Veeco Instruments Inc. (VECO) Earnings Call Transcript & Summary

Right. We'll get started. Hi. I'm Shane Brett, semiconductor equipment analyst at Morgan Stanley. So today, we have Veeco's John Kiernan, Senior Vice President and CFO. Thanks for having us. For those in the audience that may be unfamiliar with Veeco's story, could you give us kind of a brief overview of the business and kind of Veeco's story?

Sure. And first of all, thanks for hosting us here today, Shane. Very happy to be here at the conference. So yes, so Veeco, we're a capital equipment provider of semiconductor equipment. And the way we look at our business is in 4 markets: the semiconductor market; compound semiconductor market; we have a long history of selling into the disk drive market we refer to as the data storage; and then the science market. And for us, the largest market for us right now and it's about 2/3 of our business is the semiconductor market. And we're pretty happy about what we've been able to do over the last couple of years in this market. So we've outperformed WFE growth for the last 4 years. We've been -- the compound growth rate has been about 30% in that business, going from less than 40% of our business to up to 2/3 of our business. We have a number of different products that we sell into that market. The largest product for us is our laser annealing product line. Laser annealing has -- and for the most part, around 28-nanometer node. And today, we are process tool of record, all leading foundry logic customers with our laser annealing for at least one application. And now we also see opportunity to take that laser annealing technology into DRAM memory as well. So we've -- recently, over the last couple of years, we've penetrated one of the major DRAM manufacturers with our laser annealing technology. And we announced in the fourth quarter of this year after working with another one of the leading DRAM customers, agreement to place an evaluation system, laser annealing evaluation system there for high-bandwidth memory application there. So we see opportunity for growth in laser annealing. It's at the very early stages of being adopted in DRAM memory, and we're pretty excited about that. If I look at our ion beam technology, we sell ion beam technology to make the defect-free EUV mask blank. So we're part of the EUV infrastructure there. So our customers are the mask blank makers. And the way we look at that business is for every 10 to 15 ASML scanners that are sold into the market, there's a requirement for one ion beam system. They have selling prices in excess of $10 million a system. So we categorize that based upon the number of systems being sold into the market today, about a 3 to 5 tool opportunity for us. But we're also seeing some additional opportunities in that space for ion beam equipment as well. We've got tools under evaluation for our ion beam technology to sell into front-end semiconductor to address low-resistance metal films. So that's a new opportunity for us there. We are the world-leading ion beam technology company. We have decades of experience selling our ion beam equipment into the hard disk drive makers. If I then go on to our -- the last piece of our business to highlight in the semiconductor is our advanced packaging exposure. So we have 2 technologies that exposes us principally in the packaging end. We sell wet processing or cleaning equipment. We are qualified as process tool of record at an important [ COS ] step. And we also, for years, have sold lithography equipment into the back-end advanced packaging as well. So Shane, I'll move over to then the next area of our business. It's a market that we serve is in the compound semi market. We principally serve that market with our epitaxial equipment, so MOCVD, and MBE equipment there, and that's an -- also an area of investment for the company. The third market that we sell into, and as I mentioned, we've been in this market for decades, is we are the leading provider of ion beam etch and deposition equipment to the companies like Seagate, Western Digital that make the disk drives. And principally, we are exposed to the manufacturer of the thin film magnetic head for the read/write. And then lastly, in the scientific, we sell a number of different technologies into the scientific and government research markets.

Got it. So I'll start, I guess, with the semiconductor business. You sort of mentioned you outperformed WFE for 4 straight years. Why was that? And then kind of going forward, looking into 2025, what sort of gives you conviction that, look, 2025 would -- should be an outperformance year again?

Yes. So I'll start with the first part of that question. So our technologies that we've invested in have been able to gain share against legacy technology. And this has principally been in the laser annealing, has been the largest driver. So -- and as I mentioned, we've been qualified as process tool of record at advanced logic applications for all the leading logic companies. And essentially, what laser annealing does in the annealing process, we're taking share from a legacy technology, flash-lamp technology where the -- in order to anneal the wafer, it would heat the entire wafer all up at once. A laser anneal could be a more precise anneal where you can heat up higher and cool down quicker, and that's referred to as a thermal budget because we raster a laser across the top of the wafer, only heating that element of the wafer. So as line widths and spaces shrink, and as I said, principally adoption around 28-nanometer node that the thermal budget becomes more and more sensitive and opens up opportunity for us to take annealing steps that have been traditionally done by flash lamp and move it to laser annealing. And we see that opportunity now going forward as well on the memory side, where it's in the very early stages where the line widths and are continuing to shrink and seeing some of the same challenges that logic had. And we can see the ability for that same playbook to get qualified at one customer, move to the next customer, move to the next customer. So very early stages on the memory side. So we see that still as an opportunity because laser annealing is gaining share as a technology to outperform WFE. Now I will say that as a semiconductor company, we're a smaller piece of WFE, right? We're less than $500 million of business today in a $100 billion market. So as these technologies become adopted, it does give us an opportunity to outperform WFE. The second part of your question, Shane, there with respect to what can we outperform WFE again in 2025. Now we've not made a call for full year 2025 at this point in time, but we see 2 dynamics in the marketplace right now. In the one sense, we see headwinds coming from China that's been investing quite dramatically over the last couple of years, and we were exposed to that. So our expectation is that our China business, similar to peer companies, will be down in 2025. We have good visibility for the first half of 2025, and we expect the China business to be about 25% to 30%. And we're less certain about the second half of the year because we're not seeing right now the pipeline for new investments in new fabs at the pace that we saw this time last year. So we see that we're calling a falloff in that business in the second half of the year. So that's the one side. On the other side of it, we see some really good tailwinds coming from 2 pieces that we have good exposure to. First area is exposure to gate-all-around. So we started to see business in 2024 for early pilot production there. It was around a $50 million business for us last year in 2024. We see the opportunity for that business to double this year to about $100 million. So that's a good tailwind for us. And the second tailwind we see is we have exposure in advanced packaging and principally with our wet processing equipment, where we see opportunity for that business to also double in 2025. And that was about a $75 million business for us in 2024, going to about $150 million business in 2025. And the question really is, will those growth areas more than offset the potential for China downside. And I think it's just a bit early to call at this point for the full year. And if we see the opportunity for growth, we see an opportunity to perform well against WFE, but a bit too early for us to call at this point, Shane.

Got it. So I'll start on kind of the foundry logic portion. So you mentioned China down versus sort of GAA, advanced packaging up. But just on the China down portion then, can you walk us through sort of what you saw from these China customers over the last 2 years? Is it sort of the breadth of the customers increasing or the kind of specific customers increasing investment? And then could you kind of talk about if there was any impact from the regulations in December as well as sort of the revenue level we saw in '24 and how you expect to sort of kind of trickle over in '25 and '26?

Yes. So on the first part of the question, what drove our business in China is principally on the semiconductor side, although we have business in compound semi and scientific. Really no business in the data storage, they don't have operations in China there. So 3 of the 4 markets with exposure to China, but the largest piece of the exposure coming in the semiconductor market. A principal part of the growth for us in the semiconductor market over the last couple of years is really from the laser annealing being adopted in these mature node processes. And I would say a good portion was investments in new fabs for 28- and 40-nanometer process in China was the biggest driver there of the growth. Second part of the question with regard to regulations. What we came out and said is that at the time the new regulations came out, we didn't have -- it didn't have a very large impact to Veeco at that time. Essentially, we didn't have any systems backlog with customers that were newly put on the entity list, so none of our backlog at that point really got impacted, and there really were not new requirements for our products in terms of licensing that didn't already have licensing requirements. And some of our products do have licensing requirements into China and others don't. Our laser annealing does not have a licensing requirement going into China, and that's our largest product line. So from that perspective, in the very near term, it didn't have much of an impact. What the impact could be in the future? Could these have been potential customers or buying equipment for companies that have been put on the entity list? Could there be changes, further changes in regulations for equipment to make them more restrictive? That's to be determined. But in the very near term, not a very large impact to our China business.

For sort of year-over-year decline, you guys have kind of guided for the China business. Is this sort of a digestion of all these tools that these customers purchased in '24? Is it more sort of like a consolidation of customers?

Yes. I think it's really 2 factors that we see. First factor is, yes, the customers that have been buying for fabs that have -- the projects have gone forward, and they've received their investments. They took in equipment, and they've populated those fabs. But I think the bigger issue is that we don't see the number of new fabs or new funding as you've seen in the last couple of years and particularly in these 28- and 40-nanometer type fabs that is a sweet spot for our equipment.

And just one last question on China, and I want to talk about GAA next, is just on China, who are your sort of main competitors in the field? And are there any kind of local players that we should probably look out for going forward? Or should we focus it as Veeco market?

Yes. So we have one principal competitor for our laser annealing product that also has a laser annealing product. So they're a U.S.-based, large capital equipment company. We don't like to give too much airtime to our competitors there. And then what we have not seen is very much local competition. We're not aware of local competition being qualified for laser annealing. And of course, in China. And of course, their move to sell sufficiency in capital equipment surely are trying to gain capabilities in laser annealing. But to date, it's not been very successful, and we're not aware of any tools that have been qualified.

Got it. Got it. And just I'd like to move on to GAA. So I think you talked about GAA doubling year-over-year in '25. It's pretty consistent with kind of what we hear from sort of large cap equipment makers. But could you provide a bit more color on how to sort of think about the revenue cadence, what the revenue cadence looks like through '24 and how we should probably think about it through '25 and maybe going into '26 as well?

Yes. So the $50 million of revenue that we have in gate-all-around in 2024 came more in the second half of the year as customers started populating for their pilot lines. And we see that continuing into 2025. I don't have sort of an exact cadence here, but the expectation would be it'd probably be fairly consistent throughout the year, maybe a little bit more in the second half of the year.

Got it. And I'd like to move on to DRAM. Could you talk about your -- sort of DRAM exposure? And just on HBM, is your business more so driven by increased DRAM -- leading-edge DRAM capacity due to the die discount? Or is the sort of exposure that you guys have to specific DRAM -- specific HBM steps?

Yes. So we're at the very early stages of adoption here. We've been adopted by one leading DRAM customer for their HBM. So as I said earlier, we're pretty excited about the opportunity going forward, and we're excited about coming to an agreement with a second DRAM customer for our laser annealing technology to place an evaluation system midyear in 2025. And we continue to work with the third for us, DRAM customer opportunity to have laser annealing adopted there. So for us, it's been an area of growth. The one customer that we have, they've been taking a handful of tools since 2023 into 2024. And our expectation is for a similar level in 2025. And then the next opportunities will come from penetrating the other customers. Now what we have been qualified for initially is in the stack is the logic die at the bottom of the stack. And more recently for this customer, the peripheral logic in each one of the HBM die there. So we see this as an opportunity. We're at the very early stages of penetration there. And that we see, as we look out a number of years, that's the opportunity for us. So historically, we've been very tied to foundry logic, and that's been the much larger portion of the business. But we see the opportunity -- if we look out a number of years, our opportunity could be as large in DRAM as it is in foundry logic.

Okay. Interesting. And then you spoke about this AI-related revenue to grow from approximately 10% in '24 to 20% or more in '25. Can you break down what you categorize in the sort of AI revenue bucket? And sort of what are the drivers from getting to -- from getting from 10% in '24 to 20% plus?

Sure. So we've got a number of different technologies that I already described that touch sort of an AI chip, so to speak. So if we look at the logic, so the GPU, we have exposure where our laser annealing equipment is qualified for that process to make the GPU as well as our EUV mask blank tools. So some of these processes in the GPU and the HBM die use EUV processes. So therefore, that's an opportunity for some of the equipment that we sell to the mask blank makers, they're using for EUV. So that's the one area that we have exposure to. We also have exposure on the packaging, as we talked about. So we have -- we're a process tool of record [ in a COS ] step for the micro-bump flux clean, and we also are qualified for a lithography process as well. So when we look at that, we have processes on the front end of the logic, we have processes in the DRAM stack itself and we have process -- we have exposure in the packaging of the device.

So I'd assume you've guided for advanced packaging business to double from $75 million to $150 million, but that would probably be kind of categorized in the AI bucket.

Some of it, but not all of it. Exactly.

Got it.

So we don't want to double count there. So what we said is -- we had this question from investors, like how are you exposed to AI? So we put a cartoon slide in our investor package of showing a typical AI chip and saying, here are the sort of technologies that are applicable here as I just described. But then you get asked the question, hey, what's your exposure to high-bandwidth memory? What's your exposure to gate-all-around? Well, also -- well, those also are process steps within that AI chip. So we don't want to give the indication that -- a doubling of a doubling. So we see our exposure to AI doubling, and that's largely because of 2-nanometer gate-all-around and largely because of advanced packaging. I hope that makes sense.

Yes, yes. So just talk on the sort of advanced packaging portion that's doubling by itself, the $75 million to $150 million. But could you just sort of elaborate on the customer diversification strategy you have? And if you could, the kind of split between web processing and the lithography contributions.

Sure. So I'll start with the second part of that first. So it's about 2/3 web processing and about 1/3 lithography. And the growth that we see in the web processing is we are qualified at foundry. We're qualified at OSATs. We're qualified at the HBM, but the biggest part of the driver is coming from foundry business.

Got it. And just on the SAM growth as well. So you've spoken about the $350 million SAM in '25 to $650 million in 2029. Could you sort of talk about some of the assumptions behind the SAM expansion as well as sort of what you envision your market share could be in '29?

Yes. So we see opportunities now to be on this advanced packaging road map for our wet processing technology. And the growth in the -- in our SAM opportunities is in the wet processing side more than the litho side. And as these processes continue to evolve -- so if you think of things like hybrid bonding, you think about things like PR strip, wafer thinning and so on -- we have a number of different opportunities that we're working with customers on that help drive that advanced packaging road map. So a lot of the business is being driven today by new [ COS ] opportunities. But in the future, we see new opportunities that are on the customers' road map to be able to solve those [ technologies ].

Got it. I have 2 more questions, and I'll open up to the audience, but one on gallium nitride and then I have another on data storage revenue. Just on gallium nitride. So at your latest earnings, you talked about shipping a 300-millimeter GaN on silicon evaluation system to a Tier 1 power device customer. You also mentioned that you have received positive feedback. So what I want to know is, can you kind of provide -- talk about sort of the expectations for this gallium nitride market. And sort of how is your kind of customer geographic diversification that you hope to achieve in the future?

Yes. So we're pretty excited about this opportunity. So today, a good portion of power devices using gallium nitride on silicon is at 6-inch wafer size manufacturing, transitioning to sort of 8-inch. And you have a leading device maker that's looking to go directly to 300-millimeter GaN on silicon, single wafer. And that's a pretty disruptive technology there. And to be able -- this is a company today that has a silicon-based power -- large silicon-based power business and potentially looking to take that product and move a portion of their product line over to gallium nitride or GaN on silicon. And then to be able to use their 300-millimeter CMOS process there. So the opportunity is quite large. It could be more than 100 reactors of MOCVD type of equipment over a few year period of time. So the opportunity is very large. So there, we're in evaluation. Our equipment is being evaluated. There's also a competitor, so this is a head-to-head evaluation. Process, we're getting good feedback from the customer. If things go well, they would be looking to maybe start pilot line production sometime in 2026 and if successful, moving on to high-volume manufacturing at a later date. Now we're also hearing and talking to a number of other leaders in the power space around gallium nitride on silicon for 300-millimeter processing. So it seems to be picking up steam with a number of different customers at this point. So we're pretty excited about that. And then geographic diversification there, you would see that most of these -- the companies that would be investing in this area, European and North America based.

Got it. Got it. And then last question for me is the $60 million to $70 million decline in data storage revenue for '25. Just how should we think about the quarterly progression of this decline? And is there anything you can kind of comment on beyond '25 for how we should think about revenue for this business?

Yes. So we -- as I said earlier, we've been a provider of deposition and etch equipment to this space for decades, and so we have very good and long-term relationships with the customers here. Over the last couple of years, customers have taken down their utilization of their equipment. It's been a difficult market for the last couple of years. But the market is improving, and customer utilizations are coming up at this point in time. And generally, they'd look to add more capacity when they're consistently operating at 90% sort of utilization or more. So for this year, customers are not planning on adding system capacity. We work with these customers generally on like a 1-year lead time. We don't build to a forecast for these customers, so we generally have good visibility a full year out. So the progression of this year's business is really just the service and aftermarket business for us. And we see that had picked up in the second half of last year. It's off to a good start at the beginning of this year. So our view is that we should see growth in the servicing part of the business because they're bringing more tools back online, more spare part usage, et cetera, there. But clear expectation, no systems business this year. Our expectation is if they are at these 90% utilization rates, and we've been talking to the customers, they'll start to look to add capacity again likely in the 2026 time frame. So I would say by the second half of the year, we start to get a really good view of what their investment plans are for adding capacity in 2026 and going forward. Good news for us, there could only be growth for us in '26 given that we don't have any systems business for this scheduled in 2025 at this point.

Got it. Do we have any questions from the audience? Lee?

Sure, [indiscernible]. A couple of questions if I could. I thought the GaN silicon was quite interesting actually and the pivot to 300 mill, as you mentioned. So the bake-off between yourselves and the other MOCVD player, is -- does this come down to single versus multi-batch? Or is it the showerhead tech? I mean how does it...

The latter. So their tool in this bake-off is also a single wafer. So this would be their first single wafer, 300-millimeter tool. And you're correct, their exposure or their installed base is principally with multi-wafer batch tools. But for 300-millimeter wafer, it will be a single wafer. And we have a couple of tools in the market. We've had single wafer 300-millimeter. We've done a couple of early R&D projects where we've shipped tools before into the market.

Got you. Okay. And maybe just going back to the lithography, the front-end lithography supply of the ion beam technology to the mask guys. The -- is there a step difference between DUV and EUV? Always difficult for a Scottish accent to say that.

That's a great question because we have no exposure to DUV. So DUV uses PVD to do that deposition step. But PVD couldn't meet the requirements to do the steps for EUV. So ion beam is used for 100% of EUV mask blanks, and we are the sole provider today of the equipment for that space. So we did go sort of head-to-head. The defect and the requirements for the EUV mask blanks are much more stringent than the DUV, where PVD is the technology of choice. So it's interesting that for us, we have 2 use cases where the potential deposition technology was PVD, and this is making the thin film magnetic heads. And the thin film magnetic heads are all made with ion beam technology, and the EUV mask blanks are all used with ion beam technology. And we're trying to make ion beam technology the fourth deposition technology in the front-end fab. And the principal competing technology that we will compete against is PVD.

What about gallium nitride opportunity? You mentioned American customers and European customers, but there's one big Chinese IPO in December. Would that sort of -- would that customer be a possible opportunity or not really?

No. And I don't even, quite honestly, if -- yes, I don't think it's where we would direct our initial entrance. And then there could be some questions about licensing requirements there. So our focus for this product is with the Western power companies not the Chinese power company.

Got it. Is there any other question? Nigel?

I had a question about your advanced packaging business. I can't do the math on the top of my head, but you've identified the growth vectors. I think you mentioned wet processing, flux clean. Maybe first question is, I'm hearing that there's a move towards sort of fluxless thermal compression bonding, is that a headwind or is that some -- an entirely different one?

It's not a headwind in 2025 and into the beginning of 2026 from sort of our understanding. And as I mentioned, we have programs with the customers to look at our equipment for the fluxless steps as well. So that is an opportunity for us going forward. But our understanding is the build-out for the current process will continue throughout '25 and into '26.

Yes, that makes sense. And maybe also a question about time lines then. You've mentioned hybrid bonding, PR strip, et cetera, what is the time line for that opportunity for you?

Yes. I think that would be like sort of further out, probably more towards end of '26 and into '27 and beyond. So I think the -- what we see is the opportunity in '25 and the early '26 is that, that would be further out.

Maybe to round it up, I mean, if you then talk about your advanced packaging opportunity, what is the main driver sort of? Or is it more that the one comes as the other goes and you sort of build on that?

Yes. I think that's the idea is to work into like the road map and work with the customer from sort of device to device or process change to process change. So we've been successful in that in the front end of the process with our laser annealing equipment, 28-nanometer FinFET going down all the way to 3-nanometer and then 2-nanometer, gate-all-around and working with customers on 18 Angstrom and below at this point that we would look at advanced packaging now to start to be on this road map that we can work with the customers for multiyear road maps. So for products that are N+1 and N+2, similar to what we've been able to do with customer in the front end.

So I guess I'll just wrap it up with one last question. So annealing SAM, you guys have given that $1.3 billion number, ion beam deposition, you've given from the $50 million to $350 million. What are the big inflections in the semiconductor world that we should see that will sort of drive Veeco's SAM to sort of have that [indiscernible]?

Yes. On the annealing SAM, 2 things, new product for our nanosecond annealing, which opens up new opportunities for more shallow anneal. So the annealing process with a nanosecond anneal could heat up much higher temperatures, cool down much quicker and can do anneals that don't penetrate very far into the wafer. So offers up opportunities for things like backside power distribution, opens up things for 3D materials there as well and material modification. Second area of driving opportunity for SAM expansion is in memory, as we've talked about with both the laser annealing and eventually, the nanosecond annealing technology as well.

Got it. Well, thanks very much, John. I appreciate you coming. And thanks to the audience as well.

Well, thank you, Shane, and thanks for having us.