FORM - NASDAQ

Thank you, and welcome, everyone, to FormFactor's Third Quarter 2025 Earnings Conference Call. On today's call are Chief Executive Officer, Mike Slessor; and Chief Financial Officer, Aric McKinnis. 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 also with respect to the projections of financial and business performance, future macroeconomic and geopolitical conditions; the benefits of acquisitions and investments, including acquisition of manufacturing facility, anticipated industry trends, potential disruptions in our supply chain; the impacts of regulatory changes, including tariffs and changes in export controls; the anticipated volatility in 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 filing on Form 10-K with the SEC for the fiscal year ended December 28, 2024, 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, October 29, 2025, and we assume no obligation to update them. With that, we will now turn the call over to FormFactor's CEO, Mike Slessor. Sorry about that. Thanks for joining us today.

Thanks for joining us today. FormFactor's third quarter revenue, gross margin and earnings per share exceeded both second quarter results and the midpoint of our outlook range. Our outlook for the current fourth quarter builds on the third quarter and we expect to again deliver sequentially higher revenue, earnings and most importantly, gross margin. As you heard from us on last quarter's call and in meetings with many of you since then, we're focused on and committed to improving our profitability to get back on a path to the 47% non-GAAP gross margins of our target model. As you can see in our fourth quarter outlook, we expect to reach the model quarterly revenue run rate earlier than we achieved model gross margins. This disconnect is driving urgency across FormFactor in executing a program of rapid and immediate gross margin improvement actions that have already produced a 250 basis point increase from the second quarter, and we anticipate will produce an additional 100 basis point increase in the fourth quarter. We expect these short-term improvements will continue throughout next year, steadily closing the gap to the target model gross margin of 47%. At the same time, we're also executing longer-term structural initiatives that will further improve our gross margins, including developing and commercializing differentiated new products to drive increased market share and pricing as well as qualifying and ramping our new Farmers Branch, Texas facility to rapidly and cost effectively expand our capacity in a region with lower operating costs and a variety of financial and regulatory incentives. Aric McKinnis, who became our CFO on August 12, will provide more details on both the short- and long-term initiatives. While we're intently focused on improving gross margins to generate the profitability warranted by our leadership positions in both probe cards and engineering systems, we're also working to expand those positions, principally at the intersection of advanced packaging and high-performance compute. Innovations in this space like the stacking of DRAM chiplets to produce HBMs, which are then integrated with GPUs in multi reticle CoWos packages, co-packaged optics and even chiplet-based processors deployed at the edge are all driving increased test intensity and test complexity, creating increased demand in our served markets. In some of these areas, like HBM and DRAM and network switches and foundry and logic, we today have leading market share positions. In others, like GPUs, we're making steady progress on qualifications to produce market share gains and revenue growth. Turning now to segments and market level details. In DRAM probe cards, we delivered the expected double-digit sequential growth in the third quarter to a new record, primarily from growth in HBM. In the current fourth quarter, we expect to post another record primarily from an increase in non-HBM applications like DDR5 and LPDDR4, likely driven by the well-publicized recent increases in commodity DRAM end market demand pricing and customer profitability. In HBM, we've now reached the anticipated HBM3 to HBM4 crossover as HBM 3E ramps down and HBM4 ramps up. With the overall net result being total fourth quarter HBM revenue similar to the third quarter. We continue to have significant contributions from all 3 major HBM manufacturers as we execute our long-term strategy to be a key supplier to all the leading customers in the industry, thereby growing and diversifying our HBM demand profile. Consistent with the current market share split between our customers, however, our HBM revenue continues to be skewed towards our largest customer. The ramp-up of HBM4 offers some exciting opportunities for FormFactor as we look ahead to 2026. First, the test intensity for each HBM stack further increases with the transition to HBM4 16-high stacks of core die chiplets from the 8 and 12-high stacks of HBM3 and 3. As we've detailed in our past discussions of advanced packaging, each of these core die chiplets must be comprehensively tested to ensure that a single effective core die does not cause a failure of the entire stack. In addition, the I/O speeds of HBM4 increased substantially over HBM3, and our customers are now striving to exceed the JEDEC specification of 8 gigabits per second. This performance increase in greater test complexity drives competitive advantage for FormFactor as our SmartMatrix architecture is the industry's only production-proven high parallelism probe card architecture that can operate at these 10 gigabit plus frequencies, providing our customers with the unique capability to validate their product performance at these higher and more valuable I/O speeds. Shifting to the foundry and logic probe card market. As expected third quarter demand in this market was sequentially weaker than the second quarter, and we expect similar foundry and logic demand levels in the fourth quarter. Despite broader indications of the beginning of a PC recovery, we're not experiencing significant growth in probe cards for CPU applications. We believe this is because the increased demand is being served by our customers' existing legacy node designs, where they're able to employ their existing probe card fleet. As a reminder, probe cards are a device-specific consumable. And as this customer ramps volume on their new designs on new leading-edge silicon nodes, we expect to see increased demand for probe cards and CPU applications. This current situation also highlights the need to execute our strategy to be a key supplier to all the leading customers in the industry and we continue to build the foundation for market share gains at a large fabless CPU manufacturer. Having achieved qualification in a specific application with this customer earlier this year, we're now building on that penetration and qualifying our market-leading Apollo MEMS probe card technology on a mainstream CPU device that's forecasted to ramp in volume next year. And continuing on the theme of diversification and market share growth in foundry and logic, we've now met all technical requirements for a major GPU application with a new variant of the same Apollo MEMS probe card architecture and are now in the pilot production stage of qualification. Once we complete this final stage of qualification, we'll be in a position to compete for volume orders for GPU probe cards in the first half of 2026. Turning to our Systems segment. We delivered the expected sequential revenue increase in the third quarter and are forecasting additional growth in the current fourth quarter. Some of this strength stems from the typical seasonal cadence of the systems business, but we're also experiencing increased momentum in the progression towards initial production of co-packaged optics or CPO as well as the significant investments being made to advance quantum computing towards full-scale industrialization. In co-packaged optics, in addition to the multiple CM300xi systems running pilot production for our primary CPO customer at their foundry, we've now installed multiple units of our next-generation Triton silicon photonics test system. Developed in collaboration with Advantest and Tokyo Electron, Triton brings together fab level automation and integration of both optical and electrical probe and test capability for our customers. Triton represents the next step in our silicon photonics product road map as we transition our differentiated electrooptical probing technology from the lab to the fab, helping enable the adoption of energy-efficient optical data transmission in tomorrow's data centers. Before I turn the call over to Aric, I want to reiterate our continued commitment to achieving the 47% gross margin of our target model. As you can see from our fourth quarter guidance, we expect to reach target model revenue levels before we reach target model gross margin levels, but we've shown meaningful progress towards closing this gap. We plan to continue gross margin improvement in a direct factoring footprint by making continued progress this quarter and in 2026, layering on further improvement as we then bring our farmers branch expansion online at a structurally lower cost. These multipronged initiatives will improve our competitiveness and add capacity at lower cost enabling us to grow FormFactor as we meet the challenges of increased test intensity and higher test complexity associated with the adoption of advanced packaging in applications like high-bandwidth memory, co-packaged optics, and quantum computing Aric, you are up.

Thank you, Mike, and good afternoon. Before we dive into the details of our third quarter financial results, I want to express my excitement as I move into the role of CFO and begin to leverage my experience across operations and finance to drive operational efficiency and financial discipline. I truly enjoy working across varied constituents to find innovative ways of driving value and continuous improvement. I also want to spend a couple of moments to summarize how we plan to drive sustainable improvement in profitability, emphasizing the key focus areas that will guide our actions and our priorities in both the short and the longer term. As you heard from Mike, we are focused on improving our profitability to a path to the 47% non-GAAP gross margins of our target model. We are committed to achieving these improvements in a sustainable way and we believe the most critical elements of success for us in the short and midterm are to drive improved operational effectiveness, which means optimizing output from our existing infrastructure, and better financial discipline. We believe focus in these areas will drive meaningful change in our unit costs and our gross margins. This focus starts with some immediate action. First, on the labor front, we just completed a reduction in headcount, reducing costs even as we execute on existing demand and prepare for future demand. Furthermore, we implemented changes in how we manage over time in all of our manufacturing sites, immediately reducing our unit labor costs. Second, with respect to our manufacturing processes, we executed on targeted decreases in manufacturing spending. For example, we have expanded our existing precious metal recovery process to reduce waste in our manufacturing line among implementing other improvements. Beyond these immediate actions, over the coming quarters, we will continue to drive a relentless focus on reducing our manufacturing expenses by attacking the fundamental drivers of cost within our existing footprint. For example, focusing on improving yields and reducing manufacturing cycle times by smartly deploying automation, implementing more effective defect detection capabilities and implementing new factory management tools and analytics. Improvement in areas like cycle times and yields are structural and we believe improvements in these areas drive durable cost benefits that will help us to weather and partially offset the impact of inevitable shifts in product mix and headwinds presented by new challenges, such as what we have seen recently with tariffs. Even as we drive the unit cost of our products down, we are aiming to simultaneously enable higher output from our current infrastructure, reducing cycle times and improving yields enable this objective while also supporting our continued ability to be a top-performing supplier by increasing the quality and speed with which we address our customers' needs as they address rapidly growing demand in areas like high-performance compute and HBM. In addition to our laser-like focus on improving operational effectiveness, we will also exercise good financial discipline and continue to examine our overall portfolio of products, markets and businesses, evaluating all of our operations through the lens of how each best supports our target model and our key strategic priorities. We are changing how we communicate our financial results. Instead of residing breakdowns by product and market, we will continue to provide this important data in the supplemental materials on our investor website. And our discussion will focus on the key actions we are taking and the key drivers for our short and midterm road map to achieve our target model. As you saw in our press release, we are favorable to our Q3 outlook on revenues, gross margins and EPS for both GAAP and non-GAAP. Q3 '25 revenues are $202.7 million, non-GAAP gross margins are 41%, up 250 basis points from 38.5% in Q2 '25, and non-GAAP EPS is $0.33, and $0.04 above the high end of the outlook range of $0.21 to $0.29. GAAP gross margins for the third quarter were 39.8% compared to 37.3% in Q2. Cost of revenues included $2.5 million of GAAP to non-GAAP reconciling items, which we outlined in our press release issued today and in the reconciliation table available on the Investor Relations section of our website. As I mentioned, on a non-GAAP basis, gross margins for the third quarter were 41%, 250 basis points higher than the 38.5% non-GAAP gross margins in Q2 and at the higher end of our outlook range. This increase in non-GAAP gross margins is driven by improvement in both segments. The Probe Card segment was up 254 basis points and the Systems segment was up 260 basis points to 40.8% and 42%, respectively. The reductions we have made in labor costs and manufacturing spending are starting to have an effect on our financial results. This progress represents the start of a more disciplined path that will yield incremental improvement in gross margins, leading us back to our target model gross margins over the course of 2026, as you heard from Mike. Our GAAP operating expenses were $62.6 million for the third quarter, effectively flat as a percent of revenue from the prior quarter and down 130 basis points from the same period in the prior year, demonstrating continued discipline in spending across the P&L, even as we continue to invest in R&D in projects like Farmers Branch to drive innovation and future growth beyond the immediate term and even beyond our current target model. GAAP net income for the third quarter was $15.7 million or $0.20 per fully diluted share compared with a GAAP net income of $9.1 million or $0.12 per fully diluted share in the previous quarter. Third quarter non-GAAP net income was $25.7 million or $0.33 per fully diluted share up from $21.2 million or $0.27 per fully diluted share in Q2. The GAAP effective tax rate for the third quarter was 29.1% and the non-GAAP effective tax rate for the third quarter was 21.2%. We are continuing to refine our approach to key provisions of the recent tax legislation. Moving to the balance sheet and cash flows. We had free cash flow in the third quarter of $19.7 million compared to a negative $47.1 million in Q2. Remember that the reason for the negative cash flow in Q2 was the $55 million investment in the Farmers Branch, Texas manufacturing facility. Operating cash flows were $27 million in Q3, $8.1 million higher than the $18.9 million in Q2 '25, primarily driven by the improved net income on higher revenues and improved gross margins. At quarter end, total cash and investments were up $16.7 million to $266 million. Since purchasing the Farmers Branch facility, we have made excellent progress in executing our planning and pre-startup activities. This project is a good example of how we are taking advantage of our strong balance sheet to enable the next stage of growth and continued improvement in our gross margins over the long term. We expect the cash expenditures related to Farmers Branch will be between $140 million and $170 million over the course of 2026 and believe that this investment will enable further improvement in gross margins beyond our current target model. During the third quarter, we used $1.7 million to repurchase shares. At quarter end, $70.9 million remained available for future purchases under the $75 million 2-year buyback program that was approved and announced in April 2025. As a reminder, our share repurchase program objective is to offset dilution from stock-based compensation. Turning to the fourth quarter non-GAAP outlook. We expect Q4 revenues of $210 million, plus or minus $5 million. This increase in revenues more favorable product mix and the cost reduction initiatives described earlier are expected to result in a higher non-GAAP gross margin of 42%, plus or minus 150 basis points. As a reminder, we continue to see a 150 to 200 basis point impact on gross margins from tariffs. We are taking actions to mitigate the impact of these tariffs, but those efforts are ongoing. At the midpoint of these outlook ranges, we expect Q4 non-GAAP operating expenses to be $58 million, plus or minus $2 million. Approximately $3.5 million higher than Q2, mainly due -- sorry, Q3, mainly due to higher incentive-based compensation, planned spending on R&D and expenses related to the start-up of costs for our new manufacturing facility in Farmers Branch. Our Q4 effective tax rate is expected to be within the range of 17% to 21%. Non-GAAP earnings per fully diluted share for Q4 is expected to be $0.35, plus or minus $0.04. A reconciliation of our GAAP to non-GAAP Q4 outlook is available on the Investor Relations section of our website and in the press release issued today. As demonstrated by our Q3 results and our Q4 outlook, we are making encouraging progress to our target model. As recent initiatives to improve our structural costs take effect, and we are able to better leverage our fixed cost as demand increases. With that, let's open the call for questions. Operator?

[Operator Instructions]. Our first question comes from the line of Craig Ellis from B. Riley Securities.

Nice execution, guys, and Aric, welcome to the call. Mike, I want to just start with you. Thanks for all the insight on the top line and how the business segments are performing. I was hoping you could go back to some of your DRAM commentary and specifically commentary around HBM4 and the cutoff with the business having reached the cutoff in the recent quarter and flat sequentially. How would you frame up the growth gives and takes as you look out to 2026 for HBM. And if we look at the next transition without getting the cart too far before the horse as HBM4E becomes much more customizable. What will that mean for probe card intensity.

Yes. Thanks, Craig. So we did experience the crossover or are experiencing the crossover in the fourth quarter as HBM4 takes over as the majority of our HBM revenue. All three customers, as I said, are contributing to this. But we're fairly early in this HBM4 ramp. I think most people understand the timing and how it's linked up to a major high compute -- high-performance compute product launch next year. And so we expect continued growth from here in our HBM4 business and our HBM business overall. If you look at -- as we go through 2026, a couple of elements of test intensity and test complexity are increasing with HBM4, as I mentioned in the prepared remarks. Test speeds are going up, the layer counts are going up. These are all powerful tailwinds for probe card intensity. And as we move to HBM4E and then 5, those tailwinds continue, right? We expect higher speeds and higher bit counts. The other interesting wrinkle is, as you mentioned, this idea of a custom HBM base die. And I think that's -- although it's further out on the horizon, it's something we're partnered with customers very closely right now, as it combines high-end logic and memory controllers on the base die for HBM, obviously, FormFactor in a place where that can offer some significant differentiation as we're the only supplier of scale in both memory and logic probe cards.

Very helpful, Mike. Aric, the follow-up is for you. So thanks for the granularity on the levers that you're pulling to close to 500 gap -- 500 basis point gap to target GM. The question is this, as we look at things that, in my words, not yours, maybe more tactical versus those structural things that you talked about, how do those two things contribute in relative size to closing that 500 basis point gap -- and can you talk about the linearity that we should expect from where we are to 500, acknowledging that there's going to be mixed dynamics that move up and down along the way.

Yes. Thank you for your question. The 41% that we have as results in Q3 represents a meaningful improvement from last quarter. And we're already seeing the benefit of some of the actions that we have taken. We are not done, as you know. And so the restructuring actions that I referred to, we believe we'll continue to provide benefit heading into Q4 of about $1 million. And then on an ongoing basis of about $1.5 million thereafter. In addition to that, we have plans in place to continue to drive improvements in, as you noted, kind of fundamental cost structure areas, focusing on things like manufacturing cycle time and yields, and we believe that those will address both the gross margin road map over the course of 2026 as well as bring more output out of our existing facilities.

[Operator Instructions]. Next question comes from the line of Brian Chin from Stifel. Brian, you might be on mute. Brian Chin, we're not hearing you. Our next question comes from the line of Christian Schwab from Craig Hallum.

Just my first question regarding the gross margin target of 45%. It sounded like you think you will attain that level in 2026. Is that a statement regarding mix of business improving in foundry logic versus DRAM? Or is that a statement regarding the initiatives that you're doing or a combination of both.

As I mentioned in our prepared remarks, we are focused on changing the underlying cost structure across all of our products. And those underlying elements such as manufacturing cycle time and yield, those are independent of mix. Of course, there are always going to be elements of mix that impact us as well as volume. And we do expect to continue to see those impacts as we move forward. But the road map that we have in place we believe will bring us up to target model gross margins over the course of 2026, as you know, with mix independent.

Fantastic. And my second question has to lead to -- can you quantify the positive impact on 26 foundry logic from potential ramps of CPU and GPU customers, a broad range, Mike.

Yes. Christian, we haven't really quantified that. As I noted in the prepared remarks, we're making excellent progress on the qualifications and competing for business. This is a critically important initiative for us to continue to diversify our customer base and grow share in foundry and logic. We would expect a significant impact as we move through 2026. But as the selections and commercial negotiations are ongoing, it's hard to quantify. I will say the addressable markets associated with those two opportunities are significant. And if you look at one of our competitors who's been the primary vendor for it, you can see the impact of those. It's tens of millions of dollars a quarter from a served market perspective. Now we got to go compete and win on the back of these qualifications and bring that revenue and market share in.

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

Maybe first, and I apologize if I cover any ground that's been covered already between calls. But the -- for Q4, it sounds like the revenue growth, is that -- that's mainly being driven by so to speak, legacy DRAM, if I heard that right, even across the business encompassing logic/foundry. And I would even think of that being kind of a margin at best case, neutral, but probably a little bit negative, but you are guiding gross margins higher. Can you maybe kind of speak to how you're able to offset or improve that on the gross margin line? That's the first question.

Thank you for your question, Brian. I think there are some general relationships that we can draw from the market level DRAM versus foundry and logic. And in general, we do see some differentiation in standard margins across those markets. But we also need to remember that even within those markets, there's product level changes in profitability that can drive increases or decreases quarter-over-quarter as that mix changes. So there is an element of mix in there. But again, one of the main reasons why our gross margins and standard margins are improving quarter-over-quarter is really in great part, driven by the cost improvements that we're making, which is, again, mix independent, and we believe will sustain regardless of the mix of foundry and logic versus DRAM.

And I know you've announced an amount of CapEx increase for next year tied to the Farmers Branch facility and capacity expansion. Have you provided -- sorry, earlier in the queue or earlier in the call, details or some sense in terms of the timing of that deployment and sort of when some of that increased capacity will be available to the company.

Yes. So we have a detailed project plan that extends over the course of 2026 and 2027. We expect some of the initial capacity to come online late in 2026 with the majority of that capacity coming online into 2027?

Majority in 2027. Okay. Is it -- is it going to be focused mainly on like HBM market? Or have you not stated? Or is there kind of broader fungibility in terms of what you can produce in that facility initially?

That's a great question. While we are focused on continuing to drive incremental gross margin improvement, that's one of the things we have our eye on, we are also focused on making sure that as we invest additional capital into our manufacturing footprint that we're creating a manufacturing environment that's flexible and efficient in supporting our future growth and making sure that it can serve the breadth of our product lines and that we are able to move resources back and forth as the market dictates.

And our next question comes from the line of Charles Shi from Needham & Company.

Mike, Aric, by the way, Aric, welcome aboard. Looking forward to working with you. So the question on HBM, I don't recall you actually start an HBM revenue number for the third quarter in mind if you provide some color there, is that in target $11 million incremental DRAM revenue like all HBM? And what that means for traditional DRAM was still at the $20 million per quarter, that kind of a 12-ish level. But yes, I get it's going to increase into Q4.

Yes, Charles, for Q3, just to put a few more details around it. Most of the sequential growth in DRAM in the third quarter going -- second quarter to third quarter, was driven by HBM. So it was -- HBM in round numbers was $40 million in the quarter, in the third quarter, pretty close to the previous high. As you know, we now see in the fourth quarter some of the non HBM DRAM taking over, but in conversations with our customers, understanding that our lead times are still really well within a quarter, in most cases, we see some pretty strong growth associated with HBM4 as we move through the first part of 2026. And I think most suppliers who are participating in high-performance compute and the HBM4 and associated GPU and other networking chip ramp see the same strength.

Got it. Allow me -- I mean, forgive me for being a glass half empty this time, I have to do it. But looks like some commentary around your CPU customer, which disappeared from the 10% customer list this time feels like you are basically saying the revenue was kind of, I mean, getting to a pretty depressed level in third quarter. Mind if you put a little bit more quantitative color where your top CPU customer, where the revenue number was in Q3? And what's your projection into Q4 and given all the cost-cutting effort they're going through?

Yes. So you're correct in noting that our large CPU customer was not a 10% customer in the third quarter. But I'll take the glass half full position say that we still delivered revenue above $200 million, close to all-time highs. And if we talk about expectations for Q4, as I said in the prepared remarks, we're not seeing a lot of strength in the PC sector in the CPU sector, but still working very closely with that customer. They're a key partner for us. And as they go through some of their changes in restructuring and cost cutting, we're very closely partnered with them in making sure that we're a supplier that's continuing to help them through that. It's a long-term partnership. I'll also shift gears on you and talk about why it's so important that we qualify at both major CPE manufacturers. And as I said in the call, we're making good progress there generate revenue from those projects in 2026.

And our next question comes from the line of Elizabeth Sun from Citi.

I guess my first question is, Mike, you were talking about ASICs, contributed about a couple of million dollars last quarter. So I'm just curious, in Q2, so just curious in September quarter, did you see any contribution in ASICs project? And also just -- could you share us any updates on your engagement in the ASIC projects?

Yes. I think the custom ASICs space is an interesting growth opportunity for us. We're engaged with all the major hyperscalers, and we highlighted for you last quarter that we'd won a significant project that contributed to second quarter revenues. There's a little bit of contribution again in the third quarter, but I think these are long-term engagements with the hyperscalers. There is significant business there and to be perfectly transparent with you, we've got a smaller competitor who's doing a pretty good job serving that business with the two major ASIC -- custom ASIC projects in the industry. We believe that as things -- as the ASIC projects start to get closer to the specs required for GPUs, things like power, speed, density that those -- that market is going to consolidate towards the 2 top foundry and logic suppliers who have advanced MEMS probe technology. But for now, that's a hole in both our and our primary competitors revenue that we're both working to fix. I think the other point to make is when we look at the sort of the fundamental spend associated with high-performance compute in the logic space, GPU versus custom ASIC, it continues to be dominated by GPU. And that's why it's so important that we qualify for the merchant GPU business and start to participate in that.

That makes sense. And then on the Farmers Branch, I'm curious if you could share as it ramps majority in 2027, what would be the tailwind for the gross margin side?

Yes. So you're correct in that we will be ramping and investing over 2026 and 2027. We have detailed models and the detailed project plan associated with this project. And as we look at that and we look forward, we believe that our investment there will yield incremental gross margin improvement over the long term, as we move forward, and that's kind of beyond our current target model.

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

A couple of questions. Aric, I hate to do it, but I'm asking another gross margin question. When you look at the move from 38.5% last quarter to 42% this quarter. Could you segregate the impact of mix, overhead absorption and perhaps the cost reduction programs? Are they roughly the same? Or is one greater than the others?

Yes. Good question. And you're correct in the elements there. So mix, volume and cost improvement actions all contributed to the improved gross margins quarter-over-quarter. But as you could see from my prepared remarks, we are very focused on making sure that we are managing the underlying cost drivers that are going to persist period in and period out. If I were to characterize the relative contribution, I would say that the volume, for example, is the minority of the change from last quarter.

Okay. That's very helpful. And then Mike, obviously, some nice momentum again in the silicon photonics on the system side. What are the next couple of milestones that we're looking for in that space to see some progress going into the fab itself?

Yes. I think the key milestones, Tom, are that are going to be externally available or some planned product launches early next year, mid next year. And I think some of our customers have been pretty transparent about the insertion of CPO co-packaged optics into the road map. And I think if those are -- or when those are commercially successful, that will be the next catalyst for CPO. Right now, we're in pilot production, moving towards volume production -- as I noted, we've now installed multiple units of our Triton system, which is positioned for high-volume manufacturing of CPO, so we're ready to go when that begins to ramp. But probably the first externally visible catalyst is going to be some announcement coming in the early part of next year.

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

This is Steven Chin on behalf of Krish. Mike, I just had one first for you on the networking opportunity. I don't recall like from my previous conversations, but within foundry and logic is networking silicon and specifically data center type solutions. Are those a meaningful part of your foundry logic exposure today. And could you also talk about the long-term opportunities potentially with a major GPU customer versus merchant networking chip opportunities?

Yes. I think on the networking silicon side, this is currently, I'd call it an important part of the business, but some of the growth projections that we have are pretty significant. And that's one of the reasons why we highlighted it this quarter. I think you've probably heard pretty recently from some of the AT manufacturers that they see similar drivers in their business. And so as networking becomes a much more important part of the overall internal data center silicon content, that's an area where we're excited to take a strong incumbent position and continue to build revenue around that share position. I didn't understand the second part of the question. Can you repeat it for me.

Yes. Just in terms of your opportunities for future wins at, for example, a merchant networking companies versus bench opportunities at a major GPU vendor in the market just because the merchant networking companies, they -- as you mentioned earlier, they do use a smaller probe card vendor for some of the ASIC designs. Just kind of curious if that's also similar for some of these higher performing network to silicon as well?

Yes. I think it's an interesting question, right? As we've seen the ARC over the last 5 years of GPU requirements, for example, move into the space that requires an advanced MEMS probe card technology because of power, speed, pin count, pitches, there's really only two suppliers worldwide that can do that. And I think the evolution as GPUs are now firmly in this space. Obviously, we've got work to do as we qualify and start to win business there. But I think elements of these networking chips are also moving towards that space. And so in the whatever 15-odd years I've been in the probe card business, you've seen the steady progression of all kinds of different pieces of silicon migrate towards where requirements absolutely dictate the need for advanced probe card technologies. And for those of you familiar with our story, you know that, that's one of the areas of key investments and key differentiation that we have, both in foundry and logic and in DRAM.

Great. And for my follow-up, a question for Aric on the investments for Farmers Branch. You mentioned the is the target of $140 million to $170 million over the course of, I think, calendar '26. Just wondering, how does that break out between -- is that all CapEx? Or is there some component of R&D potentially in there as well?

Most of that is capital expenditures. It is a mix of different types of assets building improvements, clean room build-out and equipment. So various lives on those, but that spend is primarily focused on the build-out of the -- physical build-out of the site and the manufacturing equipment that goes along with it.

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.

Thanks, everyone, for joining us today. And let me add my welcome to Aric in his first earnings call. We're going to be doing a couple of conferences and events as we go through to the end of the year, and we hope to see you there and continue to update you on FormFactor's progress on the evolution to the 47% gross margin of the target model. Take care.