Onto Innovation Inc.

RCC 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 our financial results will be presented on a non-GAAP financial basis unless we specify otherwise. As a reminder, a detailed reconciliation between GAAP and non-GAAP results can be found in today's earnings release. I'll now go ahead and turn the call over to our CEO, Mike Pleszinski. Mike?

Thank you, Mike. Good afternoon, everyone, and thank you for joining our call today. I'll begin with the shift in timing of shipments for three lithography tools, which impacted our second quarter results and third quarter guidance. All three tools were for a specific customer and scheduled to ship near the end of the second quarter. During the quarter, we accepted a customer request for a specific enhancement package to be added to the tools and manufacturing. We estimated this could be installed and tested in the quarter, but ultimately the full verification took longer than planned. The customer has since approved this new functionality and the tools are included in our outlook for the third quarter. In aggregate, the second quarter revenue and third quarter outlook remains consistent with our prior guidance of just over $400 million for the sum of those two quarters. This reflects the weak demand from our advanced nodes customers being offset by strong power semiconductor revenue, and the more recently surging demand for our Dragonfly inspection systems to support heterogeneous packaging and high bandwidth memory customers. In fact, we expect this specific demand to drive more than $90 million in revenue over the next three quarters. I'll provide additional details about our outlook in a few minutes, but first let's dig a little deeper into the second quarter. We'll start with our specialty and advanced packaging customers, where revenue from this customer group grew by over 20% from the prior quarter. Revenue from our power device customers grew over 35% and included our product portfolio of inspection, metrology, and software solutions. This was our largest market in the quarter, and in fact, revenue exceeded total power device revenue in all of 2021. Building on this demand, in the quarter, we announced the intent to expand our metrology portfolio with the release of LMNS material metrology, and Atlas S OCD metrology, each specifically designed to address challenges in the compound semiconductor manufacturing. We've already secured orders for these tools and shipments will begin in the fourth quarter. With these additional products, we believe we are positioned to address an estimated 80% of process control steps in this high growth market, which is estimated to require 10 times the current volume of wafers by 2030. In advanced packaging, we delivered over 20 million of inspection systems to customers ramping heterogeneous packaging lines to support growing end market demand for high-performance compute. The flexibility of our Dragonfly systems, which integrate submicron 2D defect detection, 3D metrology, and our unique clarifying capability into a single system is proving to be a powerful tool for the heterogeneous packaging applications used in high-performance compute. For example, ClearFind technology is able to detect residue on dye-to-dye or dye-to-substrate interconnects and ensures good dye bonding, while our 3D metrology sensors provide critical stack height information and coplanarity data for the package. This metrology is proving essential to controlling yields and is leading to higher attach rates. In sharp contrast, revenue from our advanced nodes customers declined 43% in the quarter, However, even as capacity buys wane, R&D investments have continued, and in the second quarter, we were pleased to see further proliferation of Iris films when it was successfully qualified by a top 3D RAM manufacturer in the quarter. In addition, we delivered several Atlas OCD systems to two customers for gate all-around applications. We believe the declines in advanced nodes will reach a bottom in the third quarter, with strong demand in specialty and advanced packaging continuing in the second half of the year. Before moving to our outlook for the third quarter, Mark will now cover the financial results for the second quarter.

Thanks, Mike, and good afternoon, everyone. As Mike highlighted, we closed the second quarter with revenue of $191 million, down 26% over the same period last year, and below the second quarter guidance range of $195 to $203 million, due to the shift in timing of the three JetSTEP systems Mike commented on. Despite the lower revenue, we did achieve an EPS of $0.79 for the second quarter, within our EPS guidance range of $0.75 to $0.90. The revenue declined from the same period last year is primarily due to the decline in our advanced nodes business, which had revenue of $38 million and represents 20% of revenue. Specialty device and advanced packaging revenue of $112 million represents 59% of revenue, and software and services had revenue of $41 million, which represents 21% of revenue. We achieved 53% gross margin for the second quarter, exceeding our guidance range of 50% to 52%, driven by the favorable mix, shipping fewer JetStep systems and seeing an initial favorability due to our supply chain optimization efforts. Second quarter operating expenses were $59.9 million, within our guidance range of $58 to $60 million. We are still executing to our cost reduction activities. However, we have continued to maintain a slightly higher level of investment in R&D initiatives, as Mike mentioned, aligning to R&D engagements and planar films and power semiconductor applications. Our operating income of $41 million was 21% of revenue for the second quarter, compared to 29% from the prior year. Our net income in the second quarter was $39 million, or 79 cents per share. Now moving to the balance sheet. We ended the second quarter with cash and short-term investments of $610 million, an increase of $62 million from the start of the year, with operating cash flow of $31 million within the quarter, representing 17% of revenue for Q2. Inventory ended the quarter at $352 million, an increase of $14 million. We continue to actively manage down our inventory levels across the network. However, we had increases in lithography and services inventory within the quarter. We are projecting a decline in Q3 and are now targeting to be between $275 and $300 million by the end of the year. Accounts receivable decreased $22 million to $188 million in the quarter, and our day sales outstanding decreased six days to 90 days. With our inventory reduction goals and focus on cash collections, we expect to return cash flow to consistent performance of over 20%. During the quarter, we did not execute any share repurchases. We have $32 million remaining under our existing $100 million authorization. Now turning to our outlook for Q3. We currently expect revenue for the third quarter to be between $205 million and $225 million. We expect gross margins will be between 50% to 51%, primarily due to the lower advanced nodes revenue, which typically carries higher margins above our corporate average, as well as a shift of the jet step systems built in Q2, now shipping in Q3. For operating expenses, we expect to be between $57 million to $59 million. For the full year 23, we expect our effective tax rate to be between 14% to 16%. We expect our diluted share count for Q3 to be approximately 49.4 million shares. Based upon these assumptions, we anticipate our non-GAAP earnings to be between $0.85 per share to $1.05 per share. We are making progress towards reducing our fixed cost structure by optimizing our supply chain and manufacturing sites while maintaining our strategic priorities in several R&D programs and ensuring our ability to deliver financial performance in line with our long-term operating model. The team has made significant progress to date, expanding our second shift to drive higher absorption of fixed costs while improving cycle times and outsourcing several non-core subassemblies to global supply chain partners. We have set aggressive but achievable targets with our global partners, consolidating our supplier base by greater than 50% over the next two years while also simplifying our key components such as moving to a common automation system, which will drive a 25% cost reduction and is part of our identified savings for 24 and 25, as highlighted during our recent analyst day. And with that, I will turn it back to Mike for additional insights into Q3 and the remainder of 2023. Mike? Thank you, Mark.

As Mark mentioned, we expect the third quarter to be up 8% to 18%, reflecting an increase in both process control and lithography revenue over the second quarter. The expected increase in process control revenue is mostly driven by capacity expansions and higher process control intensity to support heterogeneous integration of GPUs and high bandwidth memory. Combining these two powerful technologies into a single package is enabling customers like NVIDIA to supply the specialized compute platforms to support the growing demand for large language model applications by hyperscalers and corporate enterprises. These assembly processes are more complex, and with multiple DAIs becoming a single package, the impact to yield of a single interconnect failure is much higher. The proven ability of our Dragonfly G3 system to reliably monitor for these failures is driving the projected increase in demand of over $90 million in revenue, which we mentioned earlier. We also project demand for our portfolio of products supporting power semiconductor customers to remain near current record levels. We still see headwinds in the advanced nodes, primarily from DRAM manufacturers, and we expect revenue to decline again in the third quarter, which we believe will mark a bottom for front-end memory overall. We're optimistic for a possible recovery in the second half next year as customers like Samsung are beginning to talk about the launch of New smartphones and PC promotions in the second half of 2023 being a catalyst for customer inventory reductions going into next year. In summary, we see strong demand for our products to support power semiconductor device manufacturers and heterogeneous packaging on both wafer and panel substrates. In these markets, we believe our portfolio of connected solutions provides differentiated value, further contributing to the adoption rate we're seeing. In the advanced nodes, though spending is down significantly this year, our customer collaborations are resulting in new tool of record positions for our iris planar films metrology and integrated metrology while progressing the capability of Atlas OCD metrology to new applications and more challenging transistor geometries. We expect these new product positions should lead to higher wallet share when investments resume in the advanced logic and memory markets. At the same time, We're implementing the structural improvements to our operations and supply chains that should result in a stronger financial foundation as we close in 2023 and prepare for the new year. I'll now turn the call over to Cynthia for questions from our covering analysts.

Thank you. If you would like to ask a question, please signal by pressing star 1 on your telephone keypad. If you are using a speakerphone, please make sure your mute function is turned off to allow the signal to reach our equipment. Again, press star 1 to ask a question. We'll pause for just a moment to assemble the queue. And we will take our first question from Charles Shi with Needham. Please go ahead.

Oh, thank you for taking my question here. Mike and Mark, I really just want to start out with some of the commentary around heterogeneous integration and high bandwidth memory. So on high bandwidth memory, can you kind of provide us a little bit further outlook beyond what you think the 90 million over the next three quarters? Do you think high bandwidth memory, the way it first starts, is going to grow? In what kind of fashion? Is it more linear as some other management team seems to believe? Or do you think it's going to be a much higher growth? Any indication from our customers on high bandwidth memory would be great. But then a related question really is about very specifically about co-ops. I know heterogeneous integration probably cover a lot of stuff, not just about co-ops, but specifically on co-ops, how much opportunity you're seeing today can quantify for us. Thank you.

Sure. So I'll answer the last one first. Regarding CoWAS, when we talked about heterogeneous integration specific for AI, I think it's pretty clear that, for instance, NVIDIA, who's driving a lot of this, is tied to TSMC and CoWAS. So that's a fairly significant driver for us and for that $90 million that we mentioned. As far as HBM and the growth of Overall, what we're seeing is a pretty rapid increase in demand right now. Part of it is capacity expansions. Part of it is, I think, a little bit of process control intensity increase. Certainly, that's the case in the COAS area. So it's hard to say how that rapid increase continues into 2024. We certainly expect additional orders in 2024, and we're already having some of those discussions beyond the first quarter. But is it going to continue on this rapid trajectory, this rapid growth trajectory, or will it start to slow down? That's going to remain to be seen, and a lot will depend on how quickly, I believe, the enterprise customers adopt AI and these large language models for applications internally.

Can I ask a second question? Maybe advanced nodes. You said advanced nodes, you believe it's bottoming in Q3. Can you clarify a little bit more? Because advanced nodes for you guys is not just about memory, but also advanced boundary logic. Between these two, I think you said memory is bottoming in Q3, but what about advanced boundary logic? Can you clarify if I heard something wrong? Thank you.

Yeah, I think where I was talking is overall, I think the bottom's there primarily from DRAM, so DRAM has been the biggest drop. Advanced logic has been a little more stable, and we've seen that have less of a big impact on our advanced nodes trajectory. So we do think advanced nodes, do we know if logic's going to snap back right away? No, but we do see several new fabs being built. Everybody's aware of Samsung and Taylor and TSMC Arizona and opportunities at Intel as well. We continue to see and seed pilot lines and deliver systems as those fabs are being constructed and starting to take some initial systems. But We have also seen the construction delays announced by several of these manufacturers. So do we hope for some growth in the second half next year? For sure. But we're already starting to see that plateauing of advanced logic as it is. So DRAM is the big change.

I see. Is the DRAM comment, is it more on the HBM-related, reaching that inflection point? Is that mainly the HBM being the driver, or do you see the standard DRAM, maybe the CAPEX is also seeing some inflection point? I just want to ask a little bit more on that DRAM comment.

Yeah, I think it's certainly the more common DRAM that's driving the drops we're seeing. You know, HBM is great, and it's great for our packaging business, and it's picking up some DDR5, 4, but it's not enough to make up for, you know, PCs and smartphones that are down, right? So DRAM overall is impacting us because of the vast rest of the market outside of high-performance computing.

We will take our next question from Craig Ellis with B Reilly Securities. Please go ahead.

Yeah, thanks for taking the questions, guys. I just wanted to see if I could start with you, Mike, and look beyond the calendar third quarter, because it seems like with the panel litho shipment timing issue that we will have at least three and maybe more tools. So I was hoping you could clarify how many tools were in the third quarter there. And then from there, can you just talk about the gives and takes for fourth quarter revenue? Not looking for specific guidance, but just want to understand how the gives and takes play out as we exit the year. Sure.

So it's actually a continuation of what we're guiding in the third quarter. So we see power semiconductors for the fourth quarter remaining fairly strong. Demand there is strong. We're going to introduce the new tools. That'll help a little bit, but that is more a driver for 2024 in the power semiconductor space. And then, you know, heterogeneous packaging, our support specifically for AI devices, i.e. NVIDIA and, you know, some of the COAS and HBM that were mentioned, that we can see continuing to strengthen straight through So those are the positives. Obviously, the litho will remain sort of stable and continue once we catch up with the shift here with the three tools. And then the advanced nodes, I mean, that's the tailwind. So, you know, as I look at Q4, it's a lot of the same stories, real strong in heterogeneous packaging, real strong in power. uh advanced nodes a bit of a a bit of a tailwind got it and then and overall we think that the sorry sorry uh craig to interrupt over overall we expect to be you know similar similar levels of of you know guidance for for the fourth quarter we'd expect it to be around the same yep real helpful mike and then mark i'll follow up with you so

So the company had a program for optimization and I think just more careful expense management. And that was about $27.5 million at analyst day. There was a $25 million program that was announced. Can you just help us understand to what extent our benefits from those programs factored into the third quarter's guide? What does it mean for the exit level of gross margin and this calendar year. So where can gross margin be in the fourth quarter? And then how much of the benefit of those two programs do we see hit in gross margin and OpEx next year? Thank you.

Yeah, no, I mean, let me start with, you know, coming off the analyst day and the comment around the 25 million, you know, of optimization, 24 and 25. I mean, that activity we're executing now. Obviously, we haven't seen The brunt of those savings are expected to be in 2024 and 2025. We are, as I commented in my prepared remarks, seeing some of that now and some slight shifts of suppliers and things like that. So, we still focused on that. I think for Q3 and Q4, we still have a lot more work to do. Certainly, we have seen some savings in operating expenses as it relates to taking costs out and offsetting the annualization year over year. Certainly in the back half, we're focused on that, those continued reduction plans. The Q3, we do have certainly some of that continue to be baked in and where we are from a gross margin and OPEX standpoint. I think as we look at the decline in advanced nodes, Fortunately, we have programs in place right now. We're looking and executing even further reductions in Q3 and Q4 as we look towards that decline and making sure that we can continue to drive gross margin accretion back up above the 53% to 54%. Our goal is, from a full-year perspective, is still to target that 53% to 54% and align to the long-term operating model of getting back to 55% plus longer term. So that's our target.

So that would imply that you should be at least 54% in the fourth quarter to get to 53% to 54% for the full year mark?

We're still, I won't comment on it specifically, but I think from a, you know, From where we are and trying to figure out what the business will be in Q4, I think, you know, our goal is obviously to continue to drive that, you know, 53, target 53 plus in Q4.

Got it. Thanks, guys. I would, yeah.

We will take our next question from Brian Chin with Stiefel. Please go ahead.

Hi there. Good evening or early evening. Thanks. Let's ask a few questions. Maybe to start with, Mike, I think you have really good, strong visibility on the packaging and specialty side of the business. Visibility is clearly not great in terms of the advanced nodes. But would you characterize the level of not just DRAM, but also memory investments, including non-volatile memory? Does it feel like it's below normalized maintenance or even sub-maintenance levels at this point? Does that give you some of the conviction in saying that Q3 looks like a bottom right now?

Meaning it couldn't go lower? It is pretty low. I don't know if I've ever seen this much of a drop for this long of a period of time. I think that Then you're starting to hear from our customers as well that they're beginning to see signs of price increase and inventories are coming down and that kind of thing. So, yes, I think it's at or below maintenance levels. When it recovers, I don't know, but I'm not expecting a very quick snapback. We're not planning for that. You know, that would be a nice surprise, but right now it's, That's why we're saying we don't expect Q3 to – we expect Q3 to be the bottom.

Okay. Yeah, fair enough. I know that the lithography has always been a bit bumpy and a bit lumpy for you guys. But it sounds like – obviously, you take the three units. If you park them in 2Q, you're kind of revenuing up. And if you take it out of the third quarter guidance, you're kind of flat. But – Given that dynamic and maybe you ship six tools in 3Q, is that kind of, you know, flattish preliminary fourth quarter outlook? Is that really saying that the X500 lithography ships maybe back to normalized levels, maybe down a few units Q1Q, and then you make that up with some of the growth that you're seeing in the other businesses?

Yeah, we think that the litho will be, you know, it depends a little bit. on timing because we're trying to catch up on things. So there may be one or two tools that would move forward into the first quarter. But it would more or less normalize at that $20 million or so that we've always talked about. So $20 million, $25 million of revenue.

Okay. Got it. And then just framing up sort of the specialty business where that's a synergy you've talked about. in terms of technology and customer synergies. And that, you know, I think you've expected bear more for this upcycle than kind of what you're able to progress on last upcycle. So if you look at kind of calendar 22 and already you're having some pretty good run rates on the specialty side of the business, how much larger, you know, in three years could that specialty revenue for you guys be relative to the, you know, the counter 22 year when you guys were at a billion dollars and that was going to be, you know, smaller amount of the, of the, of the revenue contribution to a billion, how much bigger could specialty be in three years?

Compared to current level.

I mean, it depends on spending from the customers, right? So to estimate out three years is always tough, but I can say that our, adoption rate is only beginning so we're just winning new customers i think we added five new customers in the quarter so those are just initial sales they'll buy a couple of tools then there's going to be repeat sales as they continue to grow and every one of those customers is a potential to sell more of the portfolio so we have opportunities to not only continue to add customers but continue to grow our portfolio within the customers So if we're looking at where we end up for this year, I wouldn't be surprised if we could double that revenue in three years. Given the growth dynamics in the industry and given our positions and the new products we're releasing and the value that we're delivering with these connected solutions, these integrated portfolios to solve unique problems.

Thanks, Mike.

We will take our next question from David Dule with Steelhead Securities. Please go ahead.

Yes, thanks for taking my questions. As far as high bandwidth memory goes, could you just talk about how much more inspection and intensive it is versus just a standard DDR5 product? And then as my second question, As far as your advanced nodes business goes for calendar 2023, how much would you expect the memory business to be down and how much would you expect the Foundry and Logic to be down?

Again, I'll start with the last one. For 2023, I'd expect, I don't have the numbers directly in front of me, but I would expect memory to be down probably double what Logic is down. So it doesn't help you too much, but I don't know if logic might be down 20%. Memory might be down double that, 40%. As for HBM and the attach rates, the process control intensity versus normal DRAM, it's a good question. The number of applications, so I would say it's two to three times more process control intensive, at least for our applications on our Dragonfly than normal DRAM and DRAM packaging. And that's because of the layers and the amount of metrology and inspection involved in preparing these dye to be stacked and to be connected to other dye in the stack. So there's a lot more... metrology coinciding with our inspection in order to make those packages yield. And the metrology is on our inspection tool, so that drives the utilization of our inspection tool. It's all integrated, just not to add any confusion.

Okay. And as far as high bandwidth memory goes and your outlook there, Is that – I'm assuming that's just with one lead DRAM manufacturer. I don't really think the other guys have a lot of product in the marketplace at this point. Is that the way to look at it, or are you working with more than one guy in high bandwidth memory?

We're working with more than one, but one certainly dominates more than the other.

Okay, thanks.

We will take our next question from the body shoulder with Jeffrey's please go ahead.

Hi, thanks for taking my question. So, I guess I wanted to ask about the packaging a little bit more. So you talked about the 90Million opportunity over 3 quarters. Is that more towards versus. like a co-wash capacity. Can you characterize what kind of what the split is between like a heterogeneous integration versus HBM?

Yeah, I would say 60-40. So a lot of it is on the logic side. So co-wash, let's say. And then, you know, the rest on the HBM.

Right. And then for my follow-up there, So, you know, you have TSMC talking about doubling their COAS capacity. ANCOR is sort of tripling their 2.5D capacity. So, how should we think about, like, how does this translate to the revenue opportunity for you? Is it like a direct link of, you know, doubling capacity means doubling revenues for you, or how does that equation work? So far, it looks that way.

Yeah, so far it looks that way. There's a very high attach rate, what we're seeing. So, you know, right now that would be how I'd model it. If yields improve and they start to drive, let's say, a normalization and an optimization, they may reduce the process control. But if anything, we're seeing an increase in process control as they're trying to increase improve yield. And that's why I mentioned the increase in process control intensity. So we think maybe even, you know, since the beginning of the year till now, maybe 20% or more increase in process control intensity on the logic side. So heterogeneous packaging on the logic side. Colossal.

And then kind of going, so, so, you know, if, if I think about what your install base was in 2021, 2022 for, for like a dragon fly inspection tool, most of that was for kind of smartphone markets where they do slip chip bumps, right? Um, how is the fungibility of those tools? Like as the utilization is low on maybe smartphones, are they, are they fungible as in, can they be used from in an info? process to a co-work process? Is that a possibility? Or do you need a different set of tools or different configurations?

Generally, they're in different lines, so the tools would have to be moved. And the tools are configured very differently for your traditional fan-out or just traditional bump. The amount of sensors and the capabilities like Clear Find, which is another option on our tool, those are all required for the HBM and for the heterogeneous, the logic side. So it's not really fungible. We haven't seen anyone trying to move our existing tools into those lines. It's all new tools.

Okay, so does that mean the gross margin sort of end up higher when you're addressing like a CoWAS HPM application versus what it was historically?

I believe so, but I'd have to, we'd have to get back to you on that.

Okay, thank you. That's all I had, and I'll get back in the queue.

We will take our next question from Mark Miller with Benchmark. Please go ahead.

Thank you for the question. The $90 million volume purchase order for process control, how does that order ship in terms of the third and fourth quarter? Is it equally divided, or is it going to be more back-end loaded?

That we're working on. So we're projecting the $90 million, and when it's finalized with the customers, we'll be able to provide clarity on the on the rollout of that and the follow-on, the size of the follow-on additional orders we expect. But I would expect Q3 to be a little lighter, and Q4 and Q1 to be maybe semi-equal between that ramp.

You announced 18 power customers' first-time orders. Any new power customers this quarter or the June quarter?

Yep, we had five new power customers this quarter.

Thank you. You're welcome.

And there are no further questions at this time. I will now turn the conference back over to Mr. Schaefer for any additional or closing remarks.

Thanks, Cynthia. Just a quick reminder for everybody about three upcoming events. First, Onto Management will be participating in the Needham Semi-Cap Virtual Conference on August 23rd. Second, we'll be participating in the Jefferies Semiconference in Chicago on August 29th. And third, we will be at the Benchmark Conference in New York City on September 13th. Thanks again for joining us today. A replay of the call is going to be available on our website about 7.30 Eastern time this evening. We'd like to thank you for your continued interest in On2Innovation. Cynthia, please conclude the call. Thank you.

Thank you, gentlemen. This concludes today's call. Thank you for your participation, and you may now disconnect.