Nova Ltd. (NVMI) Earnings Call Transcript & Summary

Good afternoon, everyone. Welcome to Day 1 of Citi Global Technology Conference. My name is Atif Malik. I cover U.S. semiconductors, semiconductor equipment and communication equipment stocks here at Citi. It's my pleasure to welcome Gaby Waisman, CEO, Nova Measuring Instruments; and as well as Dror David, CFO from Nova today; and we also have Miri in the audience from Investor Relations. I'm going to start off with my questions first and then open it up to your questions. Welcome, Gaby.

Thank you so much.

Gaby, I'm going to start with the topic that's on everyone's mind in artificial intelligence. When we met at SEMICON West a couple of months ago, I think it was kind of an emerging topic in terms of how the equipment makers are going to benefit from HPM, packaging, chiplets and everything that was going on in the back end. I think since then, some of your peers have seen some strength in orders. We'd love to know what you guys are seeing on the HPM and packaging side of things.

Definitely. So thank you for that. Before I start and answer this question, just as an anecdote, everybody is talking today about generative AI. So Nova actually introduced AI and machine learning into the process control industry about 7 years ago. We pioneered that introduction, and we're currently seeing tremendous adoption across the board wherever we have our metrology tool, which are augmented with mathematical modeling and machine learning to the physical one that we have. That's just an anecdote as what we have. Now in terms of AI -- and AI in that respect, these are very important milestones for the industry. It cements the road map of our customers and gives a much needed tailwind in order to drive the advanced nodes development and obviously, proliferation. Now for us, specifically, we see AI today impacting 2 areas. One is naturally the high-bandwidth memory as well as the advanced packaging in which specifically we can relate to hybrid bonding. But wherever you see that coming, we have enabled and we have positioned ourselves in a very unique and disruptive way, which is, first and foremost, related to our stand-alone OCD. Last week, we have announced the first important milestone of an order we've received from a leading foundry for the prism in advanced packaging. Now this is very unique because of the fact that we have a disruptive technology, which we introduced as an orthogonal, I would say, channel of information in addition to the traditional ones. This incorporates a spectral interferometry which is powered by a unique algorithm Swift. And that hardware and software enables specific capabilities in applications such as first silicon via and other flatness related aspects of the manufacturing, which are obviously called for when you're doing hybrid bonding and other advanced packaging techniques. Now this enabling technology not only enable -- not only powered and propelled our share gains in the front end, but gives us a unique position in displacing other slower and more expensive technologies that are not necessarily optical being used today in advanced packaging. And that, of course, is a step-up function in our business, an area that we have not been playing at in the past.

Gaby, one question I have, and I don't have the best understanding what's happening on the back end is like where do like Camtek and Onto participated? Because they've announced some orders and where you guys are coming in.

So Onto and Camtek are obviously continuing their inspection business. They've always been the major players in the inspection business and their announcements are essentially on having more inspection in the packaging and advanced packaging area. We are coming from the metrology side, critical dimensions and metrology in general in which we are not competing with Camtek and Onto. And as I said, for the stand-alone CV, we have a unique capability, which is being positioned there. Interestingly enough, we're now seeing adoption of our other parts of portfolio, whether in dimensional or in material in advanced packaging, which is also helping drive our business in that area. And that's naturally coupled by our chemical metrology business, which has a very strong foothold in the packaging and now seeing growth as a result of AI or high-bandwidth memory application space, which is growing. So we see that growth and momentum across the board with our entire portfolio.

Okay. So it's similar to the strength that you have on the front end side on the metrology. Gaby, one of the things I like covering you guys and following you guys is because you guys have kind of the best ears and eyes in terms of what some of the leading chip manufacturers are working on and whether it's a major foundry in Taiwan or microprocessor company in the U.S. And over the number of last 5, 6 years, you've done a very good job in terms of diversifying your products across the 3 major spenders globally. So if I look at where we stand right now, if you can share with us how you're seeing maybe the end markets like leading edge, logic, how do you see the trailing edge mature investments shipping out for the remainder of the year into next year and also in memory, which has been fairly depressed this year?

So memory has been down and so is foundry and logic this year. So we've seen that downturn phenomena throughout the year. And for us, with our unique position and growth in other areas, we have strived to maintain business stability throughout the year, as we've demonstrated so far, and I definitely hope we'll continue to do so. We are cautiously optimistic about growth towards the end of the year, beginning of next one. And what we foresee at the time being is that we're going to see some moderate recovery in 2024, starting with the memory and then gaining momentum with foundry and logic. And also in memory, we're probably going to see some initial recovery with DRAM, followed by 3D NAND at the later part of next year. But we do see some signs of changing of the curve, again, probably towards the end of this year, beginning of next year.

Great. And you guys have talked about dimensional intensity across the 3 areas: metrology, materials as well as devices. So kind of big picture, as we look into gate-all-around transition, I think historically, you've talked about maybe 50% rise in process intensity. Can you walk us through how has your opportunity change on the metrology side going from a FinFET architecture to gate-all-around? And what are your latest thoughts in terms of volume adoption of gate-all-around?

So when we look at gate-all-around, it's actually divided into 3. We look at the thin patterning, we look at gate formation and we look at replacing metal gate. In each and every one of those parts, we've mapped the application space for our entire portfolio. So we have, for example, with our new tools that we have introduced in the last couple of years, essentially, the in-line SIMS and the Raman, the METRION and ELIPSON. We've mapped that in addition to the OCD and to the other tools that we have. And we've identified that in each and every one of those, we have a unique position, whether it's in the position uniformity, whether it's in doping control, whether it is in high aspect ratio, critical dimension measurements. We've identified this position, and it gives us a positioning in gate-all-around, which is going to drive the business, as you mentioned. And the more or the faster proliferation would be for those gate-all-around nodes with the major players, we're going to see higher adoption of our tools, whether it's on the dimensional or the materials. And we're definitely driving also the chemical metrology, not necessarily the gate-all-around, but in the front end in order to have rapid growth and increase of the business there.

Okay. And then one for you, Dror, in terms of your long-term plan, Nova $1 billion plan. And you've talked about a 50% TAM growth. And the question I have is how much of that growth projection is coming from inorganic M&A opportunities where you guys have had pretty good success in the past in terms of adding bolt-on acquisitions? And how much of it is coming from the growth that Gaby talked about from secular drivers like gate-all-around and 3D DRAM eventually.

So we are currently serving a total addressable market of around $2 billion, and it is expected to grow to around $3 billion by the end of 2027. This total addressable market relates to the existing markets that we serve today with the existing 3 divisions. It does not include an addition of -- additional acquisitions, which could really expand this market, depending on the specific acquisition. It is important to note that within this $3 billion, we have approximately $200 million of advanced packaging applications, which were evolving in the last, let's say, a year or so. And it is including in this $3 billion. And the first, I would say, inroad into this market was done with the recent announcement that you just discussed.

Okay. And then can you remind us the profitability assumptions you're baking in that target model for 2027 in terms of gross margins and operating margins. And the reason I ask is because we went through a period of higher cost, freight cost, supply chain disruption because of COVID and all that? And how are you thinking about normalization of those trends?

So in terms of our goals for the 2027 -- year '27 with the $1 billion revenue level, and this is relating to the organic part of this business because, obviously, if we buy a company, it depends on its own financial structure. So for the organic portion of this $1 billion, which is we expect it to be north of $800 million, we expect gross margins to be between 57% and 59%, and operating margins to be between 27% and 31%. The implication of this model is growing from around $4 per share in 2020 -- around 2023 to more than $7 per share in -- once we reach this organic plan. And you are right about the fact that we do see pressure with the inflation and the cost of supply chain in recent years. And I think we are fortunate enough that together with the new technologies that we offer in the market, including software capabilities or sales and products, we are able -- and also additional features and capabilities on existing product platforms that we offer the customers once every 18 months, we are able to successfully mitigate these cost inflations and keep the gross margins elevated at these levels.

Great. Gaby, a question on China. China has been an area of strength for you guys. I think historically, because you guys were able to ship products that are sold outside U.S., but you still have half of your products in U.S. as well. But the question investors have in terms of the strength that we're seeing from mature investments in China, how sustainable these investments are? And are there real demand drivers that are driving these investments? And again, the reason I ask you is because you guys have kind of good insights into what these customers are doing in China as a metrology company.

So Nova has built the infrastructure in China for quite a few years back. We've started early, and we've built a formidable position in China with infrastructure, with coverage, and I think we are now seeing the fruits of that investment. It's not solely the broader production base. We've had a significant market share in China well before the geopolitical aspects came into place. What we do know is that wherever we provide and sell tools into, we have a strong service and application team, which supports these installations and naturally the usage of the tools. So we have a very good understanding of the fact that the tools are being used and the way that they are going to address specific applications. Now we do have a better visibility into the business in China going into 2024. And we don't see it -- we don't currently see any relent in the investments. And we definitely see the continuous nominal growth of the business in China. In percentage, we were about 30% last year. I believe that this year, we're going to be a bit higher. And once advanced nodes, investments else, where outside of China will increase, we'll see the percentage probably going a bit lower. And definitely, the percentage between trailing and leading-edge nodes. But we have a strong position in China, and we'll definitely continue and support the business over there.

I mean obviously, you don't control the mix of mature and needing because your customers kind of dictate that mix. But how do you see the mix kind of moving more towards 50-50 if this mature investments are going to sustain for a while impact your hematology intensity? And are there areas within your product portfolio that you think can help offset it if there's a negative impact?

So naturally, metrology intensity grows, the more advanced node you'll have. And the positioning of the tools for different nodes, for different customers, or trailing for advanced is different. We have legacy tools. We have more advanced tools. So we're capable to segment our propositions, accordingly. This year has been a bit slower for advanced. We saw a higher percentage of the trailing. And next year, depending on the continuous investment in China versus recovery in advanced nodes, we're probably going to see a different mix. Obviously, where advanced nodes are going to pick up and intensity is going to grow, the better it is for us. Now another factor that is important in alluding to your question, is the fact that we assume that some of the yield challenges of building new fabs actually require higher process control intensity. So that, in a way, balances the intensity of advanced nodes elsewhere. So we see a good mix and a fairly confident view of the ability to maintain this intensity going forward and even grow it.

I want to ask you about where we are in terms of some of your product adoption. You guys are, again, kind of early in terms of designing in your products because they are needed to look at the bleeding edge devices. So -- and you have a very broad portfolio of X-ray and Raman and spectroscopy and obviously, optical. So on ELIPSON, where we are in terms of adoption of ELIPSON, let's say, on the logic side and the memory side.

So let me answer perhaps with a broader context, which is a lab-to-fab strategy. We have started the lab-to-fab strategy execution around 2015. And we've seen the exponential growth of our material metrology and XPS and XRF businesses ever since. As a result of moving a tool from a lab from an area that requires lengthy cycles of analysis, that also required very trained personnel into moving it in line in an automated way that provides the results in that specific area of measurement and more across the wafer automatically through the host to the SPC chart of the fab, making it faster, more robust and definitely more effective. And tapping on the same concept, we've identified 2 additional technologies that we are currently moving into the fab. One is SIMS and the other one is Raman. Both technologies are in the labs. Now if we take SIMs, as an example, it takes today between days to weeks to break a wafer into coupons, send it to the lab, have a very well-trained PhD-level expert to look at the data and then provide it back to the fab. Now at the time it goes back to the fab, you can imagine that it's not necessarily impacting or it's already missed the train in terms of impacting whatever went downstream. And you can do it in most cases in one area of the wafer. So you broke a coupon, you send it, you analyze, you brought it back with the data. What we can do with the SIMs is basically take the same performance technologically into the fab, cutting the time that is required for the analysis from days and weeks to minutes and hours. We can, therefore, do it since we are not breaking the wafer. We are taking the wafer downstream after the measurement, we can even do it in several areas on the wafer and therefore, provide a broader view of uniformity and the results across the wafer rather than on a single spot. Now we ablate a specific area. And this area, of course, this die cannot be used. But we have proven that contamination wise and the ability to stream -- or to downstream, sorry, the wafer and production is there. So we have a lot of time savings. We can provide more comprehensive data, and we can do it in a much more effective way since we -- again, we incorporate this data, which is analyzed by the tool itself into the SPC chart. So enabling the same concept. This is very powerful. And we've seen adoption with specific customers, naturally its early stages of adoption. It takes quite a lengthy cycle in order to introduce such very disruptive tools into the fab. But we are very encouraged by where we are at today and similarly with a Raman. Now with a Raman, we've seen adoption with some major customers. Again, this is early, but we are working on that adoption not only in logic, also in memory, and we've seen adoption in memory as well. So in both cases, we are relatively at the early stages. So the more we have proliferation of advanced nodes, not necessarily gate-all-around, but advanced nodes, whether they are more advanced in DRAM or 3D NAND or even logic, we're going to see the proliferation and growth of business in both Raman and in-line SIMS.

Great. Dror, back to you on services. It is becoming a bigger piece of your total sales and for most of your peers as well. So when we look at the services piece, how do you see the growth of that piece going to next year? And just longer term, at what rate should this business grow for you guys?

Yes, definitely, service is becoming an important portion of our business. It's very important to note that currently, Nova has more than 5,000 tools operating in the field, meaning, as we speak, right now, more than 5,000 tools are measuring wafers as we speak here and sit here today. And that's in more than 200 sites on a global basis. So that's a huge impact for us on our customers and customer relations. In terms of the business of services, we expect it to grow at a rate of more than 10% a year on the strategic horizon. This is combining 2 factors. One is the growth in the installed base, which is constantly increasing because we are selling hundreds of tools which are installed every year. And obviously, the service business, service spares, service contracts are increasing over time. And the second, the company is investing in what we call value-added services, where we give our customers solutions that can enhance the capabilities of the existing tools. And again, the combination of both of these factors together should create an environment where we expect service business to grow more than 10% a year in the coming years.

And then Dror, on the CapEx plans, I know last year, you started expanding the clean rooms, both in Israel and in the U.S. as well. Can you update us where we are and obviously, the correction came as well. So where do we stand in terms of expansion to get to that $1 billion Nova plan?

So this $1 billion plan is based on our plans to expand in all 3 divisions that we have. We have 3 divisions, which are operating of 3 different territories, headquarters in Israel for the Dimensional division. The Materials division is operating of California, in Fremont in California. And the Chemical division is operating off in Germany. So in all these 3 divisions, we expect significant growth in the coming years in terms of business. Therefore, we are investing in expanding the facilities in all 3 areas. The facility in Israel is already operational and will be, I would say, fully occupied and operational by the end of Q3 this year. The innovation center and clean room in the U.S. is going to be fully operational by the end of this year. And the Germany facility, which is a totally new building in a new area is expected to be operational in 2024. In that aspect, we are increasing our facility investments in both of the years '23 and '24. And together with others, some infrastructure projects that we have, the total investment is around $50 million in both years. This investment is crucial to the execution of the $1 billion plan. But again, we do expect that all, let's say, the building blocks for executing this plan will already be available by the end of next year.

Let me pause here and see if there are any questions in the audience. Yes? [indiscernible] mic.

You mentioned that you have -- in your target model, it's $800 million is inorganic. So I was curious about the inorganic percentage, like part, what is your M&A opportunity -- like our priorities, what are the companies are looking at?

Yes. So in terms of the $1 billion plan, I'll let Gaby allude to the markets. In terms of the $1 billion plan, we expect organically to be able to grow by '27 to more than $800 million a year and the rest should come from inorganic endeavors of M&A.

So obviously, we're looking mostly at the process control areas. It doesn't have to be technologically specific. But for us, a natural expansion within process control would be preferable. But we are not excluding other growth opportunities in adjacent areas within the semiconductor. We have specific criteria in terms of being accretive, et cetera, for this type of one or more acquisitions, and we're actively searching, as we speak.

Maybe a question I often get on you guys is you've outperformed the process diagnosis control market, 300 to 500 basis points over the last 5, 6 years. And the question I get asked is, like, first of all, like what is so unique about your metrology technology portfolio? And second, where is this share coming from? Is it just more of TAM expansion? Or you're taking share from the bigger competitors, smaller competitors?

So it's both. We are growing our TAM, and we are gaining share. And the reason is always value. What we are very proud of and we are investing in mostly is bringing innovation, bringing disruptive innovation, which I personally see as the only possible route into gaining share and of course, increasing TAM. We are also seeing adoption, for example, of our material metrology portfolio, for example, XPS. We are seeing adoption on both more customers and more applications. And those are across the board in foundry, logic and memory. Now this adoption, which started in the past while having a tool per fab, for example, has become multiple tools for fab as the number of applications and utilization grew. And that obviously enables a continuous growth of that business in which we have a very unique position. And the innovation elsewhere across our portfolio has enabled both the share gains as well as venturing out to adjust in segments such as advanced packaging.

And then earlier on, you mentioned that some optimism in terms of memory demand recovering into next year, and we continue to hear about additional production cuts at memory makers as they try to clear out inventory and normalize then. And I'm curious to know if you think the demand level has fallen enough at maintenance levels, where -- it's hard for it to go down further even if you do production cuts or there is more downside to memory demand for you guys?

So I believe that we have reached the bottom in that respect. I don't currently see it going further down. So assuming we've reached the bottom, there's only one way to go. The question is when exactly that would happen. And I definitely hope that we'll see that coming in the first half of next year.

And then on the chemical metrology side, this is a relatively new area for you guys as once you acquired the German company, and I'm curious how the integration process is going? Are there opportunities for revenue synergies across materials and dimension metrology and then chemicals and then just overall observations on that business.

So we're very happy about the integration of the chemical division, the chemical metrology division. We have seen great take-up by customers. And we've also seen that we are able to synergize, especially on the front end. So in the past, most of the share was in the packaging area. And we took upon ourselves to expand, of course, the business while on one hand, defending our market share in the packaging and on the other hand, significantly increased the share in the front end. And we've seen that happening. We're very fortunate in that respect for the last 1.5 years since we've acquired ancosys to see that business grow for us. And we're very excited about the fact that we have a strong position that is also driving growth, especially with high bandwidth memory with a chemical metrology. And we are working on a few fronts in that respect. One is we are tapping into capabilities developed in the other divisions in order to support building an even more unique capability and position for our chemical metrology which is a clear synergy that we have as being part of Nova. And we are working on expanding the portfolio of the division to serve additional markets.

And then the last one on software, Dror can answer it. You talked about software sales around 10% of your total sales. And I think one good thing about metrology inspection equipment is that the attach rate of software is generally higher than the processing equipment. And we hear about TSMs, other foundries trying to use AI machine learning to their advantage and kind of look at the fab a little bit more holistically. And so just curious in terms of how do you think your customers or the suppliers or equipment can use the AI machine learning in a way where they can monetize the value from this? Is the value proposition more on your customer side? Or are the equipment makers are going to get a piece of it for software?

So we have never looked at our machine learning as supporting the fab holistically. We never took this position. We definitely invested in having machine learning, augmenting the capability of our tool to deliver better performance. And in that respect, since this is our core competence, understanding process control, understanding, of course, the capabilities of the tool and developing a deep expertise in machine learning, specifically, we were able to augment the performance -- significantly augment the performance of the tools and therefore, offer better value to our customers. That's where we have focused at. I don't see the customers focusing on developing their own AI machine learning capabilities in this specific area, but rather on understanding how they can look at the entire fab, look at the data that they gather from all of the tools and, of course, leverage on that. But within our tools, within the process control, we definitely have a core competence and ability to deliver a better machine learning solution for that specific purpose.

That's very clear. So they're probably using it to support the EDA tools or move to the next nodes faster.

Correct.

All right. We're almost out of time. Thank you, Gaby, and Dror for coming to the Citi conference.

Thank you so much for having us.

Thank you.