管理層發言
Thank you, and welcome, everyone, to FormFactor's Second Quarter 2026 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 those with respect to the projections of financial and business performance, future macroeconomic and geopolitical conditions, the benefits of acquisitions and investments, including the ramp-up of manufacturing facilities, 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. Those 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 27, 2025, 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, July 29, 2026, and we assume no obligation to update them.
With that, we will now turn the call over to FormFactor's CEO, Mike Slessor.
Thanks for joining us today. FormFactor's second-quarter revenue, gross profit, and earnings per share set all-time records. And we achieved two important milestones on the path to the new target model we introduced in May. First, we surpassed a $1 billion annual revenue run rate. And second, we exceeded a 50% gross margin. In the current third quarter, we expect to deliver sequential increases in both revenue and profitability as we extend our run of record results. Over the past four quarters, we have grown revenue by more than 30%, increased non-GAAP gross margin by 1,500 basis points, and have tripled our non-GAAP earnings per share. These improvements are the product of a multiyear effort to create and expand FormFactor's unique position at the intersection of high-performance compute and advanced packaging, while simultaneously strengthening the company's execution to enhance profitability and drive operating leverage. Our second-quarter results and third-quarter outlook show a clear path to our new target model. Looking further ahead, our Farmers Branch site is on track to come online in the fourth quarter of this year and ramp throughout 2027, providing increased capacity at a structurally lower cost. This will in turn enable us to accelerate revenue growth and generate additional profitability and operating leverage. Aric will discuss our current operational performance and future plans later in the call. The breadth of demand we saw during the quarter is particularly encouraging, with strength in major growth initiatives like high-bandwidth memory and co-packaged optics, powering sequential revenue increases in both the probe cards and systems segments. This broad participation reflects the diversification strategy we have discussed for years and demonstrates the value of serving all major semiconductors and semiconductor customers and applications rather than concentrating on any single segment, customer, or application. Currently, most of the manufacturing activity at this intersection of HPC and advanced packaging is occurring in Taiwan. FormFactor's important role in this region is evident in our supplemental materials. Second-quarter revenue from Taiwan grew sequentially by more than 30%, and the world's leading foundry was again a greater-than-10% customer. Yesterday, we announced the expansion of our multiyear partnership with Keystone Micro, a provider of semiconductor manufacturing and test services based in Zhubei, Taiwan. This latest step expands FormFactor's regional footprint in Taiwan, improving our local assembly and service capabilities and our responsiveness in supporting the exceptionally steep ramps of complex devices like GPUs and custom ASICs. Turning now to segment and market details. In DRAM probe cards, we delivered the expected sequential growth from the first quarter to set another record in this business, with increased HBM4 demand paired with sustained demand in DDR applications. HBM comprised approximately two-thirds of our overall DRAM revenue, driven by two customers' continued adoption of FormFactor's differentiated SmartMatrix full-wafer contactor technology in high-speed HBM4 applications. SmartMatrix provides a unique combination of high parallelism, productivity, and high-speed performance, enabling our customers to test hundreds of completed HBM stacks simultaneously at the 10-gigabit-plus I/O data rate of HBM4. This capability is critical in advanced packaging processes like TSMC's CoWoS and Intel's EMIB, where stacked-die test insertions ensure a known-good HBM stack before it is combined with costly GPUs or custom ASICs. Our second-quarter results show the impact of FormFactor's competitive advantage and the resulting market share gains as pin I/O speeds and overall stack bandwidth for HBM continue the relentless increase, as the industry progresses from HBM3 to HBM4 and then on to HBM5. In the current third quarter, we are forecasting DRAM revenues to be comparable to the record second quarter, but with a significant underlying shift from HBM to DDR. As you have heard recently from all our major DRAM customers, the supply environment continues to be extremely constrained, and our customers are shifting their wafer-start mix to DDR designs to capture the profit opportunity provided by significant DDR price increases. Since probe cards are specific to each customer chip design, we expect DRAM mix to track these dynamic customer product shifts while these unusual end-market conditions persist. Shifting now to the foundry and logic probe card market. As expected, second-quarter foundry and logic demand increased significantly over the first quarter. This increase was driven primarily by growth in probe cards for data-center CPU applications, building on continued strength in networking applications, initial momentum in hyperscaler custom ASICs, and steady demand in PC and mobile. In the current quarter, we expect continued growth in foundry and logic probe card revenue, driven by broad incremental demand across our served application space in this market. Compared to a quarter ago, the trend of increasing CPU compute intensity to enable agentic AI use cases is now broadly appreciated and is resulting in increased probe card demand for data-center CPU designs. FormFactor has opportunities to benefit from this trend in several ways. First, our long-term relationship and strong incumbent market share with a leader in data-center CPUs. Second, an expanding relationship with the world leader in high-performance compute, as our strong position in networking expands into supporting their GPU and CPU product lines. And third, our successful qualification and subsequent multiple design wins at a large fabless XPU customer. Each of these three customer relationships is in a different stage, but together they represent a broad-based opportunity for FormFactor to serve growing CPU demand. They also provide an excellent example of the value of FormFactor's long-held diversification strategy. Being a key supplier to all major customers for growing applications like CPUs provides us with broad exposure to the overall growth trend. Turning to our systems segment. Systems revenue nearly doubled sequentially in the second quarter, albeit off an unusually weak first quarter. This growth was driven by two components: one, a recovery in our core engineering prober business, and second, and more significantly, by accelerating growth in co-packaged optics, or CPO. A quarter ago, we forecasted our 2026 CPO revenues to reach the high end of the $10 million to $20 million range we communicated at the start of the year. We now expect to exceed that range by the end of the third quarter, and to significantly exceed the $20 million level for the year overall. This acceleration is driven by two factors: first, the growing volumes of CPO chips planned for later this year; and second, our leadership in the all-important test insertion one, which ensures known-good die on the photonic integrated circuit, or PIC, wafer before it is combined with the electrical integrated circuit, or EIC, to form the optical module in scale-up and scale-out network switches. The rapid year-to-date range growth of our CPO business is an exciting development, which we believe represents the very early stages of wide adoption of silicon photonics in the broader semiconductor industry. Traditional copper interconnect is reaching physical limits in speed, heat, and energy consumption, and photonics provides a fundamentally more efficient way to transact data within and between data centers by using light instead of electricity. FormFactor is ideally positioned to help lead tests for this new area of the semiconductor industry as our lab-to-fab strategy has produced a decade-long first-mover advantage paired with key customer and partner relationships. Before turning the call over to Aric, I want to thank the global FormFactor team as they continue to demonstrate remarkable agility in navigating the challenging supply environment by quickly resolving internal and external constraints. This agility helped deliver double-digit sequential growth in the second quarter, and we expect to grow again to another record in the current third quarter. We are well positioned as test intensity and complexity continue to rise at the intersection of high-performance compute and advanced packaging, and are excited to be making good initial progress on the path to our new target model that commits to doubling revenue and more than doubling profitability by 2030. Aric, you are up.
Thanks, Mike. Good afternoon. Q2 was another strong quarter for FormFactor. We delivered our third consecutive quarterly revenue record and drove additional non-GAAP gross margin expansion to 53.3%, demonstrating significant operating leverage and making progress on the priorities we discussed last quarter and at our Investor Day in May. At our Investor Day, we introduced our new target model with a goal of doubling revenues to $1.6 billion, achieving 55% on a non-GAAP gross margin, and more than doubling non-GAAP earnings per share to $5 per share by 2030. The current strong demand environment combined with our continued focus on operational execution drove measurable gains toward the new target model in Q2. Non-GAAP gross margin increased 34 basis points sequentially and is up nearly 15 percentage points as compared with Q2 of last year. Similarly, non-GAAP EPS increased by nearly 50% sequentially to $0.82 per share and more than tripled as compared with Q2 of last year. The quarter-over-quarter improvement in non-GAAP gross margins is driven by several factors. Approximately one-third of the improvement is from baseline cost reductions that are durable in nature. One-third is driven by the $32 million, or approximately 14% quarter-over-quarter increase in revenues. And the remaining one-third represents items that we do not expect to recur, such as IEPA tariff refunds and precious metal reclaim from our Baldwin Park site shutdown announced early in Q2. Product mix remained strong in Q2, driven by factors like record HBM revenue within DRAM. Excluding the timing items and mix favorability, we believe baseline non-GAAP gross margins have improved to around 51% at these Q2 volumes. The combination of higher volumes and a more efficient cost structure is enabling us to convert strong demand into higher gross profit and operating income. We have taken several measures to increase output in the short term, even as we prepare for the ramp of Farmers Branch starting at the end of 2026. Gains in output have been primarily achieved through yield and cycle-time improvement, even as we see more constraints across the supply chain at current production levels. Our Farmers Branch site expansion remains on track to ramp starting at the end of 2026 and continue to ramp over the course of 2027 and 2028. Bringing this capacity up on time and on budget remains a key focus, as it will enable our next phase of growth and gross margin expansion. Stepping through our results in a bit more detail: Q2 2026 revenues of $258.2 million came in $18.2 million above the midpoint of our Q2 outlook range of $235 million to $245 million, and we are up $32.1 million, or about 14%, from Q1. Systems segment revenues made a significant recovery, reaching a new record of $48.5 million in Q2 2026, up $20.6 million, or 74%, from Q1. GAAP gross margin for the second quarter was 50.7%, up from 38.4% in Q1. Cost of revenues included $6.7 million of GAAP-to-non-GAAP reconciling items, primarily related to stock-based compensation, amortization of intangibles, and restructuring charges. Details of the GAAP-to-non-GAAP reconciling items are outlined in our press release issued today and in the reconciliation table available on the Investor Relations section of our website. Q1 GAAP gross margins had included $18.8 million of restructuring costs that did not recur in Q2. On a non-GAAP basis, gross margin for the second quarter was 53.3%, 34 basis points higher than Q1 and 32 basis points above the high end of our Q2 outlook range. Excluding timing items and mix favorability, non-GAAP gross margin was 54.4%, driven primarily by higher factory utilization, manufacturing spending discipline, and improved yields. System gross margin increased to 48.5%, driven primarily by higher volumes and favorable mix. Our GAAP operating expenses were $73.1 million for the second quarter, up from Q1 but down as a percentage of revenue. On a non-GAAP basis, operating expenses were $65.7 million, or 25.4% of revenue, compared to $62 million, or 27.4% of revenue in Q1. This 200-basis-point sequential improvement in OpEx as a percent of revenue demonstrates operating leverage across the P&L and is the result of continued spending discipline even as we grow, invest in key R&D programs, and fund the Farmers Branch expansion. In Q2, operating expenses included $4.9 million of preproduction ramp costs for Farmers Branch. GAAP net income for the second quarter was $56.2 million, or $0.71 per fully diluted share, up from GAAP net income of $20.4 million, or $0.26 per fully diluted share in the previous quarter. The increase was driven primarily by higher revenue, higher gross margin, and lower restructuring-related costs. Second-quarter non-GAAP net income was $65 million, or $0.82 per fully diluted share, up from $44.5 million, or $0.56 per fully diluted share in Q1. The GAAP effective tax rate for the second quarter was 11.1%, and the non-GAAP effective tax rate for the second quarter was 16.2%. Moving to the balance sheet and cash flows: we delivered free cash flow in the second quarter of $52.6 million, compared to $30.7 million in Q1. This increase in free cash flow was driven primarily by higher cash flows from operations. Cash flows from operations were $61.8 million in Q2, up $16.8 million from Q1, driven primarily by higher net income, partially offset by working capital investments to support these higher operating levels. At quarter end, total cash and investments were up $40.8 million to $349 million. We continue to expect that the cash CapEx for Farmers Branch and capacity additions will be between $140 million and $170 million in 2026. Preproduction ramp costs are recorded as a component of G&A and are expected to be between $25 million and $30 million in total in 2026, with about $12 million incurred to date for Q2 and about $7 million expected in the current third quarter. Once production ramp begins, the cost currently recorded in G&A will become part of cost of goods sold on a go-forward basis. Upon completion of the ramp to the initial target capacity, which is expected by the beginning of 2028, we expect Farmers Branch to be accretive to gross margin. Associated with our investment in Farmers Branch, we secured certain incentives from both the state of Texas and the City of Farmers Branch that we expect will partially offset the associated expenditures. Among others, these incentives include a $24.2 million grant from the Texas Semiconductor Innovation Fund designated to fund capital expenditures upon meeting certain criteria. Overall, we continue to expect that we are largely self-funding the investment in Farmers Branch from increased profitability, a more efficient cost structure, and cash flow from operations. During the second quarter, we did not repurchase any shares. At the quarter end, authorization of $70.9 million remains available for future repurchases under the April 2025, $75 million two-year buyback program that is intended to offset dilution from stock-based compensation. In the short term, we are continuing to prioritize our deployment of cash to accelerate the ramp of our new manufacturing site in Farmers Branch. Turning to the third-quarter non-GAAP outlook: we expect Q3 revenues of $270 million, plus or minus $10 million. At the midpoint of this revenue range, we expect non-GAAP gross margin of 54%, plus or minus 150 basis points. The increase from the Q2 baseline of 51% described earlier to the Q3 outlook is driven primarily by items that are not expected to recur, specifically $7 million to $9 million, or about 300 basis points, in IEPA tariff refunds that we expect to receive in the current third quarter. These refunds represent the return of tariffs paid starting in 2025 and through the beginning of 2026, prior to these tariffs being ruled unlawful. The benefits of marginally higher Q3 revenue volumes are expected to be offset by less favorable product mix in DRAM, as Mike mentioned earlier. At the midpoint of the outlook range, we expect Q3 non-GAAP operating expenses to be $70 million, plus or minus $2 million, including about $7 million in preproduction ramp costs for Farmers Branch. Our Q3 non-GAAP effective tax rate is expected to be between 15% and 19%. Non-GAAP earnings per fully diluted share for Q3 is expected to be $0.86, plus or minus $0.09. A reconciliation of our GAAP to non-GAAP Q3 outlook is available on the Investor Relations section of our website and in our press release issued today. As demonstrated by our Q2 results and our Q3 outlook, we are already making meaningful progress toward our new target model we shared in May. As we capitalize on the strong secular trends at the intersection of high-performance compute and advanced packaging, our differentiated products are driving demand in areas like high-speed test of HBM memory and traction in new addressable markets like co-packaged optics. These growth vectors, combined with our focus on execution, cost discipline, and the timely ramp of additional capacity in Farmers Branch, have resulted in good operating leverage in our Q2 results and chart an encouraging path into the future. With that, let's open the call for questions. Operator?
分析師問答
As a reminder, to ask a question, you will need to press star 1 on your telephone. To remove yourself from the queue, you may press star 1 again. Please limit yourself to one question and one follow-up to allow everyone the opportunity to participate. Please standby while we compile the Q&A roster. Our first question comes from the line of Krish Sankar of TD Cowen. Your line is open, Krish.
Yeah. Hi. Thanks for taking my question, and congrats on a very strong result and impressive gross margins. I had a clarification question, Aric. Did you say that 51% or 53% gross margin is the new baseline? And as Farmers Branch comes online, it is going to be more accretive, so should we assume as these revenue levels go higher, the gross margins will be better than 53%?
Hi. Yes. Thank you for your question. Just to clarify what I said, the actual gross margins for the quarter were 53.3%. That included some items that we do not think are recurring, such as tariff refunds and precious metal reclaim. Those items will not recur. So the recurring or sustainable element of gross margins, we believe, is more like 51% at the current volumes and mix. So we expect that is the new baseline that you should be thinking about. Does that answer your question?
Yes. And that incorporates even Farmers Branch when it comes online?
So that is the current baseline with our current operating footprint. We do expect that Farmers Branch, as it comes online, will be accretive to current gross margin.
Awesome. And then a quick follow-up for Mike. Mike, it's kind of impressive to see TSMC being a greater-than-10% customer. I am just wondering, is some of the GPU business under that greater-than-10% customer? Is that mostly networking chips? Would Rubin and CPO for NVIDIA come under TSMC, or would it be under NVIDIA? And what is the status of the Rubin qualification?
Let me address that first, Krish. There really is no component of GPU business in the second-quarter results for this greater-than-10% customer. As we said in the past, we expect this to be a second-half event for us, and we are on track. We are qualified, as we have shared with you, and we are shipping production units for revenue here in the third quarter. The CPO piece does flow through that greater-than-10% customer, so in many cases, with the combination of the fabless-foundry ecosystem you recognize revenue in different places depending on the market segment and even depending on the individual product line. Awesome.
Thanks a lot, Mike. Appreciate it. Congrats, Aric.
Thank you.
Our next question comes from the line of Matthew Prisco of Cantor. Please go ahead, Matthew.
Hey, guys. Thanks for taking my question. First on foundry and logic, can you provide more color on the breadth of the strength you are seeing there today? And you highlighted agentic AI driving CPU demand — how are you thinking about that as a potential upside source versus your target model that was outlined a couple months ago?
The strength as we went from Q1 to Q2 in foundry and logic was primarily associated with an uptick in CPU demand from one of our historically strong customers. But as we look at our opportunity in the CPU space, which we do view as a longer-term opportunity, there are a couple of different ways that we are exposed to that. As I mentioned, there's a growing relationship with the leader in high-performance compute; they are starting to participate in the CPU business. And remember, we are qualified and now competing for share with the large fabless CPU manufacturer. So this is a great example of where we are diversifying: no matter how the CPU market-share landscape shakes out between our customers, we feel like we at least have the opportunity to participate. All of those relationships are in different phases, so if it all hits today, it will have different components for us, but it's a good example of our goal to be a broad-based supplier to all the leaders in the industry.
Got it. That is helpful. On the supply side, efficiency is coming in much better than expected. How do we think about your ability to continue driving efficiencies from here within the current footprint through the end of the year and into 2027? And more importantly, Farmers Branch coming online in Q4 and ramping through next year — how do we think about the timing of the revenue contribution there, and how meaningful can that be over the next few quarters?
We are very excited about the output we have been able to drive out of our existing footprint. That was driven in large part by the transformation we've done in our global operations team, really thinking and doing differently, and we see those benefits being driven primarily from yield and cycle-time improvements. As you can tell by our most recent results and our outlook for next quarter, we have been pretty successful, and we intend to continue to push on those levers through the end of the year in advance of Farmers Branch starting to ramp in Q4. The initial targeted capacity of Farmers Branch is roughly equivalent to our California probe-card footprint today, so we expect that it will come online over the course of the year and should contribute meaningful capacity to the extent the demand is there.
Our next question comes from the line of Craig Ellis of B. Riley Securities. Please go ahead, Craig.
Yes. Thanks for taking the questions, and let me start by recognizing the stellar execution in the quarter. Nice job, team. Mike, I'll start with DRAM: a few quarters ago we wondered if legacy formats like DDR4 and DDR5 would come back, and here they are. Can you help us understand how long-lived you see the strength you're talking about for Q3, and how we should properly think about the DRAM business from here given it seems to have near-term legs?
The DRAM market, both HBM and DDR5, is a pretty dynamic marketplace for our customers and for us. As we said, we see a pretty significant mix shift in Q3: although top-line DRAM revenue is expected to be similar to Q2, we expect the mix to shift toward DDR quite significantly. Conversations with our customers indicate this is a rational reaction to the pricing they are seeing in the market, where they can generate more profit by starting DDR5 wafers than by starting HBM4 wafers. So you'll see a very adaptive and dynamic environment where they will adjust their product mix to maximize profit. Because probe cards are device-specific consumables, as they change wafer-start mix, our mix will change correspondingly.
And then a follow-up on the breadth of demand: your greater-than-10% customer list is disclosed, but beyond those there are many big customers totaling 35% of revenue, meaning 65% comes from others. As we think about opportunities from here — you alluded to competing for business at a fabless CPU supplier and other second-half GPU opportunities at a leading GPU supplier — how many opportunities like that are there? Can you give us a Pareto list and help with timing for when those could come into the model?
There are a bunch of different opportunities, and I will ground people back to what we said at our Investor Day in early May. The segments and customers where we plan to gain more than half of the increase in the addressable market between now and 2030 are familiar areas: GPUs, co-packaged optics, custom ASICs. In expanding our HBM business, for example, we're seeing strong adoption now from two major customers. There is broadening, diversification, and share gains. You're seeing the early innings of what we described at Investor Day in our Q2 results and Q3 guidance. Those are the main drivers we expect to push growth over time.
Thank you. Next question comes from the line of Elizabeth Sun of Citi. Please go ahead, Elizabeth.
Hi. Thanks for taking my question, and congrats on the good result. First on the CPO side: Mike, you talked about CPO — total CPO revenue expected to cross over $20 million by Q3 and another step up in Q4. Is CPO revenue increasing quarter over quarter throughout the year? And ultimately how much revenue can you do for this year in CPO?
It's an interesting question. To ground everybody: we came into the year saying we expected between $10 million and $20 million in CPO revenue for 2026, and today we expect to reach $20 million by the end of Q3. How much more we do in Q4 and for the full year is an open question right now. We are seeing significant acceleration, but this is new technology for our customers and for us, so forecasting precise incremental amounts is tricky. We certainly expect to be above $20 million for the year; the magnitude beyond that is harder to judge. That said, this is a significant market: at Investor Day we said our served market is about $400 million by 2030, and we are accelerating toward that.
Got it. Second question on HBM competition: you are strong with two HBM customers. How about share dynamics at the third-biggest HBM customer? And one of your foundry-logic probe-card peers has talked about entering the HBM market — how do you see the competitive environment going forward into HBM4e and beyond?
I will be high level here. It's important to understand subsegments within HBM. Where we are doing very well and have strong share at two major HBM manufacturers is in the high-speed stack-die test — essentially the final test for the HBM stack where our customers ensure it is good before it gets shipped to the foundry for packaging with GPUs and custom ASICs. That is a very high-performance insertion and one where we have very strong share. There is competition in other parts of the flow, like core-die insertions and wafer sort-type tests. Our differentiation and share remain strong at the high-value, high-speed final-test insertions, and we continue to partner with all three customers, although our share is stronger at two of them right now. We're continuing to advance capability and differentiation as speeds and stack heights increase from HBM4 to HBM4e and HBM5.
Thank you. Our next question comes from the line of David Duley of Steelhead Securities.
Thanks for taking my question. Congratulations on getting a lot more out of your current factory footprint. At your Analyst Day you showed a chart with cycle-time improvements and a goal of reaching significant cycle-time improvement. Can you share any metrics now after several quarters of improving output? How much have you lowered cycle times or improved yields? Any metrics about where you are on that journey would be helpful.
Thanks. We do not typically share specific cycle-time or yield metrics in these forums, but I can say the drivers that will chart our path forward for gross margins remain consistent with what we discussed at our Investor Day and the target model we laid out. We still expect to drive the majority of improvements from volume, operational excellence, and transformation innovation. We are on track with that. If you look at the implied revenue run rate for our Q3 outlook, which is starting to get close to a $1.1 billion run rate when annualized, and our gross margin trajectory from Investor Day — from 49% to a normalized level around 51% at current revenue and mix — we are aligned and on track. We are very happy with the performance to date.
So would you say you are halfway on the journey to your cycle-time improvement goal, or three-quarters of the way? I don't have the chart in front of me, but at Analyst Day you implied meaningful progress.
Again, I would say we are on track. We see appropriate contribution from all the vectors we discussed. We're continuing to push yield and cycle-time improvements and expect to sustain those gains as we ramp Farmers Branch.
A second question: there's been chatter about a third hyperscale customer ramping into volume. Can you remind us where you are in servicing the hyperscaler customers, and how big you think the TAM for hyperscaler probe cards is?
Hyperscaler custom ASICs are one of our key growth initiatives. We work with all the hyperscalers; some are further along with multiple released parts, and some are earlier. We have revenue in the second quarter and have shared design wins. The big discontinuity is still to come when these custom ASICs require advanced MEMS probes. It looks like the next generation of parts, such as some TPU-like devices, will require advanced MEMS probe technology because of the power and speed requirements. The same transition happened with GPUs a few years ago, when historically they used legacy probe cards and then crossed a performance threshold requiring advanced MEMS. We are engaged with all hyperscalers on next-generation development. I don't expect significant revenue contributions in 2026, but certainly into 2027 as those parts ramp, we expect to be a key supplier.
One follow-up about broadening out in foundry and logic: besides CPU, are there other areas like GPU, industrial, or automotive that could be needle movers for foundry and logic?
Our focus revolves around high-performance compute and advanced packaging. We're levered to leading-edge devices where probe cards and wafer tests are most valuable: when yields are low and packaging costs are high. HPC and advanced packaging live at that nexus. Automotive and industrial have some exposure but are spotty and not as significant growth drivers relative to HPC and AI. From a served-available-market perspective, GPUs, custom ASICs, and CPO are the large drivers. Those are the primary growth vectors we expect to continue to drive our foundry and logic strength.
Our next question comes from the line of Dennis Pimentel of Needham & Company. Your line is open, Dennis.
Great. Thank you. On the CPO adoption outlook: initially you said this would be a 2020s story, but now you're seeing more acceleration. Are we expecting the timeline to accelerate, perhaps seeing more volume hit in 2027?
I wouldn't say there is a fundamental change. We are seeing acceleration in the near term. We talked about $20 million in CPO revenue by the end of Q3 and expect to exceed $20 million for the year overall. The difficulty is forecasting quarter-by-quarter through 2027 due to variables like overall CPO adoption, yields, and test time. We are excited about the opportunity: we have a strong position in insertion one and are partnered with leaders across the ecosystem. We expect significant growth, but making very granular quarter-by-quarter predictions into 2027 is premature.
Thanks. For a follow-up, GPUs adopted advanced MEMS probe cards a few generations ago. Is there any change or acceleration now for GPUs or other products adopting MEMS-based probe cards? Any particular products that may adopt it sooner rather than later?
GPUs have fully adopted advanced MEMS probe cards; that transition occurred a couple of generations ago driven by power and speed testing needs. Today, only a handful of companies can produce advanced MEMS probe technology. It has largely been our competitor's business, but as we've updated everyone, we are now qualified and shipping production units here in the third quarter, and we expect that business to grow. As for custom ASICs, that transition is happening now: those parts are moving toward advanced MEMS probe technology, and we expect to participate as those devices ramp.
Our next question comes from the line of Christian Schwab of Craig-Hallum. Your line is open, Christian.
Hey. Great quarter. Can you give us some idea on the fabless CPU customer where your market share is today and where you think it can go over time?
We are qualified, and we've won a couple of designs. This is a brand-new customer for us, so introducing ourselves and their supply base takes some time. Also, industry-wide capacity constraints mean we want to be thoughtful about commitments to existing and new customers. Until Farmers Branch comes online, that will somewhat govern market-share growth. I would characterize our position as low single-digit market share today, with significant opportunity ahead.
Thanks. One last question on gross margins: at Analyst Day you expected Farmers Branch to drive gross margin accretion by 2028 and beyond. Should we assume some gross-margin improvement in 2027 from lower-cost manufacturing base, or will the benefits mainly come in 2028 once ramped?
We expect the full accretion from Farmers Branch primarily in 2028 because ramps come with some inefficiency. We expect to see some of that through next year as we ramp the site. That said, we plan to offset those inefficiencies through operational improvements we've made and continue to make. Completing the ramp as quickly as possible is important, and we plan to largely offset ramp inefficiencies with operational effectiveness gains. We need this capacity; it is coming online in a timely fashion and is important to support the growth we see in our current addressable market.
No other questions. Thank you.
I would now like to turn the conference back to Mike Slessor for closing remarks. Sir?
Thanks again, everyone, for joining us today. As usual, we will be attending some late-summer and early-fall conferences and hope to see you there and answer your questions about FormFactor. Until then, stay safe and take care.
This concludes today's conference call. Thank you for participating. You may now disconnect.