ONTO - NYSE

Good day, and welcome to the Onto Innovation Third Quarter Earnings Release Conference Call. Today's conference is being recorded. At this time, I would like to turn the conference over to Sidney Ho. Please go ahead.

Thank you, Rachel, and good afternoon, everyone. Onto Innovation issued its 2025 third quarter financial results this afternoon shortly after the market close. If you did not receive a copy of the release, please refer to the company's website where a copy of the release is posted. Joining us on the call today are Michael Plisinski, Chief Executive Officer; and Brian Roberts, Chief Financial Officer. I'd like to remind you that the statements made by management on this call will contain forward-looking statements within the meaning of the federal securities laws. Those statements are subject to a range of changes, risks and uncertainties that can cause actual results to vary materially. For more information regarding the risk factors that may impact Onto Innovation's results, I would encourage you to review our earnings release and our SEC filings. Onto Innovation does not undertake the obligation to update these forward-looking statements in light of new information or future events. Today's discussion of financial results will be presented on a non-GAAP financial basis unless otherwise specified. As a reminder, a detailed reconciliation between GAAP and non-GAAP results can be found in today's earnings release. Let me now turn the call over to our CEO, Michael Plisinski. Mike?

Thank you, Sidney. Good afternoon, everyone, and thank you for joining us on our call today. Underpinning our financial results, which came in ahead of the midpoint of our guidance ranges, the Onto Innovation team made excellent progress with our strategic initiatives, including new product adoption, advancing our offshoring activities and preparing for the close and successful integration of the Semilab transaction. We expect each of these efforts will enhance our leadership position in the exciting advanced packaging and advanced nodes markets and strengthen our outlook for growth in 2026. Market growth in 2026 is likely to include increased investments in advanced packaging to support the strong demand for AI compute. So we are very pleased to announce that our 3Di technology has successfully completed the full qualification process at not 1 but 2 high-bandwidth memory customers in the quarter. Our 3Di technology demonstrated superior performance on smaller denser 3D interconnects, critical for next-generation devices. Following these successful qualifications, we started discussions for volume orders with integrated 3Di and subsurface defect inspection to support next-generation HBM devices. Another win for the 3Di in the quarter was an order from a leading OSAT to support 2.5D applications for AI packaging. To support advanced 2D inspection applications, the launch of our next-generation Dragonfly system is progressing well with the first shipment expected in a few weeks, followed by additional systems in December. After last quarter's optical performance validation by a key customer, we have since completed successful in-house wafer studies for high-bandwidth memory and hybrid bonding applications, leading to several more evaluation shipments to customers in the first quarter. In fact, the success of these demos has several customers adding the new Dragonfly to preliminary discussions on volume needs for 2026. Turning to advanced nodes. We remain on track to deliver a record year in advanced node revenue outside of China. Contributing to this performance is the growing adoption of our Iris films and integrated metrology platforms, both on track to set records for the year. Looking at the markets broadly, recent headlines continue to reflect strong and sustained demand for AI and high-performance compute. NVIDIA projects that global AI infrastructure investments could reach $3 trillion to $4 trillion by the end of the decade, potentially reshaping the semiconductor supply chain. At the core of this evolution, our new memory and logic transistors and packaging architecture supporting chiplets for logic, 3D stacking for memory and nascent co-packaged optics, all designed to increase device performance while lowering power consumption. Onto Innovation continues to play a pivotal role by working closely with our customers across this broad value chain to develop and deliver the process control solutions required to support this AI era. In the immediate term, we expect revenue growth of approximately 18% at the midpoint of our Q4 guidance range. The greatest contributor to this growth is from 2.5D packaging customers, where we expect revenue to nearly double from the third quarter, driven by strong Dragonfly system demand. We expect advanced nodes revenue will also improve with increases in DRAM and logic spending. While discussions for capacity needs in 2026 are in early stages, our packaging customers are indicating the potential need for as much as 20% more tools to support expansions and new applications for our 2D subsurface and 3Di inspection technologies. While quarterly performance may show variation, we expect sequential growth in the first half of next year with more meaningful growth expected in the second half of 2026, driven by increased contributions from new products and potential capacity expansions. Supporting this growth is our aggressive ramp of our extended factories in Asia, and I am pleased to report in the third quarter, we successfully shipped over 30% of third quarter tools from these factories. Thanks to the incredible efforts of our operations team and supply chain partners, we are now on pace to be capable of shipping over 60% of our production demand from our international locations by the end of the first quarter of 2026. These efforts will enhance our competitive position, mitigate tariff impacts, provide greater manufacturing flexibility and allow us to expand gross margins in 2026. Finally, a brief update on our pending acquisition of 3 complementary product lines from Semilab. In October, in response to a second request letter from the Department of Justice, we amended the transaction to exclude a relatively small product line. We currently expect that the transaction will close in the coming weeks and be accretive to both revenue and earnings in 2026. And with that, let me turn the call to Brian to review our financial highlights and provide fourth quarter guidance. Brian?

Thanks, Mike. Good afternoon, everyone. Third quarter performance met or exceeded expectations across key financial metrics as we work to improve our forecasting processes and implement more disciplined spending controls. Revenue for the quarter was slightly ahead of the midpoint of our previous guidance range at $218.2 million. Gross margin for Q3 2025 was 54% and includes approximately a 1 percentage point impact related to tariffs. Operating margins of 21.1% exceeded the top end of our guidance range as we maintained our focus on variable cost control in the quarter. Finally, adjusted earnings per share for the quarter were towards the high end of our guidance range at $0.92. At a market level for the third quarter of 2025, advanced nodes generated revenue of $54 million or 25% of revenue as DRAM and NAND revenue decreased as expected sequentially from the second quarter. For the full year 2025, advanced nodes revenue is expected to double to approximately $300 million as compared to $148.5 million in full year 2024. Specialty devices and advanced packaging revenue was $113 million or approximately 52% of revenue. A strong rebound to approximately $150 million in specialty device and advanced packaging expected in Q4, revenue for this market should finish slightly higher than $500 million for the full year. Software and services revenue of $51 million comprises the remaining 23% of Q3's results. The team did an outstanding job generating cash in the third quarter as cash from operations increased sequentially to $83 million from $58 million in Q2. This represents cash conversion of approximately 185% of our non-GAAP net income in the quarter. Given the pending acquisition of Semilab, we did not repurchase shares in the third quarter. Once the acquisition closes, which is expected in the coming weeks, we will pay Semilab $432.3 million in cash and issue 641,771 shares of our common stock. The value of the total transaction based upon Onto's closing price as of June 27, 2025, is approximately $495 million, a decrease of about $50 million from the original terms of the deal. Now turning to our outlook for the fourth quarter. Revenue is expected in the range of $250 million to $265 million, representing 15% to 21% sequential growth. As Mike noted, the majority of the Q4 increase is expected to be driven by strength in advanced packaging with more modest improvement in advanced nodes, specifically around DRAM and logic. At the midpoint of the revenue guidance range, we would expect to achieve approximately 50 basis points of sequential gross margin improvement in Q4. Our Q4 gross margin expectation also includes an anticipated percentage point impact of tariffs or approximately $2.5 million of cost, primarily due to inbound tariffs on raw material imports. Operating margins for the fourth quarter are expected to rebound to a range of 24% to 26%, on operating expenses of approximately $77 million. The fourth quarter, which will officially end on January 3, 2026, includes an additional 14th week, given the company's historical fiscal closing structure. The impact of this extra week is approximately $3 million in incremental operating expenses in the fourth quarter, representing approximately 120 basis points of operating margin. Starting with the first quarter of 2026, Onto Innovation will switch to a quarterly calendar schedule of March 31, June 30, September 30 and December 31. Earnings per share for the fourth quarter is expected in the range of $1.18 to $1.33 per share, assuming an estimated tax rate of approximately 13% to 15% and about 49.4 million shares outstanding. As a reminder, we are not including the pending Semilab transaction in our current Q4 guidance. And with that, let me turn it back to Mike for some closing thoughts before we take your questions. Mike?

Thank you, Brian. In summary, we've made great progress on key initiatives that will position us for growth in the coming year. On the product front, we plan to ship our next-generation Dragonfly system in the coming weeks to a leading AI packaging customer with several additional systems slated for memory customers in December. Our 3Di technology has now been validated by 2 leading suppliers of high-bandwidth memory and adoption is expanding across a broader customer base. From a broader market perspective, the long-term outlook for AI and advanced node investments continues to build, driven by aggressive infrastructure expansion plans globally over the next several years. With our differentiated product portfolio and technology leadership in advanced nodes, advanced packaging and specialty devices, Onto Innovation continues to be well positioned to serve our customers and capitalize on these secular trends. We expect to see organic growth in 2026 with momentum building toward the second half of next year. And now, Rachel, let's open the call for questions from our covering analysts.

[Operator Instructions] And we will take our first question from Craig Ellis with B. Riley Securities.

Congratulations on the good execution guys. I wanted to start, Mike, just by following up on your most recent comments regarding organic growth through the year. Can you comment on what you'd expect for your 2 big segments, advanced packaging and advanced nodes? And any color on the linearity with those businesses?

You mean for 2026?

For 2026, yes.

Yes. So I think it's a little early to provide especially linearity quarter-to-quarter kind of view. If we look at first half, second half, we think the first half is going to be sequentially better than the second half of 2025. So we do expect growth in the first half with more significant growth in the second half, driven by several different expansions that our customers are talking to us about as well as the impact from the new products coming online being more widely adopted, let's say, getting cut into volume production. So that would be the 3Di, and it would also include the Dragonfly -- the new Dragonfly system.

Yes. Nice to see the new products. And then the second question is for Brian. Brian, as we look at gross margins next year, can you just help us with some of the gives and takes? Tariffs have been in gross margin, but when can that come out? And how should we look at some of the other gives and takes with that line item?

Sure. On the gross margin front, I mean, I think we'll start to see the tariff impact start to mitigate next quarter. Keep in mind, most of our tariffs are on inbound and so they sit in inventory for a quarter and then they start to come out. So as we continue to ramp up expansion of the offshore extended factories and more tools are going from there and supply chain is going right directly to those factories, we'll start to mitigate the tariff risk. I think as we go through the transition to extended factories, we'll see a little bit more meaningful gross margin expansion as we get towards mid and the latter part of 2026. But we're certainly poised to have a good solid year of gross margin expansion.

Got it. And then lastly and somewhat clerically, once Semilab closes, do you plan to host another conference call? Or how will you update us once that's done?

My expectation is we'll just update you as part of the next earnings call. I don't think we'll provide an update right after close. We're going to take a little bit of time meeting with the team, having some more detailed discussions and then provide a more informed view of the business probably in the next earnings call.

And we will take our next question from Ezra Weener with Jefferies.

First one would be about your commentary on 2.5D packaging. You talked about sequential growth into March. I just want to make sure I heard that correctly and you weren't talking about the entire business. And then the other sequential question would be, when you talk about sequential growth in the first half, is that half over half or quarterly sequential growth?

So when I mentioned sequential growth, it was half over half. So second half of '25. We would -- first half of '26 would be sequentially stronger than second half of '25. And I'm not sure the 2.5D question on -- what was your question there?

You had talked about sequential growth immediately after 2.5D packaging. I was wondering if that comment had to do specifically with 2.5D packaging.

No. It was -- yes, mainly just, I think, AI packaging is primarily driven by our -- I think that was about Q3 to Q4 driving growth, and it's primarily due to the AI packaging and strong demand for Dragonfly systems.

But the sequential growth -- sorry, the sequential growth in the first half of '26, we were talking about the entire business, Ezra.

Understood. And then the second question would be, with the understanding that you might not see revenue until the second half from the new Dragonfly. Can you talk about the ecosystem between now and then and the timing of when you would see revenue from that?

I didn't say we wouldn't see any revenue. It's very possible we'll see revenue in the first half. It's just going to be onesie, twosies until it starts shipping in volumes, larger volumes. So second half will be more meaningful. But I would expect to convert some of these early shipments in -- as soon as the first half.

And we will take our next question from David Duley with Steelhead Securities.

Regarding the qualification of the 3Di tool at 2 HBM customers for bump inspection, is that tied to the ramp of HBM4? And are you going to be the first source or the second source there?

Well, we're not aware of anyone else being qualified through the stringent tests yet. So our hope is first choice, however you worded that. But I think the -- it is tied to an HBM4. We've been working with customers on other ways to do even existing processes that would provide a better yield impact. So applying the 3Di technology that we supply in different process steps, which would allow for some level of rework, that would help the customers if they do detect any issues to rework it and then drive some higher yields. That's something that only our tools can do, but it's a very new capability. So customers are working on what's the true impact of that across their process, do they want to make the change now in existing processes or only in the forward-looking.

Okay. My second question is more of a clarification. I just -- I wanted to make sure I understood everything you said. So as far as the Dragonfly goes, you're going to start to ship the new tool to your primary customer that you've lost share with over the next couple of weeks. Could you maybe just rerun everything you said about the Dragonfly so I get it right?

I have an entire script on the Dragonfly. So I believe what I said was we'll be shipping Dragonfly to a 2.5D logic packaging customer in weeks. So that's one. And then I said that Dragonfly -- additional Dragonflies will ship in December, primarily to memory customers. Then I said that in the first quarter, based on the successful demos we've done in the third quarter, we are shipping several more Dragonflies. So we've already increased the number of Dragonflies we intend to ship.

Okay. And it's not just to one customer, it's to multiple customers?

Correct. Yes. It's to multiple customers. Yes.

And then final question for me is, as far as your core inspection business with HBM, do you think that's going to be a growth factor in 2026? Or has it started to turn on with the ramp of HBM4? Or how should we think about that?

I think we're -- the customers are also still trying to figure that one out. So for sure, from their perspective, there's going to be some growth. What it means for the process control is not completely clear yet. Customers are working on their allocations, what they spent on process control in the prior year, how much can be reused, do they adjust sampling plans, et cetera. The normal things they always do. So that's the discussions we're having now with customers. I would say the real good, strong takeaway though is the positions of our tools are demonstrating unique capabilities for the next-generation devices. So as customers start to bring those next-generation devices into more higher volume, primarily in the second half of next year, we should see an outsized positive impact from that.

We will take our next question from Matthew Prisco with Cantor.

So I guess, first, I'll stick on Dragonfly. It seems like making good progress in the quarter. So kind of after you shift these initial valuation tools, what milestone should we look to from here? And how should we think about the timing of customer adoption decisions both at TSM and all these ancillary opportunities? And then the transition from those adoption decisions to revenue, how quickly can those kind of start shipping for production?

So I think it would be reasonable to assume second half. In fact, we said that in the second half, we would expect to see more meaningful revenue from the new products that we're shipping. So as we're qualifying, and that's the good news about shipping to a variety of customers, we're getting those qualifications started such that they're also intending to try and ramp into the second half to cut these technologies into the second half. So as long as we continue to execute, as long as the tools perform as well as they're demonstrating now, yes, I would expect, like I mentioned earlier, incremental revenue in the first half, that means closing out some of these initial tools with more meaningful revenue in the second half tied to the volume adoption that cut into production.

That's helpful. And then maybe going to that 3Di, those new qualifications. Can you kind of give some more color on what drove the wins there? How you're seeing your competitive positioning in that technology today? And then how do we think about the translation of those wins to the P&L and potential magnitude of impact there?

So I don't think there's this huge, big jump into the 2026 for the 3Di. There will be an incremental improvement, and then we'll see a bigger impact even in 2027. By incremental, I mean, tens of millions of dollars will be driven by 3Di. So -- now why is -- why are we winning? Why is the 3Di so important? There's a couple of things going on. One, in the last call, we talked about several -- well, let me start with the technology. The technology is differentiated and that it uses the laser-based coherent light. Coherent light allows us to focus it in between the dense smaller bumps. And we can do that at a throughput and with the precision that the customers require, which is extremely stringent as they're moving from HBM3, 4E and beyond. That's one thing. So that's one of the reasons we've won those 2 qualifications and it was obviously a very stringent evaluation period. The second is because of this capability, where the technology and what it's providing, we're opening up several new applications. So -- actually, at least 3 new applications. One is the one I described earlier, where we can potentially apply 3D bump metrology at a different step in the process, which will allow customers, let's say, to do some rework and provide a better yield improvement. That's one. Two is there's 2 other new applications where the technology has a speed and the precision to provide other types of metrology across surfaces. So across die, die warpage metrology as well as some specialized metrology for 2.5D packaging.

And we will take our next question from Brian Chin with Stifel.

Maybe first, just to follow up on the 3D discussion. From the sounds of it, this sounds like it's the pre-reflow bump metrology step, Mike. Is that correct? And -- is this an additional or new metrology step for many of these adopters?

Correct. And it is a new metrology step because previous technologies couldn't reliably measure at that pre-reflow step. The light would scatter too much is what we're told.

Okay. And would you expect -- in terms of the amount or the metrology time required kind of post reflow, would that kind of potentially decrease that?

I think you would eliminate. So you'd be -- the customers, their intent and our discussions are around shifting the metrology to pre-reflow, so you don't need to do a post-reflow. There's a strong correlation there, but you have the ability to rework.

Got it. And then for a follow-up question, kind of weaving in sort of, again, your view of advance to WFE markets as well as advanced packaging. Q3 this year kind of a low quarter. You're up in Q4. And so sequential first half versus second half, it's helped by that kind of low 3Q. But when you kind of think about whether the quarters without giving a specific quarters, 1Q, 2Q, whether they'd be kind of linearly up or there could be sort of up or down. What are the kind of the key swing variables you're looking at around either advanced packaging, CoWoS investments, maybe also kind of changes in sampling plans, reuse, et cetera, and also advanced nodes, where there does seem to be some pickup now in the DRAM side of that spending. And maybe you can even kind of like even advanced foundry front-end spending. So kind of all the kind of variables that you think about in terms of how the revenue could trend in the first half next year.

So I think it really depends on the customer spend plans, right? So -- and that's always the case. That's why I'm struggling how to answer it. A lot of customers are opening and have spoken about factory expansions, but most of them are going in for second half. We, of course, get process control systems in early. So we get in, let's say, in before that the larger WFE starts to rise and pick up from that. So -- and that's normal process control spending patterns. That's one. On the AI packaging front, there is some level of capacity digestion, process optimization, that's both positive and negative. So some of it is, hey, because of the challenges with the new technologies, some of it is tied to, hey, we overbought in one area. We need to optimize some processes there, but wholly macro, we have struggled in another area. And we need the new technology. So some of these 3Di applications, I mentioned, are new technologies. Some of the subsurface applications are new where we're seeing a lot of demand are for new applications. So it's very hard for us to quantify that. And that's why we said there could still be -- so to answer your question in one way, we don't expect linear. We do expect some variability quarter-to-quarter. It's kind of natural. But the secular trend, we think, is positive, and we also think we're well positioned.

And we will take our next question from Edward Way -- from Edward Yang with Oppenheimer.

Just to clarify on a couple of numbers, the tens of millions of 3Di you're talking about, that's for 2026, correct? And I also heard you say something about shipping more tools. Was that related to AI packaging?

Correct. Yes. And it's based on initial discussions. So the early indications -- kind of going back to Brian Chin's question, we're getting early indications of a certain level of demand, but that can change. And we're still in the early stages of discussion. So it's really hard for us to provide any kind of real guidance on that. But the 3Di, correct, that was for 2026.

And the 20% more tools comment you had in the script?

It's for primarily AI packaging. Correct.

Okay. And that would account for the fact that you wouldn't be shipping or recognizing much revenue from your new high-res Dragonfly G5 tool until the second half of 2026, right?

Correct.

Okay. Wonderful. And could you just comment or provide your thoughts on the tightness and the strong pricing your customers are enjoying in memory markets and how that might pertain to 2026? You had a $69 million DRAM VPA starting this year. Does that cover this current market strength that you're seeing in those memory markets? Or do you expect more orders coming down the pipe?

I believe the existing VPAs covered this year, and I believe we've worked through those. I don't think there's any -- we've had actually a pretty strong year in advanced nodes. So I believe most of that is already covered. And the discussions we have now are all new VPAs for next year and slightly beyond.

Okay. And my follow-up would just be on Semilab again. On the amended terms, were you surprised that regulators had wanted some scrutiny around that? And could you just provide your updated confidence in approval with the changes that you made and the timing on that?

Well, since it's not closed and not wanting to ruffle any feathers from potential regulators, I will say, yes, we were surprised, but we worked cooperatively with Semilab team, with everybody to find a very reasonable solution.

And we will take our next question from Vedvati Shrotre with Evercore ISI.

The first one I wanted to understand was the first half '26 sequential increase that you talked about, like where do you have the most confidence, advanced node versus packaging? Is one doing better than the other? If you could give any directional sense on that?

I think -- I actually don't have the numbers in front of me, but my impression is that it's probably advanced nodes in specialty devices continuing to show some strength. Advanced nodes also okay, but advanced nodes, many of those factories are second half. So I think that will be much stronger in the second half.

So sorry, banking nodes will be more stronger in the second half, is that...

Yes, I think -- so basically, the advanced packaging segment, advanced packaging, specialty device, probably coming out in the advanced nodes in the second half. But again, as Mike has kind of pointed out, a lot of those discussions are still early days as we work through the exact timing of spend with customers and all of those pieces.

Understood. Okay. And then I think last quarter or the whole of this year, you sort of had that view that one of the HBM suppliers isn't qualified and that's kind of limiting the visibility you have on HPM progression. How has that conversation changed last quarter versus this quarter? Has the visibility gotten much better now that the demand dynamics have changed so significantly for the memory suppliers?

Well, I won't -- the visibility is improving for -- with regards to our discussions with the suppliers, but who's been qualified and who's -- what the allocations are, I'm not so sure that, that visibility has gotten much better.

Understood. Okay. And then the third thing I kind of wanted to understand was, if I sort of take the larger caps, like the Lam, not necessarily your competitors, but a lot of them are already seeing like strong quarter-on-quarter growth on HBM and DRAM like fourth quarter versus third quarter. Are you seeing any of that?

First, you mean in the fourth quarter?

Yes.

In the fourth quarter, yes, we mentioned that we are seeing strength in the -- well, more of our -- yes, strength in the memory. Yes, DRAM. DRAM then some logic. Obviously, NAND is still very weak.

And HBM is muted. Is that kind of fair?

Well, then HBM goes to the packaging side. In packaging, I think we've just -- we mentioned that it's an extremely strong growth from the AI packaging side. I think we said almost 50%.

Right. So the packaging growth, my interpretation was most of it was driven by essentially your OSAT plus kind of the foundry logic piece growing. Does that have HBM element in it as well?

For sure, HBM as part of our Q4 forecast is driving most of the growth. I mean I didn't break it all down into that, but it's for sure AI packaging, and that includes HBM.

[Operator Instructions] And at this time, we have no further questions. I would now like to turn the call back.

Thanks, Rachel. We will be participating in a number of investor conferences throughout the quarter. We look forward to seeing many of you there. A replay of the call today will be available on our website at approximately 7:30 Eastern Time this evening. We would like to thank you for your continued interest in Onto Innovation. Rachel, please conclude the call.