Onto Innovation Inc. (ONTO) Earnings Call Transcript & Summary

Good afternoon everyone. Thanks for joining us today. My name is Elizabeth Sun and I'm with the semiconductor equipment team here at Citi Research. And in this session, it's my honor to have Mike Plisinski, CEO of Onto. Welcome, Mike.

Thank you.

I have a list of prepared questions. And -- but for those in the audience, if you have any questions, feel free to raise your hand any time. So first, Mike, I'd like to start with a high-level question about the market outlook. What's your view on the end market growth into next year across various segments you are in, both on the front end and back end? It's a big question.

Everybody always wants to know that. So I don't -- I think it's too early for us to have a very informed view of 2026 right now. We're still talking to customers. We're actually starting the process of engaging with customers on their expansion plans for next year. They're talking to their customers, what demand drivers look like. So we'll know more towards the end of the year. But for us, we've never really that obsessed with where is WFE going and where that overall market is going. Our focus is really on more, what are the underlying waves of growth that are happening within the market and how do we get more and more products onto those waves. So with that, I think we're positioned pretty well for 2026. And that's in both new products as well as the markets we're serving. So AI is continuing to show no signs of abatement. It's still growing. The expanding technology, both on the HBM -- both on the memory and the logic side is creating new opportunities, new capital intensity demands. We talked about some of them in the last earnings call, subsurface inspection, et cetera. But then we have new products as well. So new products in 3D metrology that are moving well through qualification at large accounts and also being already adopted by smaller accounts. We mentioned OSATs last earnings call and co-packaged optics as well. And then, of course, the critical films, big, huge market, $500 million market, making very good progress moving the Iris G2 through that. And the other one for next year is, of course, the Dragonfly G5, the next-generation high-resolution tool.

I'll ask more on that later. I guess next year, you are well on track to maybe outperform the market wherever the market is.

That's our view.

Okay. Great. And on the leading edge, I just want to ask what's your visibility now because some of your large OEM peers are seeing some seasonality among nodes in maybe second half. But it seems like you guys saw some like adjustment in Q2 and maybe to some extent in Q3 as well, but you are expecting Q4 to be back to peak levels. So just could you help understand the disconnect here?

So I'm not sure the disconnect per se. So if you look at our growth this year in advanced nodes, it's nearly 2x, nearly 100%. I don't think any peer is talking about 100% growth. So it's really hard to say peer-to-peer comparison when we're significantly faster. And why are we so much faster? The market is not clearly growing that much. But it goes back to what I talked about earlier, the different products we put on the wave. So as DRAM expansion was required because of the HBM, we weren't just getting OCD revenue anymore. We were getting films business. We were getting integrated metrology. So we added more to that wave of growth. Similar to another example would be the new gate-all-around customer we added. Instead of just an OCD order, we significantly increased the size of just that small pilot line by adding the films business as well. So I think when you look at it from -- we're not just talking about growth due to market -- end market demands, but also what we're doing to enable the revenue, and that's why we're seeing this kind of an outperformance. The -- for sure, the revenue in advanced nodes for us was second -- first half weighted. We saw the biggest growth there. We always predicted the Q3 drop. We talked about that early on. And we still expect some kind of a Q4 improvement, not as big as it was, but AI packaging is now bigger than we expected for Q4. So net-net, about the same.

Okay. Great. And we hear some people -- some of the international peers are talking about maybe 3-nanometer can be larger -- like 2 nanometers next year might be smaller than 3-nanometer investment going into next year. Do you agree with this view?

2-nanometer can be smaller than 3-nanometer. Meaning FinFET versus gate-all-around?

Yes.

Not from what we see. So I'm not sure where that comes from, but at least our customers and TSMC has been public about this. They see a very long and sustained demand for the gate-all-around node. And I think the performance benefits from the -- both the speed and the power consumption, the reduced power consumption of the gate-all-around transistor versus FinFET are pretty compelling. So I think once certain level of critical mass starts adopting it, the rest have to follow or risk being left behind.

Okay. That's fair. And on the packaging side, now there is a large competitor in the market and gave you some pressure in the 2.5D side. And then there is a smaller competitor on the front-end who has been also very vocal about gaining share in the packaging side. So I guess first question was, what is Onto's competitive advantage here? How do you maintain your leadership position in the back end? Is there any like pricing pressure you're seeing on that side, too?

No. As far as pricing goes, it's the opposite with the entrance barely of that one competitor, we're seeing an opportunity to increase price. So as the -- and it's really not so much them. As that customer required much more advanced inspection. It's a chance for us to move a tool that we've been developing over the last 3 to 4 years for front-end applications quickly adopted for the back end. So that's -- that gives us an opportunity to significantly increase pricing. The -- and the value associated with that. From the other side, the other front-end company you mentioned. I would say, I'm not sure about the share. I think what we're seeing is more the pie is growing. There's a lot of different applications in -- that customers need resolved and solved, and they're solving some. So we don't see any share -- you worded a share loss. We don't see any share loss to that second competitor. So we're continuing to grow our films metrology business. We're continuing to add additional metrology capabilities, the subsurface inspection. We're continuing to grow. We do know that they've solved some unique requirements in TSV and maybe some others that I'm not even aware about, based on some of the capabilities of their sensors that maybe we don't have. So I think that just shows that the complexity and the importance of the market.

Okay. And I guess I have to ask what gives you the confidence to gaining share back next year on the twin hub packaging?

So I think we've had a strong market share for a very long time. We'll continue to develop that strong market -- the technologies that the customers need to maintain that strong market share. We are with G5. We clearly demonstrated that in the early part of this year, where the customer is very close into wanting to decide one way or another and then deciding, please, pull in the G5, that's the tool we really want. So that's one piece of confidence. But the other is just the performance. We're seeing the application studies that we're getting, so the wafers we're getting and customers want to see results. The results are impressive. And the feedback we're getting is very impressive. So the tool wasn't designed for just solving this one issue, but several generations ahead, and that performance is quite frankly, compelling. There's also some new capabilities in there that we're also now starting to see emerging applications and hybrid bonding that could benefit from those new capabilities. So I think we have a very compelling story on this tool.

It's good to know. And could you share what is a -- you provide -- you said you're passing some milestones in our last earnings call to get like requalified, I would say, with your key customer. So can you share what is the latest? And when do you expect shipments to ramp for that customer?

So the really nothing changed. Between the Q2 and the Q3 call, really nothing changed. We said when we announced the share situation that the customer was already engaged with us sending wafers and working with us on developing and accelerating the new tool. And then the next quarter, I just said, yes, and it worked. I never had any doubt it wasn't going to work, or I wouldn't have ever mentioned it. So really, all we did was continue to execute along the plans and we are going to continue to execute. And I'd mentioned, I guess the one big piece of real news from last quarter is we said we'd ship one tool, and now we're going to ship a few tools to several different customers. So clearly, demand for the technology is growing.

Okay. Good. And you touched a little bit earlier about what are the newer opportunities for next-gen Dragonfly, anything specifically? You talked about hybrid bonding earlier but anything else?

We talked about hybrid bonding because there's going to be -- and we've known this for a while working with our customers. There's a real high amount of concern over voids in the hybrid bonds. And even I think I talked about last -- maybe last year, even just 1 micron void when the -- between the wafer stack or dye stack, the next stack on it with that compression that 1 micron void makes a crack that pops through the entire dye stack sideways. So that wipes out the whole stack. So it's a very big concern. So what causes those voids, 2 main areas they're focused on. One, the flatness of the pads. So if it's like a cup and you push, you have void. It's not totally flat. The other is very, very tiny particles. So even a small particle can potentially cause a void. So that G5 will have the ability to inspect for those types of particles. So that creates a new opportunity for us and it's very high resolution. So it's a relatively slow inspection, and we think it's going to be high sampling. So maybe 100%, but certainly high sampling. So that creates new opportunities. What was the first part of your question?

Any other opportunities?

On the other -- yes, I know there's one more point I wanted to make. It was the front end. Of course, there's the whole front-end opportunity, which is what the original intent of this. The front end market for macro inspection is probably around $400 million, $500 million. It's served only by 1 competitor. And the customer feedback is pretty positive about the capabilities we're providing, not just with the G5, but other sensors we're combining in there to enter into that market. So that's another area. It will take some time to develop. So I would say in 2026, we should see some movement and then real revenue maybe in 2027. But by movement, I mean initial orders.

Okay. Got it. And what is your expectation on hybrid bonding on the timing? It's a moving target I know.

It's a moving target. I'm getting the sense that the customers will -- I might be wrong and Sydney yells at me later. But getting a sense of customers will delay this transition for as long as they can. They keep pushing it out, pushing the yields, pushing the capability of the existing technology before making that cut to the hybrid bonding. So who knows what innovations they have in the pipeline. But for sure, hybrid bonding is on their road maps. For sure, everyone's get investments, meaning all the 3 main memory guys have investments to be ready, but I also see them continuing to push and push and push.

Is the main concern, the cost or the yield or like supply chain...

I think they're tied. So the cost and the yield, if the yield is low, the cost becomes much higher. So I would say they're somewhat tied. And yes, it's the yield.

Okay. Got it. And on the HBM side, it looks like the third HBM customer is finally getting some traction. So could you talk about your market share position across all 3 HBM customers? And if the third customer get qualified, how does that mean for Onto?

So it's not just Onto, so if the third customer gets qualified, the industry, right, so now qualified -- the second part of that is also how much allocation do they get. So if they get allocated a small amount and the other 2 have to increase, then that's good for everyone. If they get allocated a large amount to take up all of their excess capacity, then the other 2 have to expand just a small amount. And because they have excess capacity, not just from us but from -- they had lines ready to go, that will kind of be a bit of a digestion period for the industry.

Yes, that's fair. And what is your share position with all 3?

It's very similar across. So extremely strong, extremely high with 2D inspection and activities in all 3 with 3D inspection, 3D metrology, so I should say. And it's the same with some of the other technologies, subsurface inspection, things like this.

I'll pause here to see if there's questions in the audience. Okay. I'll keep going On the new products, I know you'll be happy to talk about it. So can you talk about the progress you are making with the new products, especially with the 3Di metrology, but maybe others you want to share as well?

So for 3Di metrology, we talked about having 10 customers now already adopting the technology, including new applications and co-packaged optics, which we think is another future wave, fairly meaningful wave of growth, given the extreme growth in AI and the power problems tied to AI. So I think that's going to be a big one. And the OSAT, so a Tier 1 OSAT for bump applications. The big thing, the other area that we're focused on is qualifying for some of the Tier 1 semiconductor manufacturers. So particularly the memory guys. So there, the process is much more rigorous. It's taking time, and we're working through it. By time, it's really around stability testing now. So making sure these metrology tools can measure the same way every time for a month at a time. And any deviation, we have to restart the clock. So we make some changes, improvements, and then the clock restarts. So that's taken some time, but I think we're moving pretty well through that process, and we should be seeing light at the end of that tunnel. And then the question is, okay, well, how quickly will they adopt the new technology moving into next year? For sure, the performance is proving out what we expect and what the customers were hoping for. So it's definitely more sensitive, it's definitely more repeatable. It's faster than the incumbent technology. We also are opening up new applications. So in the bumping area -- it's a long story. Do you want me tell?

Yes. Go ahead.

All right. So in the bumping area, there's -- most of this metrology happens after solder bump, solder caps are put on the copper pillars. That's not an ideal step, but it's because of the existing technology, that's the only way it was able to measure. Otherwise, there's too much scattering on the top surface of the copper pillar bumps. We're showing in partnership with an OEM that also selected our tool early on. We're showing that we can measure actually that pre-reflow step, and there's a big yield benefit to customers if they shift when they do the measurement. So they can do more rework and save yield. So that's a fundamental improvement that customers can gain from the tool as well.

It's good to know. And I want to ask about the Semilab acquisition. Could you just add more color around the pending acquisition? And what is the market you're after for this acquisition? And any early thoughts on the synergy opportunities? I know you are like probably waiting for the deal to be closed, but anything you can share right now.

Yes. Synergies will leave to the side, at least the quantification. I think there's some synergies we talked about with our Ai Diffract, so our modeling software, being able to enhance the capability of some of their tools, so adding more value to our customers. The big advantage, though, is that their technologies are applicable to a wide range of markets. So it wasn't very a market-specific deal. It was really us bringing in some new and unique technology that we felt we could add, we could further enhance and we can enhance them in 1 -- in 2 ways. One, our channels to market. So we see opportunities. We talked about this on the call in the surface charge metrology, which is going to be a growing concern for -- as chiplet architectures continue to migrate and become more widely adopted in the marketplace by designers. That's going to be a critical source of yield loss. There's no other way to measure it, except this tool. So that's going to be important. And we have good channels to market to help enable that transition. The other is on the materials characterization side, and that's where we can add some enhanced capability, also potentially with some of the other metrology systems that we have as well. So not just the software but also some technology.

What do you see -- how do you see the market of materials characterization or materials metrology? Because I think it's kind of a smaller market, but it's very dominant by another player in this space. So how do you see the market growth going forward?

I think there's a lot of -- so it's a broad term, and I think there's a lot of growing interest from the customers as the use of exotic materials and more materials in the advanced nodes and advanced packaging continue to add complexity. It's not just about dimensional metrology, it's about how these materials are interfacing with each other. And there is one player who's doing well with X-ray systems in that space. But it doesn't solve all the materials characterization challenges that customers have. We have one little piece with the element, but that's only composition. That's only one piece. So customers are looking for more solutions. This team has some interesting technology. They don't have the service and support infrastructure, not as broadly as we do specific to these high-volume manufacturers. So we think that, that will be -- eliminate one barrier to a more wide adoption of this type of technology. So it's an area that we need to continue to look at.

Okay. Great. I guess, next...

I do think it's an area that's growing.

Definitely. And I guess I have to ask about panel-level packaging. So what is your latest expectation for panel substrates? And is there a moving target as well? Like how does it compare with what your expectation was maybe last year or like 2 years ago when you held the Investor Day?

I would say it's somewhat of a moving target, but really tied to the enterprise. So if you look at enterprise server market, that hasn't really recovered yet. And that was -- the AMDs, the Intels, they're heavy users of the panel substrate technology. So they pushed the substrate manufacturers to increase capacity quite a bit. We've delivered a lot of tools and the utilization is still -- so I think that will recover. That will return. But the other big inflection that's happening, and we always said it was unclear when the transition would be is the move to glass. So that we see continuing to build momentum. So it's not like stalled out. It's not like just a few research. We see more and more players. Our apps lab is completely the pace, the application center of excellence we built for that, very full with engagements from customers, many, many customers. So I think that area is still -- it's still early, but also very transformative. So when you think about being able to drive -- if you're limited to about 5 microns or so on an advanced IC substrate, being able to take that down to 1 micron and being able to put all the RDL on a single surface has much better thermal capabilities than the IC substrates, there's a lot of compelling advantages to glass. Challenges, technical challenges, but those are being worked through.

Yes. That's fair. And on gross margin, you are on track to achieve $1 billion in revenues this year, maybe not far from that. And then it seems like gross margin is a bit below your long-term target model of I believe it's 56%, 57%. So -- could you talk about puts and takes there? Like how should we think about -- how do you improve the profitability going forward?

You hire a new CFO. Seeing if anyone is awake. So several areas. We have been behind on improving our margin profile. And I think we are taking steps to address that. But beyond a little bit of a joke is the moves we're making on the supply chain. We've talked about common EFEM and bringing that. So we had 7 or 8 different EFEMs across the organization. That's a huge inefficiency and it also limits our ability to extract value from suppliers. By moving to a common EFEM across all of our platforms, that's a big savings. The shift to overseas manufacturing and the ability now for us to have far better business continuity, but also a supply chain based in Asia, several areas in Asia. That's also going to provide help. And then just better, let's say, operational discipline, better efficiencies. Back to the discipline we had in the past. And I think we'll see -- certainly the long-term operating model is well within our reach, if not even improving on it. The other piece is also, some of these new products that we're bringing to market, the G5, the Iris critical films, they will all be margin accretive. They're significantly more advanced than the prior generations. And Semilab will be margin accretive. So that's another one. So yes, I think the margin story will be a positive.

Okay. And if you look at Onto over the next 3 to 5 years, what kind of key technology transitions you're watching closely? And what is the most exciting driver for your business?

That's a good question. So I think from a technology transition for us, we're very careful not to get obsessed with any particular transition. For instance, we won CoWos 7, 8 years ago, went nowhere. If we are obsessed with that, okay, we'd wait 7, 8 years, then we'd finally be heroes. But no, then we were on to InFO and we're onto the mobile and the RF, 5G transition. So for us, it's -- and now, of course, AI packaging explodes, and we're right there. We're on top of it. So for us, it's more about not trying to pick and watch the winners and losers, but make sure we continue our strategy of expanding our footprint across the value chain and the depth of our footprint in each of those accounts. So that's the beauty of the types of products we produce. The macro inspection is broadly applicable. The films is broadly applicable. And then we have various additional capabilities in materials characterization, other sensors that are also broadly applicable to power and front end and packaging markets. So then when the different waves of growth occur, we're right there, we're on them and we ride them. Power semi is another great example. Pretty flat for most people, but we've been having very strong growth in power semi the last several years. This year is kind of flattish, but off of a record year last year. So that's why we don't obsess over 1 or 2 particular things. But not to be facetious and to answer your question a little more specifically, packaging is clearly an area that we're excited about. We see so much investment happening in packaging, finally, an area that we have a very strong position in. And it's not just in, let's say, the traditional shrinks that you see in the front end, but we're seeing the 3D expansion, but we're also seeing the heterogeneous -- the adoption of heterogeneous dye and the complexities associated with that as well as the move to panel. So if you have heterogeneous dye, you can't process it very effectively on a round silicon wafer. So the shift to panel is there as well. So a lot of new challenges in the packaging area tied to many different markets. So whatever end market drivers you're looking at, enterprise server, enterprise market, AI, power semi it's all going to tie into these packaging advances.

That's a great answer. On capital allocation, after Semilab, what's your priorities on capital allocation? Are you doing more M&As or...

Yes. I think so from a capital return perspective and from an increasing shareholder value perspective for a company of our size, for the markets that we're serving, we still think M&A is the best source of return, the best allocation for the capital. But of course, every quarter, as we've said to investors, we're looking at the M&A pipeline, comparing it to the stock and the opportunities, and we're always making the decision, where best to deploy that capital. We're a strong cash generator. So we're lucky that we continue to increase our capital so that we have the ability to deploy it in the way that's most advantageous for the shareholders.

How do you look at M&A opportunities? What is the criteria?

Yes. A lot of investors look for hurdle rates or particular numbers that we're trying to hit. I think that's too limiting. So for us, we look at first the -- essentially the synergies we think we can extract from it. And the closer to the center of our bullseye, which is going to be around the broad markets we serve and the software and the optics capabilities. So that -- those are our core competencies. To the extent we expand on those core competencies, we strengthen all the businesses, not just the 1 business we might be acquiring. So that's the strategic nature of it. From a financial nature, Obviously, we're not here to dilute our margins. So there's many deals that we could do that we walk away from just because they're too dilutive, and we don't see enough synergies to actually improve. So there's not a good fit. The Semilab is a great example of a company that we worked with and we had to cultivate and it wasn't one that was obvious to people. But that's an example of why we kept telling investors. We're working the pipeline. Pipeline is getting better and stronger, and it is. Our M&A pipeline, now that we focused on it, I guess, starting 3 years ago, is pretty robust. And it's not through the run of the mill, everybody knows the same names.

That's fair. And then we have a couple of minutes left. So just any final remarks that you think investors are missing in the Onto story?

Final remarks. Well, I think there was an overreaction to our ability to compete. So I think that's a little hard for me to understand because that competition there from all those sides that you mentioned, has always been there. That's the nature of competition. And we've always been able to maintain it. So I think as we look forward and as we continue to drive the G5 adoption and demonstrate the capabilities that we'll be back in the good graces.

Okay. I like that. All right. Thanks, Mike, for coming to our conference. Thanks, everyone, in the audience.

Thank you.