Thank you, and welcome, everyone to FormFactor's Fourth Quarter 2024 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 include 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 new technologies; the impact of global, regional and national health crises, including the COVID-19 pandemic; anticipated industry trends; potential disruptions in our supply chain; the impacts of regulatory changes, including the recent U.S.-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 risks 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 filings on Form 10-K with the SEC for the fiscal year ended December 30, 2023, 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, February 5, 2025, and we assume no obligation to update them. With that, we will now turn the call over to FormFactor CEO, Mike Slessor.

Thanks, everyone for joining us today. As expected, FormFactor reported sequentially lower fourth quarter revenue, gross margin and non-GAAP earnings per share, driven by the forecasted reduction in Foundry and Logic probe-card revenue. This was partially offset by growth in DRAM probe-card revenue, which set a third consecutive quarterly record. Like most broad-based equipment and consumable suppliers to the semiconductor industry, FormFactor's fourth quarter and full year 2024 results reflect two very different market dynamics. On the one hand, the explosive growth of generative AI and high bandwidth memory, which more than quadrupled over 2023 and added $100 million in revenue for FormFactor. On the other hand, lackluster demand persisted in important high unit volume markets like client PCs and mobile handsets. This overall theme is continuing in the first quarter of 2025 and we expect sequential reductions in demand for both non-HBM DRAM probe-cards and systems. As we move through 2025, we do expect an overall increase in demand for FormFactor's products and I'll provide some more details on these growth areas later. Before diving into segment and market-level details, I'd like to highlight two significant strategic developments that further strengthen FormFactor's leadership position in enabling the adoption of advanced packaging. Earlier today, we announced that together with MBK Partners, the largest private equity firm in North Asia, we're acquiring FICT Limited, a leading supplier of complex multi-layer organic substrates and printed circuit boards to semiconductor test and high-performance computing customers. This acquisition will solidify FormFactor's access to FICT's technologies and products, which are an important component of our advanced probe cards. And early last month, we announced technology partnership and share purchase agreements with Advantest Corporation, the market leader in automated test equipment. These agreements build upon a multi-year collaboration between our companies, including FormFactor's 2020 purchase of…

Thank you, Mike, and good afternoon. As you saw in our press release, Q4 revenues were $189.5 million, $0.5 million below the midpoint of our outlook range, and non-GAAP gross margin of 40.2% was 0.8 percentage points below the midpoint of the range. These, together with operating expenses slightly lower than the midpoint of the outlook, resulted in a non-GAAP EPS of $0.27, $0.02 below the midpoint of our outlook range. Fourth quarter revenues decreased 9% from the all-time record revenues reached in the third quarter and increased 12.7% year-over-year from our Q4 '23 revenues. For fiscal '24, we recorded revenues of $764 million, up 15.2% from $663 million in fiscal '23. Almost all of this $100 million growth came from HBM revenues. Probe-card segment revenues were $150.3 million in the fourth quarter, a decrease of $21.9 million, or 12.7% in the third quarter. The decrease was driven by lower Foundry and Logic and Flash revenues, partially offset by a record-high DRAM revenue. For fiscal '24, we recorded probe-card segment revenues of $626 million compared to $498 million in fiscal '23. The system segment revenues were $39.2 million in Q4, $3.4 million higher than the third quarter, and comprised 20.7% of total company revenues, up from 17.2% in the third quarter. In fiscal '24, we recorded system segment revenues of $137.6 million, $27.6 million lower than the $165.2 million revenues in fiscal '23 with a decline due to the sale of our FRT business during the fourth quarter of fiscal '23. Within the probe-card segment, Q4 Foundry and Logic revenues were $83 million, a 22.5% decrease from the third quarter. Foundry and Logic revenues decreased to 44% of total company revenues compared to 51.7% in the third quarter. DRAM revenues were a record $63.3 million in Q4, $3.1 million, or…

[Operator Instructions] Our first question comes from the line of Charles Shi from Needham & Company.

Mike, Shai, thank you for letting me ask a couple of questions. I think you provided segment color for your Q1 outlook, I think, I mind if you kind of unpack a little bit the Foundry & Logic, DRAM between non-HBM and HBM -- and of course, you also mentioned about system and we will unpack that a little bit for us. But I do want to specifically ask you about the non-HBM DRAM probe cards since it looks like you're anticipating a sequential reduction in demand. But if I really back out 6 CSMT, the China demand that was there before the actual controls, if I back that out, your DRAM revenue, non-HBM DRAM, I think has been run rating at $1 million per quarter for almost 2 years, almost like there's no recovery in that part of the market. Mind if you educate us a little bit of what's going on there and when could we see the outlook for that part of the market to become a little bit more positive?

Sure, Charles. So as we bridge from Q4 to Q1, you've identified the biggest factor in the sequential decline between our Q4 results and our Q1 outlook, and that's non HBM DRAM. We expect HBM DRAM to be approximately flat quarter-on-quarter and by far, the largest component of the non-HBM-DRAM reduction in Q1 is associated with advanced DRAM probe cards shipping into China. If you look at our results in the second half of 2024, you saw some significant growth there, and it was primarily associated with DRAM probe cards. With the recent export controls that have gone into place from the U.S. Department of Commerce, we don't believe we can ship to those customers. And as a result, we've taken our outlook for DRAM probe card revenue into China down to zero in the Q1 outlook, and we'd expect to have that persist throughout the year unless there's some dramatic change. That's a pretty significant impact. We were running in round numbers at about $10 million per quarter with that customer with China DRAM and advanced probe cards. So that's definitely a headwind. Some of the other pieces, we often see seasonality in the systems business. So a Q4 to Q1 step down is not all that unusual given our customers' budget cycles. And then we see primarily a flat outlook Q4 to Q1 overall in the Foundry & Logic business. There's a few moving parts underneath that with various customers and various projects, as there always is inside any segment. But by far, the biggest component causing the Q4 to Q1 sequential reduction is the removal of non-HBM DRAM into China.

Yes. Sorry, maybe I asked the question a little bit too long, but the second part of the question, I do want to repeat here. In non-China non-HBM business?

Okay. The rest of the world.

Yes.

Yes. DRAM, let's continue to call it, non-HBM DRAM. So primarily DDR5 and some graphics DDR as well continues to operate at those cyclical low levels, around $20 million a quarter, and we don't see much activity there. There's new designs as our customers continue to release and ramp in limited volume, especially associated with DDR5. But as you noted for the past couple of years that business has been operating at cyclical lows. HBM continues to be the highlight in the DRAM probe card business and really is what powered the three consecutive record quarter.

And our next question comes from the line of Craig Ellis from B. Riley Securities.

Mike, I wanted to start off just with a higher-level question. I think over time, as you and I have interacted on the business, 1 of the things we've talked about is the business being very product cycle and node transition dependent with its product demand and much less volume dependent and yet it seems like there is quite a bit of volume impacting the business' REIT exit 2024 and going to 2025. So can you just come back and kind of square that circle? Why is it that volume is such a big impacting factor in the very near term here?

Sure. So for probe cards, the fundamental demand driver is new design releases by our customers. But those new design releases need to ramp in volume, Right? what drives the fundamental change from one set of probe cards to another is the design change or the mass change, things like going from HBM3 to HBM4 or a node transition from 3-nanometer to 2-nanometer those things definitely drive probe card demand but also individual designs inside those nodes also drive proper demand. But the number of probe cards we sell, right, is dependent on the ramp volume associated with that. How many test cells are going to need to run, each of them needs a probe card. And so it's both. One thing that drives the refresh cycle, no question about it and why we're bullish about 2025 is things like the transition to HBM4, which we expect to drive growth in 2025 over 2024. But we also need all of these things to ramp in volume. The wafer volume, the test cell volume drives the number of probe cards. And so with markets like PC and mobile operating at continued cyclical lows, we're not getting the volume out of those markets that really drive the business.

That's helpful color. And then as a follow-up, which is a take-off really on that last point, there's a large U.S. customer that did not appear on the 10% list in the fourth quarter. And from time-to-time, they just go through a period of probe card digestion. Can you address to what extent was that at play versus the demand levels in the prior couple of quarters versus their own share loss that may be impacting the volumes that would drive business with FORM or other issues?

Yes. I think it's fair to say that we see fairly broad-based weakness across Foundry and Logic. But certainly, our microprocessor business, as you noted, is especially weak in the fourth quarter. No Foundry & Logic customer, including our large microprocessor company appeared as 10% of revenue. And so I think that's indicative of the weakness there. As we look at some of the initiatives for the company, obviously, that raises the importance -- increases the importance of something we've been committed to for a long time, and that's making sure we're qualified at all the leading customers across the industry. Now you've heard from all the customers in the x86 space that client PC is awfully weak. So I don't know that would have made a really big difference to either our Q4 or Q1 outlook, but it remains a strategic imperative for us to qualify and be competing for business at all major customers. And as I said in the prepared remarks, we're making some good progress there and expect those things to positively impact FormFactor's growth in 2025.

And our next question comes from the line of Brian Chin from Stifel.

Well, I guess, in terms of the statement that you do expect an overall increase in demand as we move through the year, it sounds like you're saying that you do think HBM will be one of those drivers sequentially through the year. And when you think about the foundry logic business, is it reasonable to expect that even last year was not exactly a banner year in terms of phone and PC unit sales, for example, but there was release of new designs and levels of seasonality in production? Would it be reasonable to expect that you can get back to sort of that $100 million a quarter kind of run rate in foundry logic if we have a similar kind of profile this year? And that's also incorporating, I guess, some of the progress you're seeing in terms of fabulous microprocessor and GPU?

Yes. I think about -- if I look at 2025 overall and the reason we're confident in growth, HBM and the transition to HBM4 is definitely one component. We're going to higher layer stacks, which amplifies the test intensity going from 8 to 12 to 16 in stacks and then the higher test feeds amplify the test complexity. Those things historically have always ramped up guest spending and we saw it in 2024 with HBM3 and HBM. I think if we look at Foundry & Logic, it's certainly reasonable to assume a return to the let's call it, 100 million-ish levels with some activity associated with a PC refresh cycle, more activity and higher unit volumes of these new designs in PC and mobile. And so I think that's a reasonable place to think about the Foundry & Logic business returning with some Form Factor specific drivers being qualifications at some of these new customers adding to revenue and balancing and diversifying the overall customer base.

Okay. Great. And maybe asking about the FICT acquisition. I guess it sounds like they've been a key supplier to you historically. And in terms of the motivation here, is this reflective of just how fast those product cycle roadmaps are moving in the AI data center arena and kind of that need to sort of exert some influence over key ecosystem partners, maybe still with an eye towards not fully internalizing or verticalizing some of these partners?

Yes. I think that's a good way to characterize one of the objectives. FICT has been a very important supplier to us and the overall probe card industry over the years and remains so. And with the accelerating complexity associated with high-performance compute and what our customers are asking us to do, we felt like we wanted a larger influence for that important supplier. The structure we executed here with MBK Partners allows us to have that influence in that closer coupling without taking on all of the financial and execution risk of a full cross-border acquisition. So we think this is a good step forward to both solidify the supply chain, our supply chain but also offer us some increased access into advanced packaging. Runs a similar theme to some of the partnership announcements you've seen from Advantest, Advantest investing in us in and one of our competitors. Technoprobe investing -- or Teradyne investing in one of our competitors, you see these closer couplings across the industry just because of the complexity of what customers are requiring of us and the speed with which you need to deliver that complexity.

And our next question comes from the line of David Duley from Steelhead Securities.

As far as the HBM outlook goes, so I think I heard you say you have designs in hand or you should have revenue in hand from two customers more significant revenue from two customers now. So does that mean you have two customers ramping up in HBM4?

Well, we've got a mix of HBM3 and HBM4, but I think the point you're keying on that we now have a significant contribution in our HBM revenue from a second customer is correct, right? So I want to make sure that doesn't get lost. Our HBM revenue is still mirrors pretty well our end customers' HBM market share in the end market. Dominated by one, a second customer coming up and taking more share and producing more volume and the third one, we expect -- we're qualified and competing at all 3 of these. But our share, again, our contributions reflect their shares in the end market. But we've now added a significant second component to our HBM revenue in both HBM3 and HBM4. So it's still a mix. But we really expect HBM4 to ramp in significant volume as we go through the middle part of 2025.

Okay. And then I was curious, when you talk about qualifying new customers I'm assuming one of those is the CPU customer that you don't have. And then I think you've probably got qualifications or business in hand with the hyperscalers? Could you elaborate a little bit more on what's going on with the hyperscalers?

Yes. So this is an interesting new business. These are generally these custom ASICs that the hyperscalers are developing in-house sometimes with a design partner, sometimes directly with the foundry and we've got good engagement there. A couple of projects that we're shipping the initial designs for here in the first quarter but that's also a business that we expect to strengthen as we go through the year and kind of a new component of the overall probe card business. We used to just either deliver to the IDMs or the foundries, here's a case where the hyperscalers are taking these new chip designs, mostly custom ASICs for AI applications and starting to develop that silicon themselves. And so a new growth area for us, one that I wouldn't say is going to materially affect us in the first half, but we're laying the groundwork to make sure as those new designs ramp as we go through 2025 that we're in a strong incumbent position to capitalize on these new opportunities.

Okay. And just a follow-on to that. I'm just kind of curious as far as inside your foundry logic business, I'm sure that's where you find the networking piece. I think you have exposure to the [Avago] food chain there and perhaps the NVIDIA food chain. Could you just talk about what you're seeing in networking?

Yes. Networking, we do have good exposure there and that is an area where with those customers has offered some pretty significant growth over the past several quarters. Part of it is due to their product strategies. For example, you see some of these customers selling things like GPUs together with the networking components and so the system level view that our customers are taking kind of drags along the networking chips with some of the higher-value GPUs, and that's caused those volumes to grow significantly and their product roadmaps to accelerate significantly as high-performance compute gone to this 1-year roadmap cadence. So the networking business, although it's not the volumes of something like HBM continues to be a strong and growing part of our overall Foundry & Logic business despite the weakness in areas like D.C.

Okay. So it kind of sounds like you'll have a second half ramp in multiple different markets. And I guess I would assume that Q1 is going to be the bottom of revenue in this calendar year?

That's certainly the way we currently see it. And if I point you back to 2024, we saw a very similar Q1. I'm not going to imprint seasonality on this, but there are often lumpiness in this business as things go through different product cycles and ramp cycles and I'd take you back to 12 months ago, where we had a very similar and grew significantly over that. Remember, in 2024, we grew the top line 15% over 2023. I'm not suggesting that we're going to grow 15% again. But all of these factors, whether it's HBM4, PC recovery, some of the new customer qualifications, we do expect to drive year-over-year top-line growth for FormFactor.

[Operator Instructions] Our next question comes from the line of Krish Sankar from TD Cowen.

Hi, this is Robert Mertens on the line for Krish. One, just could you provide a bit more color on what you're seeing within the systems business? The last 3 quarters have done quite well with silicon photonics being highlighted but I know there are a lot of other R&D-driven portions of that business, whether it's silicon carbide or quantum. But looking at the March outlook, could you just sort of partition how much may be due to typical seasonality versus near-term in digestion and then just further out in the year, if there are any sort of areas within that business, that would be the most meaningful drivers to growth?

Yes. So the systems segment often does have this seasonality, again, because of our customers' annual budgeting process, often their fiscal years and calendar years online. One of the things that's really got us excited in the Systems segment is the transition of silicon photonics and co-packaged optics to pilot production. This is an area where we've been engaged in the lab with leading customers for a long time and have an installed base of over 100 systems. And we're leveraging that know-how and those customer relationships to work with partners like Advantest to drive this to high-volume production. We expect that to start to happen in the second half of 2025. We're working very actively with customers, with foundries right now on making sure we're ready for that ramp. I think if you interpolate our customers' roadmaps, it's probably really a 2026 volume event. But we definitely see some growth there. I think the other one to highlight is quantum, right? It's probably, again, not a 2025 event, but we're trying to build a leadership position in test and measurement for the quantum computing industry. It's in an investment phase right now for us and for the overall industry. And we are seeing some interesting opportunities and growth prospects there. The overall systems business growth is going to be driven by those factors but also our customers' continued investment in innovation and R&D in things like high power, in things like new silicon nodes. It's a very broad-based business. But I think probably the most exciting tangible growth opportunity in 2025 and into 2026 is co-packaged optics and silicon photonics.

Great. That's very helpful. And then maybe just if I can sneak one more in. I know we've gone over the high bandwidth memory quite a bit, but it looks like maybe the DRAM probe cards could benefit in the second half margin-wise with new non-high being with memory probe cards being introduced. Have you ever given out a breakdown sort of between what the margins of high bandwidth memory are versus more of the traditional DDR designs?

I will take this one. We didn't specify the differences between HBM gross margins and other DRAM gross margins. But we did say that HBM revenue because the cards are differentiated because our technology there is unique is as higher margins within the DRAM gross margin family not all the way up to the Foundry & Logic gross margin, although it could be similar or equal to some low-end Foundry & Logic design gross margin.

And our next question comes from the line of Tom Diffely from D.A. Davidson.

Shai, just following up on that previous question. If you see Foundry & Logic and maybe microprocessors being a smaller percentage of the business going forward. How impactful would that be to your kind of corporate gross margin going forward?

So it's a good reminder that we are committed to our target model. Our target model is 47% gross margin at $850 million of revenue. But we are currently delivering revenues with a very different mix than when we put our target model in place. Now there are a few things that we are working on to improve that together with a few things that really needs to happen in the industry for us to achieve the target model gross margin. We are developing, as Mike mentioned in the prepared remarks, a lower-cost DRAM architecture that can improve the DRAM gross margin, bring it closer to the Foundry & Logic margin. We're also targeting a higher market share in the higher-margin Foundry & Logic market. And we also have a few internal initiatives to improve margins all targeting improving the margins across the different products. But an important part is we need the overall end markets to recover since we need the higher volume to achieve our target model.

Okay. That seems to make sense. And switching gears here, Mike, when you look at the acquisition, FICT, is the main thrust there just to keep an uninterrupted supply or maybe exclusivity of what you're currently getting or is it really to kind of mold or bend the work going forward to what you really need to have to happen?

It's much more the latter. If we wanted exclusivity and vertical integration, we would have bought the whole thing. This is more about partnering to accelerate our product roadmap, FICT's product roadmap and as I said in response to one of the other questions, really make sure we're meeting the challenges associated with high-performance compute on the accelerating roadmap that high-performance compute is on. So the company is going to operate independently. We'll obviously have a seat at the table with a 20% ownership stake. But our goal is to grow that business and better align their future product roadmap with ours to produce more competitive products for FormFactor.

And so when you look at the design cycles for those type of products, if you get involved today, is that for products that are 1, 2 years out then?

Yes. I mean there's 2 design cycles nested inside each other. There's if you like, an architectural design cycle where we're meeting maybe requirements for, let's call it, 1 nanometer logic. But inside that, there's a whole bunch of multiple designs that need to be customized and the integration between FICT's products and our products is a much higher frequency design cycle. So 1 to 2 years for that nanometer transition, if you like, but much tighter cycles for collaborating on the individual customer design. So we should be able to exert some impact on this as we go through 2025.

And our next question comes from the line of Christian Schwab from Craig-Hallum.

I had a question on high bandwidth memory. There's varying third-party research on growth rate, but the one, I guess, I'm looking at 26%. Given increased capital intensity on HBM4, if that 26% is accurate, is it safe to assume that high bandwidth memory should grow even faster than that in '25 over '24?

Yes, Christian, I think that's a fair assumption, right? When you look at the dynamics of moving from HBM3 to HBM4, more layers, and that means higher test intensity. We've already seen that in the move from HBM2 to HBM3 to HBM3. And you can see that we added over $100 million in revenue associated with HBM in 2024. So a good proof point there. I think the other piece that goes into it is complexity, and complexity in moving from HBM3 to HBM4 really is higher test speeds, higher chip clock rates, which are a more complex probe card and a more valuable probe card. So we do expect that we're going to be able to outgrow the overall HBM industry when it's in this phase of ramping quickly with high complexity and the need for our customers to push parts of the door.

[Operator Instructions] And this does conclude the question-and-answer session of today's program. I'd like to hand the program back to Mike Slessor for any further remarks.

All right. Thanks, everybody, for joining us today. We'll talk to you soon.

Thank you, ladies and gentlemen, for your participation in today's conference. This does conclude today's conference call. You may now disconnect. Good day.