FormFactor, Inc.

Thank you, and welcome, everyone, to FormFactors' first quarter 2023 earnings conference call. On today's call are Chief Executive Officer Mike Slessor and Chief Financial Officer Shai Shahar. Before we begin, Stan Finkelstein, the company's VP of Investor Relations, will remind you of some important information.

Thank you. Today, the company will be discussing GAAP P&L results and some important non-GAAP results intended to supplement your understanding of the company's financials. Reconciliations of GAAP to non-GAAP measures and other financial information are available in the press release issued today by the company and on the investor relations section of our website. Today's discussion contains forward-looking statements within the meaning of the federal securities laws. Examples of such forward-looking statements includes those with respect to the projections of financial and business performance, future macroeconomic and geopolitical conditions, the benefits of acquisitions and investments in capacity and in new technologies, the impacts of global, regional, and national health crisis, including the COVID-19 pandemic, anticipated industry trends, potential disruptions to our supply chain, the impacts of regulatory changes, including the recent US-China trade restrictions, the anticipated demand for products, our ability to develop, produce, and sell products, and the assumptions upon which such statements are based. These statements are subject to known and unknown risk and uncertainties that could cause actual results to differ materially from those expressed during this call. Information on risk factors and uncertainties is contained in our most recent filing on Form 10-K with the SEC for the fiscal year end of 2022 and in our other SEC filings, which are available on the SEC's website at www.sec.gov and in our press release issued today. Forward-looking statements are made as of today, May 3, 2023, and we assume no obligations to update them. With that, we will now turn the call over to FormFactor CEO, Mike Schlepper.

Thanks, everyone, for joining us today. FormFactor's first quarter revenue was comparable to the fourth quarter's, exceeding our outlook range due to shipments pulled in from the second quarter. We also generated significant sequential improvement in gross margin and profitability, achieving results at the top end of their ranges. In the second quarter, we anticipate similar results, as we executed against the demand environment that remains relatively stable at diminished levels. While this cyclical downturn continues, we're carefully balancing short-term results and long-term investments with discipline cost control to preserve quarterly profitability and protect our strong balance sheet. We continue to invest in R&D for new product innovation and competitive differentiation, as well as in the long lead time facilities and equipment components of our capacity increase plans. These investments are designed to produce market share gains and above industry revenue and profit growth when we emerge from the current cyclical trough, positioning form factor to achieve and then surpass the levels of our current target financial model. Turning now to segment level details, In our largest business, Foundry and Logic Probe Cards, revenue increased significantly in the first quarter as we benefited from multiple technology-driven wins at major Foundry and IDM customers. We expect some moderation in Foundry and Logic demand in the second quarter as customers digest these recent shipments in their initial production ramps of five nanometer and three nanometer mobile application processors, tile-based QIUM microprocessors, and high-performance compute chips. Advanced packaging processes like Foveros and 3D Fabric are increasingly being adopted on these leading-edge foundry and logic chips, providing an exciting opportunity for form factor. As we've noted in the past, these integration schemes drive both higher test intensity, which expands the number of probe cards required for good die-out, and higher test complexity, which raises the performance requirements for each probe card. Probe card architectures like FormFactor's MEMS technology are essential to meeting these challenging performance requirements at a compelling cost of ownership, while also fulfilling the short delivery lead times that our customers' rapid and dynamic production ramps require. In memory, as expected, we experienced weak first quarter demand for both DRAM and flash probe cards, as each of the major memory manufacturers reduced wafer starts to consume elevated inventory levels. However, even at these lower production levels, each of our major memory customers continues to advance the roadmaps with new design activity and innovations like DDR5 and third generation high bandwidth memory, which are driving a modest expected sequential increase in second quarter DRAM revenue. This provides insight into the unique characteristics of probe card demand. Since probe cards are a consumable specific to each customer's individual chip designs, early low volume production activity for new DDR5 and HBM designs is generating demand for new probe cards, albeit at lower volume levels than we typically see in a cyclical upturn. This new design activity extends beyond DRAM and is broad-based across our customer base and serve markets. As a result, FormFactor's design engineering teams are currently operating at full utilization as we work with customers' test engineering teams to adapt our MEMS probe card architectures to their upcoming new chip designs. These diverse design projects span 3 nanometer processors, RF bond saw filters, and HBM DRAM, and foreshadow demand acceleration for form factor when the industry returns to growth and these new designs ramp in volume. Similar to Foveros and 3D Fabric in Foundry and Logic, The growth of HBM is an exciting opportunity in the DRAM market as well, as the die stacking advanced packaging process increases both test intensity and test complexity of HBM products compared to standard single die DRAM architectures. This in turn increases both the number and complexity of probe cards required for good die out. HBM is also a key enabler for generative artificial intelligence. with one of our major DRAM customers recently stating they expect their HBM sales to increase more than 50% this year, with expectations for further growth next year because of AI applications need for this higher speed memory. With ProbeCard lead times of typically less than a quarter, our visibility remains very limited, but we continue to be encouraged by the relatively stable overall demand for our products, along with the strong new design activity we're supporting across the industry. As is evident from the details of our outlook, we are experiencing underlying quarter-to-quarter shifts in spending between customers and markets, but our broadly diversified lab-to-fab product portfolio across Foundry and Logic, DRAM, and Flash probe cards, together with our system segment products, enables us to compete for business across these diverse demand pools at all major customers, producing the relatively stable overall top-line results we've demonstrated for the past several quarters and expect again in the second quarter. Shifting now to systems, our systems business continues to deliver strong results, with revenue and gross margin at near-record levels again in the first quarter, and our second-quarter outlook assumes record revenue. This strength is driven by our market-leading test and measurement products for early development of applications like silicon photonics, quantum computing, and advanced packaging metrology. Silicon photonics is an important and exciting driver in this segment. Our turnkey electro-optical measurement systems, built on our CM300 and Summit 200 engineering probers, are tightly integrated with subsystems from leading partners like Keysight Technologies and PI. Together with FormFactor's electrical probe cards, these systems are enabling our customers to rapidly characterize, analyze, and debug new silicon photonics devices in early development and pilot production. Independent market research forecasts a 30-plus percent growth rate for silicon photonics devices over the next several years, driven by the data center speed and power efficiency improvements that derive from co-package optics. And FormFactor is well-situated to capitalize on this opportunity as silicon photonics undergoes the transition from the lab to the fab. Finally, we continue to manage export restrictions in serving domestic China semiconductor customers. As a U.S.-based supplier with significant exposure to the leading edge foundry and memory technologies and customers affected by recent U.S.-China export controls, these regulations are a headwind in all of our businesses. We believe this U.S.-China trade headwind will persist over time and do not anticipate any relaxation of advanced semiconductor export controls. This, of course, provides a strong incentive for domestic China customers to onshore their supply base and de-emphasize foreign suppliers like FormFactor to ensure their business continuity. I'd like to close by reiterating that in the short term, we're encouraged by the stabilization of demand across our diversified product and technology portfolio in the first half of 2023. Over the longer term, we remain excited and confident in the growth prospects for form factor in the industry overall, driven by the fundamental trends of semiconductor content growth and innovations like advanced packaging and silicon photonics. These are trends where form factor is well positioned as an industry and technology leader, and we're confident that our commitment to invest in R&D and capacity will position form factor to emerge from the current cyclical downturn a stronger and leaner competitor. enabling us to achieve our target model that delivers $2 of non-GAAP earnings per share on $850 million of revenue. Shai, over to you.

Thank you, Mike, and good afternoon. As you saw in our press release, and as Mike mentioned, Q1 revenues slightly exceeded the high end of our outlook range, and non-GAAP gross margin and non-GAAP EPS were at the high end of the range. First quarter revenues were $167.4 million. a 0.9% sequential increase from our fourth quarter revenues, and a year-over-year decrease of 15.1% from our Q1-22 revenues. Q1 revenues were slightly above the high end of our outlook range, mainly due to shipments pulled in from the second quarter. ProbCard segment revenues were $127.3 million in the first quarter, an increase of $3 million, or 2.4%, from Q4. The increase was driven by higher Foundry and Logic revenues, mostly offset by decreases in DRAM and Flash revenues. System segment revenues were $40.1 million in Q1, a $1.5 million increase from the record fourth quarter, and comprised 24% of total company revenues, slightly down from 25% in Q4. Within the Probe Card segment, Q4 Foundry and Logic revenues were $101.6 million. a significant 24% increase from Q4. Foundry and Logic revenues increased to 61% of total company revenues compared to 50% in the fourth quarter. DRAM revenues were $19.8 million in Q1, $7.4 million, or 27%, lower than in the fourth quarter, and were 12% of total quarterly revenues as compared to 16% in the fourth quarter. Flash revenues of $5.9 million in Q1 were $9.1 million lower than in the fourth quarter and were 3.5% of total revenues in Q1 as compared to 9% in Q4. Gap gross margin for the first quarter was 36.5% of revenues as compared to 27.2% in Q4. Cost of revenues included $3.3 million of gap to non-gap reconciling items, which we outlined in our press release issue today, and in the reconciliation table available in the Investor Relations section of our website. On a non-GAAP basis, gross margin for the first quarter was 38.4%, 6.7 percentage points higher than the 31.7% non-GAAP gross margin in Q4. We achieved significant improvements in gross margin in the probe cart segment and also delivered a higher gross margin in the system segment. Our probe cart segment gross margin was 34.3% in the first quarter, a significant increase of 8.8 percentage points compared to 25.5% in Q4. The increase is due to the net effect of the following four factors. First, four percentage points of the increase relates to improved manufacturing expenses, such as lower wafer processing costs and better production years. Second, three percentage points is due to a more favorable revenue mix, with higher Foundry and Logic revenues and lower memory revenues. And third, three percentage points relates to a lower access and obsolete inventory reserve. Partially offsetting these three positive factors is a two percentage points decrease related to lower factory utilization. In terms of labor costs, we achieved the full quarter of restructuring plan savings in the first quarter versus about a third of the savings in Q4. In comparing Q1 to Q4, these additional savings were mostly offset by lower PTL taken in the first quarter and the typical beginning of the year impact of the annual benefits reset. Our Q1 system segment gross margin was 51.7%, 130 basis points higher than the 50.4% gross margin in the fourth quarter, reflecting slightly higher revenue and a more favorable mix. Our gap operating expenses were $61 million for the first quarter, similar to the fourth quarter. Non-gap operating expenses for the first quarter were $51.2 million, or 30.6% of revenues, as compared with $47.9 million, or 28.8% of revenues, in Q4. The $3.3 million increase is primarily due to higher performance-based compensation, annual benefits reset, and less PTO taken. Company non-cash expenses for the first quarter included $9.3 million for stock risk compensation, $2.9 million for the amortization of acquisition-related intangibles, and depreciation of $7.6 million, all at similar levels to the fourth quarter. GAAP operating income was $0.1 million for Q1, compared with GAAP operating loss of $16 million in Q4. Non-GAAP operating income for the first quarter was $13.2 million, compared with $4.8 million in the fourth quarter. GAAP net income for the first quarter was $1.3 million, or 2 cents per fully diluted share, compared with a GAAP net loss of $13.7 million, or 18 cents per fully diluted share, in the previous quarter. The non-GAAP effective tax rate for the first quarter was 13.9%. 7.4 percentage points lower than the 21.3% in Q4. In our previous earnings call, we discussed that we expected our non-GAAP effective tax rate in 2023 to decrease to 6% to 9% due to the expected benefits of the new Advanced Manufacturing Investment Credits, or AMIC. However, in accordance with the IRS's recently issued proposed regulations, we plan to record the AMIC benefits as a reduction to our depreciation expense over the life of the assets, rather than lowering our tax expenses. We have thus updated our outlook for the annual non-GAAP effective tax rate to mid-teens, between 13 and 17%. The benefits above the line, to gross margins and operating expenses, are not expected to be material in 2023, since the reduction in depreciation will be recognized over the service lives of the qualifying assets. Maintaining profitability at the current reduced demand levels continues to be an important goal for us, and the actions we took during Q4 to reduce our costs had the full impact in Q1, contributing to a first quarter non-GAAP net income of $12.5 million, or $0.16 per fully diluted share, compared to $4.1 million, or $0.05 per fully diluted share, in Q4. Moving to the balance sheet and cash flows. We had negative free cash flow of $7.3 million in the first quarter compared to negative free cash flow of $5.4 million in Q1. Sorry, let me repeat that. We had the negative free cash flow of $7.3 million in the first quarter compared to negative free cash flow of $5.4 million in Q4. Cash provided by operating activities in Q1 was $12.3 million, lower than the $20.7 million in Q4. The main reason for the decrease in cash provided by operating activities is timing of both tax payments and payments from our customers. We invested $19.7 million in capital expenditures during the first quarter compared to $26.2 million in Q4. With the core drivers underpinning our strategy still in place, we continue to execute on increasing our long lead time facilities and equipment portions of our capacity increase plan a bit placing equipment in service at a lower rate, at a slower rate, to ensure capacity does not significantly outpace demand. We expect CAPEX for 2023 to range between $50 and $60 million. During the quarter, we received an $18 million grant from the state of California, which we plan to record as a benefit to the P&L as we make progress in achieving the five years plan goals of headcount increases and capital investment. No P&L benefit is expected in 2023. Overall, at quarter end, total cash and investments were $240 million, down $2 million from year end. As of the end of the first quarter, we had one term loan remaining with a balance totaling $15.2 million. Regarding stock buyback, during the first quarter, we did not purchase additional shares under our $75 million two-year buyback program. At Q1 quarter end, $18.6 million remained available for future repurchases. Turning to the second quarter non-GAAP outlook. As Mike mentioned, we expect Q2 revenues to be similar to Q1, with lower Foundry and Logic revenues, partially offset by higher DRAM revenues, and expected record revenues from our system segment. This demand results in a Q2 revenue outlook of $162 million, plus or minus $5 million. Second quarter non-GAAP gross margin is expected to be 38%, plus or minus 150 basis points. The impact of slightly lower revenue and the less favorable product mix within the probecart segment is forecasted to be offset by increased higher margin system segment revenues and lower projected E&O reserve. At the midpoint of these outlook ranges, we expect Q2 operating expenses to be $52 million, plus or minus $1 million, with the $1 to $2 million increases compared to Q1, mainly due to additional investments in R&D. Non-GAAP earnings per fully diluted share for Q2 is expected to be $0.12, plus or minus $0.04. Reconciliation of our GAAP to non-GAAP Q2 outlook is available in the Investor Relations section of our website and in our press release issued today. With that, let's open the call for questions. Operator?

Thank you. To ask a question, you will need to press star 1-1 on your telephone. Again, that's star 1-1 to ask a question. We ask that you limit yourself to two questions and then re-queue. Please stand by while we compile the Q&A roster. Our first question comes from the line of Brian Chin of Stiefel. Your question please, Brian.

Hi there. Thank you. Good afternoon and appreciate you letting us ask a few questions. Maybe to start with, on gross margins, I know that NICS plays a role certainly in how gross margins trend from here, but what quarterly revenue level do you need to bring gross margins to sort of the 40% or let's say low 40% level? And I know, again, the second part of that is that although all caveats typically apply in terms of limitations to your visibility, what would be the more likely drivers that would drive a sequential pickup that's needed to get to those revenue levels or maybe even beyond in the second half of the year?

Sure, Brian. I'll take that. I think if I try to draw a trend line, I would say that the current revenue levels, $160 something, $165 million a quarter, we expect to be at the high 30s. We saw 38.4%. We're talking about midpoint of 38% in Q2. If you go back to Q3 of last year, once we get back to these revenue levels of around 180, we expect gross margin to be around 42%. And once we get to the 200 million level on a quarterly basis, we expect to see gross margins around our target model of 47%, as we saw in the first half of 2022. We need that increase to come mainly from Foundry and Logic, which is our higher margin market within the prop cost business, to support these increases over this trend line. It's going to continue to fluctuate based on mix, as you noted, but this trend line, I think, represents our thinking on this subject.

Okay.

And maybe even to expand on that, and maybe also pull Mike in here as well, I guess the commentary you saw, you know, a little bit of pull-ins, pull-forward in Q1, you know, drove some higher revenue, additional sequential for Foundry Logic. There's a right way to think about that. If some of these initial ramps that those probe cards are against in terms of, you know, flagship smartphones, you know, PCs, et cetera, if those ramps progress, then there's potential for sort of follow-on orders or activity that could happen maybe in 3Q, and that would sort of be the catalyst to start to drive revenues higher in the second half of the year? What's kind of maybe the right way to think about how revenue can move from here?

Yeah, Brian, it's Mike. I'll take that. You know, the pull-ins from what we currently thought were going to be Q2 into Q1 that led to revenue above the top end of the range really came from three places. One, you know, we have some very short lead time products. RF is a good example where there's perishable demands. to meet a specific customer wafer out schedule. And if we can ship it on time, we get the business. If we can't ship it on time, we don't. And so those are ones where we had a little bit more in Q1 than we thought we might. We had a few specific customer requests to accelerate delivery on specific designs, mostly associated with the ramps that you talked about. There's some pretty prominent ramps in the industry, flagship smartphone APs, a tile-based client microprocessor that are going on right now. And some of those wafer out schedules accelerated a little bit. And we had some recognition of deferred revenue in Q1 that we didn't think we'd get until Q2. You know, as we look forward now to the second half, as you noted, we have very limited visibility given the lead times we have. But I think if we see in demand in some of these areas, whether they be mobile handsets, whether they be client PCs, start to pick up a little bit, and our customers start to more aggressively ramp those particular devices, I think we could see some upside. Right now, we're, as we said, reasonably pleased that we've sized the business for the current levels, which, as we go through the middle part of the year, appear to be kind of a baseline stable demand level that we're operating to. But the factors you talked about could lead to some upside. We don't have any visibility of that upside at this point, But at longer term, it's part of the reason why we're continuing to invest in capacity in R&D. Okay.

All right. That's helpful. Thank you. Thanks. Thank you.

Our next question comes from the line of Tom Disley of DA Davidson. Your line is open, Tom.

Yeah. Good afternoon. I appreciate the chance to ask some questions. I guess, first of all, it sounds like you had some really strong design activity during the quarter, and I'm curious, are the new designs more design-intensive on your part going forward, and are you capacity-constrained on design at this point?

Yeah, Tom, we did make the point that our design teams are very busy working with customer test engineering teams, essentially on the next generation of devices, things like high bandwidth memory, DDR5, and some of the three nanometer foundry parts. This is essentially feeding the future for us. As you know, probe cards are specific to each customer chip design. And although things are not ramping in significant volume right now, all of this design work that's going on, I wouldn't cast it as being much more complex than we've had for the past couple of years, but our design teams are very busy across the customer and markets, essentially getting first articles ready, making sure our technology will work for these future designs, and shipping a handful of probe cards that then customers can validate so that once these new designs do ramp in volume, when we get to an upturn in growth in the industry, we're able to produce them and produce them in a high-quality, predictable way.

So how much of your activity is based on designing to a node versus designing to a very specific chip itself?

It's all the latter, right? You basically get qualified for a node or a set of requirements. Three nanometer foundry is a good example. And you need to have the right mechanical and electrical characteristics to essentially accommodate the envelope of different specific chip designs that are going to be produced on that node. But then our design team goes to work with the foundry, with the Fabless customer to customize that architecture, that MEMS probe architecture to each individual chip design so that we're ready to go when those individual chip designs, for example, you know, a three nanometer application processor that's ramping in the middle part of the year. We're able to produce that probe card, as I said, in a high quality, predictable way. And the architecture is validated on it.

Right. Very helpful. And then from follow up, moving to China, how big was China for you in the quarter? And it sounds like you expect that business to kind of decline over time. What kind of tail do you think there is there?

Yeah, it's an interesting question. So we do post our regional revenues in on our investor relations website. So you can go look at how the world broke down. I think we were something like 27 million in China in the quarter. Now, we haven't been that low in quite a while. That's composed, as it always is, of two components. One is shipments to the multinationals that operate in the region. We have some large customers, large global customers, who operate both memory and foundry and logic operations in the region. And so they'll have a ship to those factories, depending on where their production activity is. know the part i'm really referring to uh declining over time is supporting the domestic china customers and the timeline for that um you know we're still trying to figure out clearly with the export controls export restrictions with us as a u.s supplier we're limited in our ability to provide and support leading-edge advanced semiconductor uh projects and really form factors leverage to advanced nodes and the leading edge. If you think about our systems business, we're enabling new capabilities there. The probe card business really has high test intensity on leading edge nodes, new designs, complex designs, where customers don't have great yields and need to improve those yields. And so where some of the WFE manufacturers have been talking about trailing edge opportunity in China, We're not exposed to a lot of that. And so I do think over the next couple of years, let's call it, we're going to see that business drop to pretty de minimis levels.

Okay. I appreciate your time, Mike.

Thanks, John.

Thank you. Our next question comes from the line of David Dooley of Steelhead. Please go ahead, David.

Yes, thanks for taking my question. I have a couple technology questions. As far as in your Foundry and Logic probe card business, I know one of your big customers now is in production with a chiplet module type product. I was wondering now that you've seen that in production, what sort of increases in probe card intensity are you seeing from that chiplet or module strategy, whatever you'd like to refer it to?

Yeah, this is one, you know, we've talked about in the past and I think some of the ATE manufacturers have talked about in the past as well. Typically on a like for like basis, we're seeing somewhere between a 20 and 25% increase in what we call test intensity. And loosely speaking, that's the number of probe cards required per good die out. And that comes from a couple of different reasons. One, fundamentally you're breaking what used to be one ship up into multiple chiplets. And so that requires an individual probe card for each of the chiplets. Now, if that was done perfectly efficiently, there wouldn't be an uplift there, but you know, you need buffer probe cards and different elements like that that do cause an increase. The more significant part of the increase really is associated with the move towards something close to what people call known good die. You can imagine if you're going to stack, eight chiplets together you want to have very high confidence that each of those chiplets is good before you commit it to the stack because it turns out there's not a lot of rework or redundancy in these in these chiplet architectures so customers are testing these chiplets much more comprehensively than they might test a single die package and that's really what's responsible for the increase in intensity of something like 20 to 25 percent that we're seeing

Okay, as my follow up along the same lines, I think you referred to it in your prepared remarks where you're seeing kind of a large increase in probe intensity in the high bandwidth memory, specifically for AI type applications. I think Micron might have said five or seven times increase in the number of memory devices in one of the servers for an AI server or application. Could you just talk or elaborate a little bit more about that?

Sure. Well, HBM has been a pretty interesting opportunity for us over the years. We're now working on new designs for the third generation of HBM. And as you know, we've had a couple of our customers talk about the demand for high bandwidth memory in AI modules. So it's going to be a driver as we move through this adoption cycle of AI. The interesting part for us as a provider of test consumables and probe cards is that HBM is built essentially on a chiplet architecture. You have multiple DRAM die up to 16 DRAM die that are stacked up to build this end HBM product. And each of those die needs to be individually tested. Again, the same theme associated with something close to known good die being required for each of them. And you have extra insertions just because customers want to check the stack as it gets built up over, you know, as it goes from four die to eight die to 16 die. And so you've seen our DRAM in previous quarters in past years, our DRAM revenues get up close to and exceed $40 million. A lot of that has been driven by adoption and ramps associated with HBM, with high bandwidth memory. So another interesting opportunity where advanced packaging driving probe card intensity up.

And I'm sorry, just as a follow up to that, do you think that your high bandwidth memory portion of your memory business is going to see a big uptick from AI since there's so much more per server?

It's hard to say at this point, right? We're early in an adoption curve. There's no question directionally, you know, that AI requires more high bandwidth memory. We talked about one of our customers talking about a 50% growth rate in their HBM business. But, you know, how much that is in relation to our overall DRAM probe card business and what that adoption curve looks like over time, I think are still open questions that we're going to see.

But there's no question the direction is positive for us. Thank you. Thank you. Thank you.

Our next question comes from the line of Christian Schwab of Craig Harlem. Your question, please, Christian.

Hey, great. Thanks, guys. Just a quick follow-up on that, Mike, on the high bandwidth memory. Okay. If your largest memory customer continues to dominate that market, you know, obviously we're taking off 20% of production and then, you know, after they get supplied in demand and balance in the memory market, you know, then they're going to increase utilization rates and they're going to tweak the equipment before they buy a ton more capacity. So probably a longer cycle, right? But on the high bandwidth memory side, your commentary, you know, should AI accelerate as some people assume it does? I mean, you know, could that leading customer, you know, drive your DRAM business back into the 30 million category, even if we're still in a lull in memory capex spending for adding new wafers?

I think there's that possibility. Again, we're seeing and forecasting a sequential uptick in our DRAM revenue here in the second quarter off pretty low levels. I think in round numbers, we were about $20 million in DRAM revenue for the first quarter, expecting to come up off of that. But obviously, WFE is not coming up in Q1 to Q2. And it's another interesting example where Probe Cards is a design-specific piece of tooling, a design-specific consumable, can be decoupled from WFE. If everything goes right and we do see a strong ramp associated with HBM and it becomes a major part of DRAM wafer starts, sure, even in a cyclical memory downturn, that could provide enough uplift to get us up into the 30s per quarter. But again, it's not something, we're not calling a turn in DRAM here, we're just highlighting one of the interesting opportunities that we have here in the second quarter and in front of us that really is helping our overall probe card revenue.

All right. And the other thing that's kind of helping you, correct me if I'm wrong, in DRAM is the fact that, you know, we finally have working silicon from everybody out there for DDR5, right?

Yes.

For DDR5 for... I don't know, a long time. But now we finally have the working silicon out there, you know, which should certainly will accelerate the movement to there because the memory prices for the manufacturers will be materially higher than DDR4. So that's more wind at the back. Is that more important than HBM or is HBM more important?

I feel like HBM is more important because of the test intensity associated with the advanced packaging structure that HBM is built on. You know, if we can see that become a significant part of wafer starts driven by AI, I think there's a lot more uplift there. You know, DDR5, that transition is good. As you note, you know, the ASPs are higher, but it has been probably the slowest data rate transition, DRAM architecture transition in my career. We're still doing a reasonable amount of DDR4 in both mobile server and PC on DDR4. Okay.

And then my last question, you know, talking about slow migrations. Your largest customer appears to possibly have fixed their ALs. And it's in the process of rolling out five new generations of chips. in four years versus trying to have a 14 nanometer chip compete with a 7 nanometer chip and lose market share everywhere. So, you know, I saw that you guys saw some sequential growth there. And in your prepared comments, you suggested that certain trends in the industry, you know, you could surpass your previous model expectations. it's just kind of math. If they're successful, that's materially good news for you, correct?

Sure, of course. You know, we have worked hard to diversify our customer base in business by growing elsewhere and growing in different segments. The systems business, a good example, now up, you know, in the neighborhood and above $40 million a quarter. But How our largest customer goes, right, is obviously a key driver for form factor success. I'd also argue it's pretty important for the industry, right? They've always been a technology pioneer, a technology leader. And as they work on things like tile-based architectures, you know, I think the whole industry will benefit if they're able to accelerate and really execute on pushing the Foundry and Logic technology platform forward.

Great. Perfect. No questions. Thanks. Thanks, Christian.

Thank you. Our next question comes from the line of Craig Ellis of B. Riley. Your question, please, Craig.

Yeah, thanks for taking the question, and thanks for all the colors so far, guys. I wanted to start off just picking up on the last question with the largest customers. It was great to see the On a mix increase about 350 basis points quarter on quarter. The question, Mike, is of that 350 basis points and knowing that you are trying to win back share at that customer this year, how much of the increase in mix is that customer just executing better on their own roadmaps, more consistently to the prior point, more consistent with their nodes, and how much of that would have been share gains? or form in programs that you may not have had last year, maybe new programs this year?

Yeah, I think market share across the board has become a very important initiative for us, right? It's no secret, certainly to the people on this call, that we lost share in Founder & Logic, our biggest business in 2022, to our primary competitor. But it's also, we've been pretty transparent with you that we've recommitted ourselves to Sharegate. Made some organizational changes, some people changes, some compensation changes to really focus the team on market share. As you note, in Q1, we did see a good improvement there. But it's been our experience over the years that market share is not a one-quarter event. You can have different designs that you've won, different designs your competitor won. that can shift share in any given quarter, depending on who's spending and who's the incumbent. So, you know, this is going to continue to be an area of focus, reform factor, both at our largest customer, but across the customer base as we continue to compete with our major customers, our major competitors in founding logic, in memory, and in systems.

So is it fair to say, Mike, that to the extent there was short gain in 1Q, that you think there's still A considerable ways to go till you're happy with the share you recaptured to that customer.

Absolutely.

Right.

I think, you know, the way the way we think about it is you need to make small steps forward. Seems like there was some progress forward in Q1, but in no way would I call us satisfied with what we've accomplished. This is very much a work in progress.

OK, clear message. And then the second question, I know there's been a lot of inquiry on DRAM. I'm going to try and frame it up a different way. and just see how you might reply. So at least in my model, and the model could be off, DRAM revenues in the first quarter were at seven-year lows. And frankly, fundamentals in the DRAM industry are probably at 20-year lows. So revenues are outperforming industry, but we're at historically low levels. The question is, this is, as you look at the potential of getting back to more normalized high twenties to mid thirties level, can you just, um, can you just talk a little bit more about how you'd build the layer cake? You know, high bandwidth memory, I think is a pretty small part of total sort of server DRAM memory, but obviously very pro part intensive. How much is that? How much is just getting a lot more DDR five volume? on PCs and into smartphones and how much is going to be other things? And not looking for particular time, but as you see getting back to a normalized shipment level, how do the pieces stack up?

Yeah, I think if you build the layer cake, as you said, it's going to require a couple of things. Obviously, DRAM in markets are very weak, and I think it's clear that all of our customers are under-shifting in demand to try and consume some of the persistently high inventory that's still out there. So the first thing that has to happen is that DRAM supply and demand get back into balance and essentially our customers start to ship to the actual demand that's out there. They all seem to have a slightly different view of when that's going to happen, but that very much is probably the first piece of the layer cake in getting back to the mid-thirties. The second piece, you know, some of the stuff that Christian talked about in his question, DDR5 will drive new memory designs. New memory designs require new fleets of probe cards. And certainly there's DDR5 in the mix, a significant amount of DDR5 in the mix right now. But I think more DDR5 is probably a piece to that if we see adoption of that take off a little bit. And then the third piece is we stack up the layer cake. And the part that you probably inferred from my comments I'm most excited about is HBM. And it's exciting because it is yet another example of the industry moving to backend integration and advanced packaging to drive its innovation roadmap forward. It's cool to be helping enable AI, which is a fundamentally game-changing technology. technology for the world and semiconductors are underpinning it. And then finally, you know, when you think about our advanced packaging, and as you noted, the high probe card intensity associated with advanced packaging, it's an outsized opportunity for us on a per wafer start basis. So I think it's probably the icing on the cake right now, but probably becomes a bigger part of the cake if all these things line up for us.

Got it. And then the last question, just looking at the supplemental materials, I noticed Samsung wasn't listed as one of the 10% revenue customers. Is that simply because they haven't been such over the last five quarters or was there any other reason why they dropped off? And can you just comment on how you're feeling about both on the DRAM and then on the FoundryLogic side of that customer's business?

Yeah, we had a few customers that have been on the 10% list that were awfully close, right, in Q1. And just the way the numbers worked out, they didn't make the 10% list. So I wouldn't interpret it as anything significant. Getting back to the share topic, the comments I made about share and founder and logic apply across the board. We've refocused ourselves, refocused our processes, our incentives. on making sure that we're delivering technology that drives share gains, right? These are segments where you only have really two suppliers that are able to deliver into the advanced requirements. And, you know, those are, so we're, we're investing in R and D. So we have compelling products that we can charge a premium for drive gross margins up, deliver value to our customers and gain share at the same time.

Got it. Clear message. Thanks for the help, Mike. Thanks, Chris.

Thank you. Again, to ask a question, please press star 11 on your telephone. Again, that's star 11 on your telephone to ask a question.

Thank you.

Our next question comes from the line of Krish Sankar of Cohen. Your line is open, Krish.

Hi, this is Robert Mertens on for Krish. Thank you for taking my questions and congrats on navigating through a difficult industry environment. I guess just to start off, I know we've talked a bit about it, but on the gross margin side of things, it looks like you're expecting 40 basis point contraction to the margins in the June quarter with sales slightly down in the mixed shift, more DRAM and less the foundry logic. compared to the first quarter. Could you just give some of the puts and takes into the June quarter gross margin guide? Are there any incremental margin drivers expected from the cost reduction plan or manufacturing efficiencies that are baked in there?

So you listed most of the items. The restructuring cost reduction is already baked into Q1, so continue to impact us in Q2. We're talking about similar production levels, so no changes there. The things that move the 38.4 in Q1 to a midpoint of 38 in Q2 is the factors you talked about. I can repeat them. We're talking about overall slightly lower revenue. 162 is the midpoint of the range versus 167. So that has a negative impact. The lower foundry and logic and higher memory also have a negative impact. And these are partially offset by the fact that we expect the record systems revenue, which means a higher margin impact because the system segment has the highest gross margin for us. And we also expect less E&O in Q2 after high E&Os in Q4 of last year and still elevated in Q1, but kind of normalized in Q2.

Great. Got it. And then just one more if I can. In terms of the systems business, obviously it's been performing really well and capitalizing on the lab-to-fab strategy. I was just trying to see if there's any way to parse out the current demand strength from a handful of technologies or specific R&D projects, or is it sort of more broad-based and a lot of different components in that side of the business? Thanks. Sure.

The systems business does tend to be pretty broad-based. You're dealing thematically with new technologies. We've talked on this call about silicon photonics. Last call, I think we talked about silicon carbide a little bit. Our investments in cryogenics for quantum computing tests are in that bucket, metrology and inspection for advanced packaging in there as well. So it's a pretty broad spectrum of applications. But the common theme to them is you're working with customers in their very early development, things that won't come to production, you know, for several years from now. And, you know, that gives us, I think, some strategic value out of the systems business because we get foresight into some of these applications which are going to go to production and get us better equipped to deal with them when they get to the fab. You know, things like CMOS image sensors, Vixels have been nice examples of that. where we've taken systems business engagement from the systems business and turned them into production probe card products. You know, the other piece obviously is, as Shai noted, it's a pretty good financial contributor and we've seen some, the nice growth out of it, but nice growth with good profitability, especially on the gross margin line, help buffer the cyclical downturn in the overall semiconductor industry and the probe card business, which is heavily levered to, new design activity and ramping that new design activity, which, you know, given the current downturn across the industry, customers have pulled back pretty significantly on ramping new products.

Got it. Thank you. That's very helpful. Thanks. Thank you.

Our next question comes from the line of Charles Shee of Needham. Your question, please, Charles.

Oh, thank you. Thank you for taking my question. Hey, Mike and Shia, I think you went through several very detailed questions about quite a few corners of your business with various analysts asking questions ahead of me. I'm going to ask a little bit higher level question. I think in the second half last year, On the way down, in terms of the business, you mentioned about design starts, design projects getting optimized, some got reduced, and the existing, the surviving design starts, actually the volume got adjusted downward. So just want to ask you as to where we stand in early May. What do you see in terms of those two vectors, the design starts and the volume specific to each design starts? Are design starts coming back? Obviously, volume may not come back yet, but I just want to get some thoughts on that regard. Yeah, thank you.

Yeah, it's an interesting question, Charles, and we tried to touch on it in the prepared remarks. we are seeing very strong new design activity. So this is where our design teams are working with customer test engineering teams to make sure we have first articles, validated probe cards for when they do ramp their new designs in volume. Now, you know, the things that are ramping in volume, you can count on, well, there's not a lot of them and they're not ramping at high volume, as you can see from overall industry levels from, our customers' narratives from our revenue. But we wanted to point out to people that we have seen a return of this new design activity. Even though these things aren't ramping in volume, they position us well for when the industry does return to growth to really be ready to ramp these new designs in volume. The quantitative piece of it, our design teams, which we really didn't downsize at all, in our restructuring in Q3 are operating at full capacity. And so that's a very different situation than it was in the second half, as you say, on the way down. Strong design activity. We're going to be ready when these designs ramp, but they are not ramping now. And as we said, we feel like we're in a trough through the middle part of the year. When growth does return to the industry, if it's late this year or next year, I'm pretty optimistic that we've won some designs that are going to help drive our revenue forward towards the target model.

Thanks, Mike. I want to ask specifically about the part of your revenue exposed to mobile. Obviously, I want to include a discussion around microprocessor and high-performance compute here for a bit. On the mobile side, I think I did hear you some pull-ins for an RF side of your business, but any other green shoots are you seeing right now in the SOC side of the mobile business for you? Maybe let's exclude the iOS side of the business for a bit because there seems to be some seasonality in Q4, Q1, but going into Q2, do you see some moderate recovery in the mobile business, either on the SOC side or RF side? Thank you.

Yeah, so mobile for us is pretty broad-based, and if you look at form factor over the past several years, Mobile handset growth and 5G adoption have been a big driver for us. Obviously, that's pretty weak right now, both from a component standpoint, but also from an end handset market standpoint. One of the things we're working on pretty aggressively here in the first part of the year is making sure that we've got the technology in place to ramp on some of the major mobile application processor designs in the industry. You know, as they move to three nanometer foundry platforms, there's some new requirements there that we've had to deliver. Some of that was a contributor to Q1. We said we expected a little bit of digestion here in Q2, given the volumes we ship. But I think more broadly, we need to see whether it's SOC or application processor or the RF component side. I think we need to see a return to broad based handset growth. in both the Android and the iOS ecosystems to really drive that business forward.

Got it.

Maybe lastly, if I may, Mike, the DRAM business, I thought you were expecting another lockdown in the middle of the year when a quarter ago, but looks like that lockdown is already here looking at the DRAM revenue numbers. Just want to ask, are you still expecting another lockdown or you think maybe Q1 would be the low watermark for the DRAM business? Maybe the modest growth into Q2, there is more lack to go into the second half of the year. Thank you.

Yeah, we're not, to be clear, we're not calling a turn in DRAM. You know, this really, the move, the sequential move up Q1 off a very weak Q1 into Q2 is really associated with some of the specific DDR5 and HBM activity that we talked about. You know, with the lead times that we have and with the trajectory of inventory consumption and new product releases and ramps in the DRAM industry, I think the jury's still out on whether You know, this is now a return to growth. At best case, it's gradual growth. Feels to me more like we're going to be kind of moving along the bottom of a trough here until the inventory situation gets in better shape with our customers.

Thanks, Michael. As always, for the great colleagues. Thank you. Thanks, Charles.

Thank you. Once again, to ask a question, please press star 1-1 on your telephone. Again, that's star 1-1 on your telephone to ask a question. Our next question comes from the line of Gus Richard of Northland Capital Markets. Your question, please, Gus.

Yes, thanks for taking the questions. You talked a little bit about the transition to three nanometer, and I'm just wondering, you know, your designers have made Sort of intersecting, is that transition broadening out beyond apps processors to AI ASICs or GPUs, CPUs? Can you just talk about the mix and does apps processors have more or less probe card intensity than some of the other types of silicon?

Yeah, 3 nanometer right now for us is pretty apps processor focused. On the intensity question, it really depends on the details of how the customer's testing and what their overall test strategy is. A great example of that, I'll go back and tie it to advanced packaging, right? If, you know, three nanometers used for a set of chiplets, that's going to have relatively higher test intensity, whether it's an HPC part or a mobile part. compared to a monolithic dive because of the trend towards known good dive that we talked about earlier in the Q&A session. So, you know, I wouldn't say there's a broad-based rule of thumb for HPC being more intensive than apps processors or vice versa. But I would say that, you know, if any of those chiplets end up in advanced packages to drive then the end product, we are seeing a significant increase in test intensity.

Got it. And then in your systems business, I think you highlighted silicon saponics in your prepared comments. And I'm just wondering, you know, how close is that moving to commercialization? You know, is a handful of customers or is it just light bender? You know, any color around, you know, how you see that unfolding over the next, you know, six to 18 months?

Yeah, well, that's probably reaching, you know, there are a couple of customers that are already in production. Now they're pretty niche applications. I think the exciting part for us is we're seeing more and more potential customers in this market as people pursue, you know, the value of co-package optics and integrating them into data center architectures. The power benefits, the speed benefits are awfully compelling. But, you know, if you want to look at it, you know, when's it going to materially affect our customers' P&Ls and probably our P&Ls, that's going to be closer to 18 months than it is six months in the timeframe you gave me. Okay.

Okay. And it's a handful of customers that are doing this at this point.

Yes.

Very good. Thank you so much.

Thanks, Gus.

Thank you. I would now like to turn the conference back to Mike Schlesser for closing remarks.

Great. Thanks, everybody, for joining us today. Thanks for some of the questions that allowed us to describe our view of the industry and its current state. We're signed up to do several sell-side conferences in the month of May and June, so we hope to see you at one of those and be able to answer more questions and update you on form factor progress as we work through the cyclical downturn. Have a great day and take care.

This concludes today's conference call. Thank you for participating. You may now disconnect.