Hello, everyone. Welcome to the 26th annual Needham Growth Conference. It is good to see so many people in the first in-person day. Joining me here is Nova's management team. Thrilled to have here Gaby Waisman, President and CEO, and Dror David, CFO.
Hi.
Thanks for coming here and joining me today. I really appreciate your time, and-
Thanks for having us.
Looking forward to the next few minutes of our discussion. I know you've been coming to this conference for quite a few years. My pleasure hosting you for the first time as a covering analyst of Nova. Okay, maybe to kick off the discussion, right, I want to start with a high-level question. So we know the business of process equipment companies is quite straightforward, right? The customer needs capacity. They buy the process equipment, process tools. I mean, you can kind of correlate how much capacity they're building to how many tools they're going to need to buy, right? But for process control, equipment companies, metrology, inspection, looks like it's not always as straightforward, right?
Almost feel like you need to constantly prove out the capability, why it's helping the customer, either, like developing the technology, improving yield. It's not as straightforward. So I guess my question is: what do you think is Nova's... As a process control equipment company in the, in the metrology space, what is Nova's specific strength relative to your competition, and in terms of providing that value to the industry, and, of course, making sure you get rewarded in that process, for the innovation you do, for the industry?
What I typically tell customers is that when they buy a Nova product, they actually harness the brainpower of the best process control team in the industry. Buying the tool is just the first step, because you need to develop applications, and as you said, you need to address challenges that the customer doesn't even know at first. And when they come across those challenges, you need to cater for specific solutions that are definitely leveraging the capability of the tool but require a specific development. And this is a perpetual process, that once you do, and once you gain customers' trust, and the customer understands that the team has this capability, they're actually presenting additional challenges and applications, and that drives tool adoption. That's one.
Second, our challenge as a company is to understand what's going to be the next-generation process, derive the process control challenges, and develop core technologies that could solve those process control challenges. Now, we understand that, and to a certain level, what we do, because we don't fully understand the range of applications, is we develop a core technology that admittedly, at first, we don't understand the full spectrum of the potential of that core technology. But once it's out there, you can definitely get rewarded by that innovation in gaining both market share and an additional range of applications. I'll give you one example. We have a standalone OCD tool. Now, a few years ago, we developed a unique capability, spectral interferometry. That is a different capability to our competition.
When we did that development, we didn't fully realize the potential, for example, in addressing hybrid bonding or advanced packaging applications. Customers essentially said, "Well, you have this capability. You can separate the data noise of underlayers. You can do direct measurements of through-silicon vias. You can understand line topography using that tool. Why don't you try and solve our challenges in advanced packaging?" We did, and now we're seeing adoption of that capability and displacement of other slower, more expensive technologies by having this capability that first and foremost was developed for the front-end, but customers adopted also for the back-end and packaging. That's an example of what we do and how we tackle and address those challenges.
Mm-hmm. Mm. So, I mean, we're going to get there, but it does feel like for a lot of people who don't know about Nova, I mean, it kind of your company kind of strikes me as constantly developing new technologies, new capabilities, that there's a lot of customer push. It seems like you're doing. But we'll get to some more specifics there.
Sure
... definitely. So now let's dive into those kind of details, right? I want to ask you about specific product lines. I mean, no, I mean this is an exhaustive list, but I think the main products, I mean, especially the newer ones, Nova brought to the market over the past few years include, I mean, the newer version of the XPS, X-ray photoelectron spectroscopy. Okay, I got it correct. Okay. The material, that's a material metrology tool, right? And the Prism, that's the optical CD or optical dimensional metrology tool, right?... Elipson, the Raman spectroscopy, right? That's a lot of technical terms here, but—and of course, there's a Metrion. You guys just made a pre-announcement, right?
Metrion is something called a SIMS, and that's short for secondary ion mass spectroscopy. Okay. That's a lot, but in the context, right, you guys have had this lab-to-fab strategy, right? All these tools, I mean, some of them personally, I actually have used as a PhD student. I mean, when I was doing my work on material science, but you guys are bringing that into the fabs. What do you think you are in terms of the product cycle for each of these new products? Can you kind of give us a rundown? I mean, that's three or four new products there.
So let me start by perhaps putting all of those tools in particular baskets. What we are doing is addressing dimensional metrology challenges, mostly with optical capabilities. We're focused on material metrology, in which we're very unique, using mostly X-ray-based technology and the Raman, new laser-based one, and we're doing chemical metrology, which is a completely different plane. In each one, we're driving innovation and a roadmap that is intended to increase the overall TAM and capture as much market share as we can. And, and I'll elaborate about each and every one of those products, but this is essentially our claim to fame. And we also understand that the only way to gain that market share, and of course, drive margin, is to bring a very differentiated value. This is the only way we can succeed.
Now, in terms of our dimensional metrology, the spectral interferometer is one technology that I mentioned that is driving standalone OCD market share. We are seeing now, dozens of tools already deployed, and I believe that by the end of this year, most advanced customers, if not all of them, will have that tool. In terms of integrated metrology, which is addressing the polishing or the CMP segment in the market, and in general, we're adding imaging capabilities to our metrology in order to further cement and increase our leadership in that domain. In the material metrology, as Charles said, our strategy is based on lab to fab, and let me say a few words about that because this is a very interesting capability in which we are very unique. We started with lab to fab with XPS, and we've seen great success that I'll talk about.
But I want to try and explain why it is so important. So today, when you have high volume manufacturing, both R&D and high, high volume manufacturing, you need to take the wafers or break them into coupons and send them to a lab, an external lab, which could be owned by the customer, but it's not part of the production line, in order to understand some of the specific merits of the process steps and whether or not you have met the process window. In order to do that, you send it over, either by an overhead transmission ability, or just take the wafer to the lab. You have someone in the lab, typically a PhD, to look at the benchtop tool, analyze the wafer, understand what are the spec requirements, and then send the results back to the fab.
Now, this is obviously a lengthy process. It involves human abilities to analyze and then send the results back. It's typically days till you get the results back, and that's typically on a single location, a single sample on the wafer. This is not effective, especially when you want to ramp up in R&D stages, or you have high volume manufacturing, in which the cost of mistake is very high. So the idea is to automate that entire process and bring that capability in line, and not only bringing it in line, but you don't need a person to analyze. It's automatically connected to the SPC, to the process control chart of the fab. And not only that, you can extend your abilities and have a few samples on the wafer, understand the within-wafer variation, and do it much faster.
So from days, you're moving into minutes or hours. That's a revolution. So we've started with XPS a few years ago. About 40% of our business, 30%-40% of our business now is a result of the success of XPS moving from lab to fab, and we're doing the same with two additional tools. One is the Raman-based tool, which is doing stress, strain, and crystallinity. The other one is the Metrion, which is the in-line SIMS. And what we have seen is that from four customers that we already have today, and this is a new tool, we're going to double the amount of customers, and we are now in evaluations in two major customers that are going to be added in 2024 in order to cover the advanced node manufacturing. So we definitely see adoption. We are just in the beginning.
This is early stages, but we are seeing adoption for both those tools. In the Elipson case, the Raman case, we have already two out of the three major ICs. One, in one case, we are the PTOR for the production of DRAM in the most advanced memory customer in the world. So we are very, very, I would say, satisfied from the ability to drive this innovation and have adoption and proliferation in our industry.
Thanks. So I think you may have touched upon a little bit, right? But, but, I want to ask you about the go-to-market, right, go-to-market part. It does appear to me that, you guys are following a very consistent, rhythm, I would say, a path, I mean, in terms of go-to-market. To start with the leading foundry, I mean, very often, it sounds like, proliferate to all the leading foundry logic customers, then maybe second-tier foundries, that's, that's probably the easier one, I mean, than, than especially those in China. But then, of course, very important, the advanced memory side of the customers, right? So my question is this: we do see that, Nova may have the highest, revenue exposure to the leading foundry in Taiwan among your peers.
Of course, not every year is the highest, but if I look at the range, looking at like a SPC chart, it does look like very high. Like can be somewhere between 20%-30% of your total sales in any single year. How do you feel about this dependence on this, on your largest customer? Is there still room for you to, like, further proliferating your products to other customers, so maybe you can reduce a little bit, of that dependency? Or maybe that dependency is not exactly a bad thing. So what, what do you think?
I'll start with saying that we're very proud and very honored to have a partnership with all the leading manufacturers in the world, definitely with the leading foundry. We have published that the third quarter of last year, the first adoption of Prism over there for both back-end and advanced packaging. But we've seen a paradigm shift in that respect. With the new products that I mentioned, we have adoption in both logic and memory. Our ability to drive that innovation is not secluded to that leading foundry, but we are definitely seeing that shift in adoption across the board. It's true that we first and foremost pursue the more advanced nodes, both memory and logic foundry, because the value is there.
The adoption of the rest follows once you have the PTOR position, the strong position with those advanced customers, the leading five. So we are targeting those at first, and then we are seeing the adoption on the trailing nodes, et cetera. Now, you may ask me that later, but on the trailing edge nodes, we have seen the adoption and success as the result of the introduction to that leading foundry and others. In future, I believe it will be the same for both memory and foundry.
Got it. So kind of, kind of more, this is maybe, asking the same question from the, from a different perspective, right? I mean, I think investors looks like, they may be largely seeing Nova as a foundry play. However, when I look at your revenue split by application for the past few years, right, that, your revenue exposure, I mean, between foundry logic and memory, seems to be tracking actually very closely to the overall WFE mix. And I mean, 2018, memory was high, foundry logic was low, was like probably 60-40, and, and you did a 55-45 in that year, 2018. Last year was the other way, right? Yeah, last year was like foundry logic, 80%, memory, 20%.
I mean, you only kind of gave us the first three quarters, but looks like you're tracking in that ballpark, right?
Mm-hmm.
So in a way, kind of feel like your exposure is actually pretty balanced. But is that a result of some choices you made, or is there some coincidence? Because there's always a little bit historical path dependent in terms of a company's exposure to different parts of WFE. And shall we still be expecting Nova to maintain that kind of balance, the mix, going forward?
So I'll start with saying that, we increased, dramatically increased the number of customers. We were focused on a few customers just five or six years ago. In 2023, we are reaching about 60 customers. So that's one. So the split is definitely much more even and includes many more segments. Second is, I'd say that in 2022, we had about 50/50, memory and, logic, foundry, and our long-term model is 60/40, 60 for logic foundry, 40 for memory, simply, simply because metrology intensity is higher with logic than foundry. Definitely as a result of the investment patterns, WFE in the beginning of, in early stages of 2023, we've seen kind of an 80/20, split. In the third quarter, we already had 65-35, so it became more balanced.
You can see by those numbers that we are quite even in our presence and market shares, and definitely across product lines between memory, logic, foundry. That's where we want to be. We don't want to be skewed to one area rather than the other. We want to be split among as many customers as we can and obviously increase the market share across the board.
... Great, thanks. So maybe I want to ask a little bit more about numbers. So, we're at an investor conference, so of course, we want to be a little bit forward-looking, right? But when I look back, 2022, Nova grew the product revenue by almost 40%, right? When the WFE was up only by 10%. But last year, I mean, 2023, right? We're in 2024 already. The product revenue, I mean, kind of based on your Q4 guidance, of course, I took a guess, what could that be at the split? But it seems to be a little bit like it's going to decline by roughly low- to mid-teens %. But WFE was probably down, depending on whose estimates you're looking at, maybe down single-digit %.
Can you kind of help investors here to understand why you had a massive outperformance in 2022? Seems to be a little bit underperformance in 2023, but what should be the right expectation going forward?
So the jury's still out about the exact numbers of 2023. On the WFE, of course, we need to exclude the litho that grew in 2023, and we need to look at the process control segment. I'd say that, of course, we're in a quiet period. Our mid-guidance is $127.5, so we can realize or understand where we are. But, and that's a big but, I expect to still outperform process control and WFE as a company. That's one. Second, in 2024, what we foresee is a growth of about mid-single digits of WFE, probably starting with DRAM, then gaining more momentum with logic, foundry, anywhere in the middle of the year and towards this Q3, Q4, we're probably going to see 3D NAND joining the pack.
But overall, about 5%, and we definitely expect to outperform that number.
Great. That's very interesting color, 'cause you're kind of expecting the sequence of the recovery being DRAM, foundry logic, and the NAND. But it seems like you're confident, you feel optimistic that's all gonna happen in 2024.
We are fairly confident in our ability to outperform-
Okay.
WFE in 2024. Yes.
Okay, got it. So maybe let me ask some longer-term questions. One of the reasons why, I mean, the stocks of your peers, I mean, Onto and Camtek, unfortunately, I, I mean, probably that fortunate for them, I mean, they have been doing better than for, than, than Nova last year. I mean, well, well, I mean, the reason may be that they were traditionally in the, in the packaging business.
Mm-hmm.
Right? And, but now packaging is kind of seen as the bottleneck for, for generative AI-related, chip demand, GPU, TPU, AI accelerator, et cetera. So the general view seems to be that Nova may be somewhat missing out on that trend. I'm sure the team will dispute this idea, but, can you kind of elaborate a little bit how Nova is positioning itself in advanced packaging? You kind of touched upon a little bit, right, when we talked about, I mean, deploying Prism for advanced packaging applications. But, but want to get your general thoughts on how Nova is positioning in packaging, and what's the expectation of growing your advanced packaging business from here?
So Nova is exposed to advanced packaging in around 10% in 2023, overall, and we've seen very significant growth in that respect, and we still anticipate and expect growth of strong double digits in 2024. Our exposure to advanced packaging is across our divisions and across our product lines. On the chemical metrology, we are the market leader in packaging, and we are seeing significant growth as high-bandwidth memory, as well as other advanced packaging, comes into driving that business, similarly to the other companies that are in that business. We are also investing in our roadmap for having additional products, essentially in wets and cleans, that are going to add about $100 million to our TAM in 2027.
In the dimensional metrology, I mentioned both the integrated metrology and the standalone OCD as being adopted in advanced packaging, and I want to put some color into that. So first of all, in terms of integrated metrology, we started about two years ago in customizing and modifying our tools for advanced packaging. For hybrid bonding, you need flatness, complete flatness of the wafers. In order to do that, you need to polish, and you need to have control on that polish, and integrated metrology is the best-suited product and tool to do that. But in order to do that, you need to customize the tool. We did so, and currently, I would say all of the advanced packaging customers and high-bandwidth memory have integrated metrology tools, whereas we hold, I would say, the vast majority of the market share. So that's integrated metrology.
On the standalone OCD, with the Prism that I mentioned in the beginning, what we're seeing is a displacement of either traditional optical metrology tools or displacement of more expensive, higher, or let's say, higher cost of ownership, more expensive and slower tools by the technology that we introduced, for through-silicon vias, for edge roll-off, for the flatness of the polishing. So these are adopted in advanced packaging in a fairly, I would say, impressive way, and we're still expecting to grow that business. And last, but definitely not least, we're seeing the beginning of the adoption of material metrology for advanced packaging. So I would say that all of our divisions are taking part.
If I need to quantify, I'd say that materials and dimensionals alone are going to add $200 million of TAM in 2027, on top of the initially projected TAM for our strategic plan of $1 billion by that year. So I think we have a fairly good position in advanced packaging, and we're definitely going to see very significant growth in that segment in 2024.
Thanks. So lastly, maybe I wanna ask this China question.
Sure.
China as a region, I mean, quite consistently, right, I think 18%-20% of Nova's sales over the past few years. How are you, how are you positioned in China, on the lagging edge, foundry logic, and the DRAM? I mean, I'm not excluding all the discussion about NAND for now.
Sure
... relative to your competition, and these two segments still seems to be expanding.
Yes.
How should investors think about... I mean, that's a related question. That's a two-part question, sorry. How should investors think about the impact from the potential tightening? I mean, if there's tightening. We're not making any prediction here. If there are tightening of the U.S. export control rule, which may or may not come, but how should we think about when it relates to Nova's business?
So I'll start with saying that the geopolitical rift is driving the business across the board. We see this in having Europe, Japan, China, and the U.S. all driving domestic production, with TSMC and Samsung building fabs in Arizona and Texas. We see Micron investing in both Boise and later in Clay. We see the JV between Sony and TSMC in Kumamoto, in Japan. Europe, which is going to have Intel and probably TSMC fabs across the board. So the geopolitical situation is definitely seeing semiconductor as a core ability that needs to have domestic roots, and that's driving parts of our industry.
Specifically in China, we've worked for the last seven years or more in order to be able to serve that market, which is geographically dispersed, and the nature of the beast in our business is to be close to the customers wherever they are. So we see China, from Dalian to Nanjing, Guangzhou, Shanghai, Wuxi, and Chengdu, covering China, and we've built the company in order to address those challenges, and we see that investment bearing fruits in 2023. In 2022, we had about 30% of our business in China. I guess in 2023 will be a bit higher. But the market share that we have, and our ability to provide the service to those customers is definitely very impressive.
Now, it also has to do with the fact that when we deliver tools to fabs in China, we know what the tools are doing. We have the service engineers there who are developing the recipes with our application team, so we understand how the business is evolving in China across those fabs, logic foundry, DRAM. Naturally, we're following and adhering to the regulation, so the entity list is impacting us as much as it impacts Applied or KLA or Lam. But we do see our position and market share as, as strong, and, this is definitely helping us drive the business. We do see in 2024, still solid investments in, in China. We have, good visibility into the business in China this year, so we are very confident about our ability to deliver and continue to grow.
Thanks. I think I've run through all the questions I have. Now I think it's time to open it up for the audience to ask a question. Any questions? Yes.
When flash memory recovers, how many months afterward in terms of the lagging impact on when you see the start of equipment purchase?
So if I understand the question correctly, you're asking: What's the time lapse between having an either upturn or a downturn in memory to WFE, right?
Yeah.
So we've seen in the last quarter of 2022, we've seen the memory getting into a down cycle, and that had about one or two of delay on the WFE. We expect the same with the upcycle coming, let's say, the end of last year towards the first quarter of this one.
I just wanna clarify, you're expecting, like, maybe a quarter or two delay relative to the-
Sure
... the memory chip recovery. Okay.
Yeah.
Any more questions? I think, maybe my last... I just wanna check, if anything you feel like we should have addressed, but we, I didn't ask anything, and I think just leave it open-ended here.
So the last thing I perhaps I'd say is that we are seeing 2024 as an elevated year for R&D investments. I think that at least I tried to explain why innovation is so important, and this is continuing in 2024. We're investing across our product lines, whether it's adding high-energy XRF to the XPS in order to have additional species and expand on our application space. It's investing in deep learning and machine learning in order to drive both XPS capabilities as well as our optical product line. I mentioned the wets and cleans, and we have additional abilities that we are adding across the board in order to maintain that competitive edge, not only through the horizon of the current strategic plan, but obviously much beyond that.
I just wanted to emphasize that part of our strategy for 2024.
Thanks. All right, I wanna give another chance. Any questions?