA, look at the gross margins, the gross margin profile, we segment report. It's pretty clear about how we're performing from a revenue point of view. Our competitors don't necessarily have to do that, also the margin position within those businesses. We're seeing strong margin, and I think it's an indicator. The volume's been good, which has been a driver of leverage, also of the differentiation of the products in the market. I think that leverage continues. We've seen some gains from the volume, I would expect that we'll continue to see high single-digit type growth over time. I think semiconductor equipment with rising capital intensity will grow faster than semiconductor revenue.
We ought to be able to drive through 1.5x EPS growth of the revenue growth rate. Feel very confident about that moving forward. Our capital deployment strategy is pretty clear. We're big believers in capital getting valued only if it's deployed productively. We'll continue to increase our dividend and grow our dividend in line with the growth in our free cash flow, return most of the cash flow that the company generates. We're pretty excited about the future, both in terms of end markets. I think that the revenue opportunity, the product positioning, and the overall financial and operating leverage in the business.
Excellent. Well, I think we've run out of time. Gentlemen, thank you for everything today. Really appreciate it.
Sure. Thank you, CJ. Thanks again for having us.
Thank you. All right, next time live.
That'd be great. Look forward to it.