All right, everybody. Why don't we get seated and get ready. We're at the top of the hour. We're gonna kick it off with a short video. Enjoy the show.
Well.
Oops. Okay. Now enjoy the show.
Okay. Well, semiconductors are an interesting class of materials and devices. They're things that don't really conduct electricity, but don't really block electricity either. That unique property allows them to be built into switches. These switches are called transistors, and they're essentially the fundamental building block of laptop computers, of mobile phones, even the heart of artificial intelligence that everybody's talking about today. These miniature building blocks are really at the heart of what we at FormFactor and the broader semiconductor industry do every day. Moore's Law was an observation and prediction coined by Gordon Moore, one of the co-founders of Intel, in the 1960s. What he observed was that the number of transistors on a chip was doubling every two years, and predicted that it would continue on for years in the future. It's been the heartbeat of the semiconductor industry for five-plus decades.
It turns out that it's getting very, very difficult to build smaller and smaller transistors, and as a consequence, the industry is running into challenges continuing the pace of innovation we've had for the previous 50 years. Instead of making smaller transistors, 'cause we're an innovative industry, we're resorting to other techniques. Advanced packaging is one of the ways the industry is counteracting the slowing of Moore's Law, and the basic idea is you break what would've been in the past a single chip into smaller chiplets. That allows you to then pick and choose which kind of production processes you're gonna use for each chiplet and fine-tune them so that you can use the process that's most advantageous to each block of the overall chip. FormFactor's at the center of this because each of those chiplets needs to be tested before it gets reassembled into a chiplet.
We build a variety of products and services that allow customers to check whether the chip is gonna work as it leave the factory and verify that it's gonna work over the life of the chip. For example, in a mobile phone or a PC or, even more importantly, a car or a pacemaker.
FormFactor is a unique company in the sense we have a very comprehensive set of product and portfolio solutions that addresses pretty much every device that can be tested in the market, particularly in advanced test needs that is needed by our customers. For example, whether it is a GPU or a CPU or high-bandwidth memory, we call our portfolio as a lab-to-fab solution, essentially because we have a very comprehensive set of differentiated solutions. That kind of sets us apart.
With our technologies, we are able to provide automation, which actually reduce manual interaction of operators and improve time to market dramatically.
Production cost is part of our daily job. What that means is controlling the cost and trying to keep as cost-effective as possible. One of the things that we do is we deploy advanced, efficient manufacturing processes. We produce a lot of products here, right in Livermore, California, and we produce millions of probes, so we get scale economies. We work strategically with our supplier and our supply chain to get cost efficiencies. Also, we ensure when we design, we design with cost in mind so it can be produced at a cost-effective manner.
Diversification helps address many different customer needs and technological challenges that we have in the industry, but it also helps us support various applications that we have in the industry today. Take it for example, automotive, high-performance computing, high-bandwidth memory applications, advanced packaging, quantum computing, and now even artificial intelligence. Diversification, it makes us resilient to market fluctuations and also technological changes that's needed in the industry.
AI is driving a lot of demand for computing power and also high-bandwidth memory. To optimize that, silicon photonics is the new approach to reduce power as well as increasing bandwidths. Developing silicon photonics chips is a different approach than normal electrical test. You have to connect optically to the chip, which is a unique way of testing it. FormFactor has developed a solution for the last eight years with customers together and introduced a technology which reduces test time dramatically. This enables our customers to go to market faster with more reliable chips.
FormFactor being a key enabler in the semiconductor ecosystem, we actually enable AI through testing of high-bandwidth memory and AI-specific logic devices working with our customers. At the same time, we're also consciously using AI-driven solutions to really help accelerate and collaborate on the product development side, as well as improving our manufacturing efficiencies. That's kind of where our predominant focus is now and for the foreseeable future.
As one of the key suppliers and key enablers of test and measurement technologies to the industry, we feel like there's strong growth in front of us. That's why we're doing things like investing in our team, investing in facilities worldwide, and making sure that as, most of all, we're investing in R&D and new innovative products to help the industry and our customers move ahead.
The FormFactor employees are the key to our success. Without their knowledge, expertise, dedication, and hard work, we would not be in the leadership position where we are right now. Working very close with customers, developing tailored solution together to really address their need is really engaging and motivating, not only for me, but for the whole company.
All of the things that we are doing, we want to make sure that while we are focusing on technological innovations, while we are growing significantly in so many different areas with diverse products, we want to grow and motivate people to the values that we are standing for.
I really like being part of an industry that continues to drive advancement and innovation that touches almost every human life on the planet.
Thank you everyone. Welcome to FormFactor's Investor Day here at the Nasdaq MarketS ite at Times Square. My name is Stan Finkelstein. We have a great agenda for you today. We have a great lineup of speakers. We're gonna kick it off with our President and Chief Executive Officer, Mike Slessor. Mike will outline his vision for the business for the next three to four years. After Mike, we're gonna have Sudhakar and Jens cover their respective businesses. Sudhakar runs our probes business. Jens runs our systems business. Both of them will cover strategic priorities. They will focus on long-term growth opportunities for their businesses, as well as some of the operational initiatives, more near-term focused operational initiatives. We'll take about five minutes break. After the break, we will have our Chief Commercial Officer, Aasutosh, kick it off with customer perspective.
It's all about the customer. Aasutosh will hand it off to Missy, the head of our global operations, and Missy will cover both some of the important operational initiatives that were implemented over the last year or so that resulted in very significant improvement in our financial performance. As well, she will provide an update to Farmers Branch expansion, our capacity additions, and some of the very innovative ways that her team is approaching this capacity expansion. After Missy, we will hand it off to our Chief Financial Officer, Aric McKinnis. He will provide view of our financials, talk about capital allocation, and close us off with our target model. Before I hand it off to Mike, I want to bring your attention to forward-looking statements and GAAP to non-GAAP financial measures.
As I have this slide on the background, I want to run through a couple of housekeeping items. First, the presentation itself will probably take about 1 hour 45 minutes-1 hour 50 minutes. We will take a break in between, and then we will have an hour for questions and answers. Unfortunately, we will only be able to take Q&A or take questions from people in this room. For those of you online, I apologize, but obviously you're welcome to stick around and listen to Q&A part of the presentation. We will have Crystal and Kelly, who will join us later, handle all of the mics. If you guys have a question, raise your hand, you'll get a mic. Please keep your questions to two to three so we can have broader participation.
The last point, this presentation will be available on the Investor Relations section of our website. It will be posted right after we're done here today. Again, welcome and thank you for joining us. I'll hand it off to Mike now.
Thanks, Stan. Thanks to all of you for joining us for a few hours today to learn some more about FormFactor and the bright future we have ahead. I'm Mike Slessor, FormFactor's CEO. I've held that role since 2014 after joining FormFactor as part of FormFactor's acquisition of MicroProbe back in 2012. I'm gonna start with the punchline today. By 2030, we plan to double FormFactor's revenue, and through operating leverage, more than double our non-GAAP earnings per share. We're gonna do this by executing at a unique position that we've created, the intersection of high-performance computing and advanced packaging, two of the most powerful trends ever in the semiconductor industry. From that position, that unique position at the center of high-performance computing and advanced packaging, we're redefining what's possible.
Redefining what's possible in terms of technology, in terms of products, in terms of scale, in terms of operational execution, and of course, financial performance. Our plan's built on three foundational themes. First, starts with customers, focusing on building sustainable long-term relationships as a leading supplier to the top customers in the industry. I'll get back to that in a minute and explain why it's so important in our business. Second, continuing to build and grow our diversified position as a leader in wafer test, creating scale and a diversified set of revenue streams, both organically and through M&A. Finally, realizing the benefit of that position of scale at the center of these two trends with a really renewed capability and focus on the ability to operate and execute at a world-class way.
Today, our team's gonna share how we get there and why Form's a compelling investment opportunity. I want to start with the idea that we're a key player in enabling high-performance computing. Let me do that by mapping our major products onto a high-performance computing module or chip. You see this thing at the center. I think everybody's familiar with the architectures by now. High-end compute logic devices like GPUs and custom ASICs at the center, surrounded by high-bandwidth memory, a variety of networking and interface chips, some bridges and interconnect that tie all this together, and in the very near future, nearly the present, co-packaged optics, which fuses photons together with electrons and allows the industry to take the next step forward. Let's take our products. Starting on the left here, probe cards.
We're the world number one leader in probe cards and have been the number one leader for well over a decade. We operate in both DRAM and foundry and logic probe cards. Up at the top are our DRAM architecture, which has been the real growth driver for our leadership in high-bandwidth memory. Down the bottom are Apollo foundry and logic probe cards. These got their start in applications like PC and mobile. You've now seen recently us extend that application space to networking and moving on to GPUs and custom ASICs. We're the only supplier of scale in both foundry and logic and DRAM probe cards, and that gives us the ability to create synergies, technology synergies in the basic R&D we use for things like the advanced MEMS probes that Sudhakar is gonna share with you.
It gives us operational synergies in giving us more scale and more leverage operationally. As you'll see in the future, as certain applications like custom HBM start to dominate the landscape, this fusion of DRAM and foundry and logic or logic IP on a chip, we're uniquely positioned to capture that opportunity. Probe cards are an interesting device, an interesting product in that they're a device-specific consumable. Probe cards are specific to each customer chip design. When a customer changes a mask set, no matter how subtle, changes the die size, changes the specs, that means new probe cards. It drives a fundamentally different demand characteristic in our business than capital equipment, even capital equipment that operates very close to us like ATE.
This is why we call it a design-specific consumable, and it's one of the very unique flavors of the probe card business, and as I'll get to, some of our competitive advantage. Shifting over to our systems segment, Jens is gonna share with you how we're advancing the systems segment with respect to serving co-packaged optics. This is an area that's a perfect example of our lab-to-fab strategy and the benefits of that strategy. We've been working with leading customers for almost a decade on silicon photonics, co-packaged optics, and various photonic-based applications. We've taken that learning, that co-development, that refinement of the technology, again working with key customers, and now driven it into a production-worthy platform, collaborating with Advantest and Tokyo Electron. As I noted on the last earnings call a couple of weeks ago, we're now beginning to see the acceleration of production adoption here.
The point is, you can't get on this train right now. We had to be on this train a decade ago, working with leading customers on some real fundamental challenges associated with integrating optics into the overall production testing scenario. You can see across all of the major applications driving high-performance computing, each of these pieces of silicon that are essentially the engine room of generative AI, FormFactor has a strong and relevant position. Backing up for a minute and talking about how we got here and where we're headed, I talked about customers, and you see some names across the top. These are customers that have recently been in our SEC reports as at least 10% of revenue. This is a set of names that we're proud are leaders in the industry.
We work with these customers every day, not just in serving their production requirements, but in serving their next-generation R&D requirements, this lab-to-fab strategy. It's critically important. We have, essentially, the largest R&D budget in our served markets. These customers every day help us prioritize what we're working on, how we're developing it, because the industry is moving so fast, they don't wanna leave anything to chance, right? They're dependent on our technology roadmap, execution, and then operational execution to get to their next challenge. That's why these durable, long-term customer partnerships are so important for our success. I talked about expanding, growing, and diversifying the business. We've done that both organically, you'll see some examples in a minute, but also through M&A. Here's a list of some of the major M&A events in company history.
These have both been strategic acquisitions, but we also hold ourselves to the bar of them being financially accretive transactions. The most recent one we did at the end of 2025 was the acquisition of a small company called Keystone Photonics, who has an innovative optical probing technology that we think is a key enabler for driving the CPO roadmap forward. A couple of milestones or a few milestones here. Again, this notion that engaging early with these key customers in long-term partnerships is central to success. We shipped our first HBM probe card to who's now our number one customer in 2012, the year I joined FormFactor.
We've worked with them and other DRAM customers to optimize this technology where it's now capable of astounding things in serving HBM4 and HBM4E, and in the future on to HBM5. Sudhakar will touch on how this architectural development has occurred and why FormFactor has a competitive advantage that will continue to persist in this important market. Same story from silicon photonics, Jens will walk through the timeframe associated with that and some of the key milestones as we're poised on the brink here of scale, production at scale for CPO. The point again here is this is a multi-year, almost a decade engagement in refining the technology together with some of the customers you see on the top line.
The other major milestone that I want to draw your attention to is a few years ago, we decided we had to change the organizational structure and some of the leadership of the company. You're going to hear from some of these leaders that we've added as part of FormFactor's executive team today. This was a really fundamental shift from the organizational structure and people that had led FormFactor to that point. It's basically responsible for the dramatic improvement in financial results we've had. The underlying financial results or underlying operational results have improved similarly. We look ahead, we're making some very large investments this year in Farmers Branch. Missy's going to walk us through how we're doing that, right? Why it's going to drive some of the attributes of the financial model, doubling revenue and more than doubling profitability.
This is a big organic initiative for us. Poised at this intersection of high-performance computing and advanced packaging, we've got the company in a very strong competitive position. We also need the capacity and capability to build all of these products, and that's why Farmers Branch and the global operations team is so central to FormFactor's future. I wanna come back and motivate the other side of this equation, advanced packaging and why it's important for a semiconductor test company. It's not obvious at first, again, if we take one of these example high-performance computing modules, a variety of GPUs, ASICs, HBM, networking, co-packaged optics, each of these subcomponents and the total collection is built using these techniques of advanced packaging, things like TSMC CoWoS, Intel EMIB, right?
These are the techniques the industry's using, right, as you saw in that video, to kinda make up for what undeniably is a slowing Moore's Law. It turns out that this is true not just in semiconductors, but in almost anything, there's some pretty universal math associated with breaking something apart into its subcomponents. A little bit of math on the chart, I wanna use one of the subcomponents of this high-performance computing module, HBM, to motivate why test and probe are so important in enabling advanced packaging and therefore high-performance computing. I'm gonna contrast something we did years ago, right? An HBM2 probe card, for, to test a four high stack, right? This was in the early days with our number one customer. HBM was just developing and hadn't really garnered the momentum associated with AI. You take four chips.
If you have an incoming defect rate of 1%, so of those 4 DRAM die that are coming into the stacking process, let's assume that 1% of them are not gonna work, don't work. That results, as you can see on the chart, in a 3%-4% yield loss in the stacking process. Not a disaster, right? Not great either. As we move to today and 16 high HBM4, that math compounds the 1% incoming defect rate into a composite defect rate for the stack of something like 15%, an additional 10% yield loss just by stacking higher. Again, this is simple universal math. In addition, you can imagine that the HBM4 16 high stack is a lot more valuable than the 4 high HBM2 stack.
A bunch of advanced node silicon, a lot of process technology has gone into building this stack. This is an expensive thing to scrap. We've got a situation where not only is the probability of scrap going up, but the cost of scrap is going up. Something needs to be done. That something is reducing the 1% incoming defect rate, the way to do that is through wafer probe and wafer test, screening for something close to what's been called known good die in the industry. If you do that, you can move left on this curve and work your way back up, right, to higher yields after the stacking process. That's why wafer probe and wafer test are so important in enabling advanced packaging. Of course, the roadmap has only more of this happening, right?
HBM is gonna continue to increase enabled by copper hybrid bonding. Future GPU products go from 1 GPU to 2 GPU to 4 GPU, right? This math is a fairly profound effect on the economics of the industry, and it's why two things are happening to our business. One, what we call test intensity is going up. Test times are longer, more probe cards are required. Basically, the chips are being tested across more conditions and with more test vectors going through the part. It also means higher test complexity, 'cause each of these chiplets needs to be exercised across a broader process parameter range.
If you think about things like speed, things like thermal conditions at different temperatures, to make sure each of these chiplets is good before it goes into the stack, driving down that 1% defect rate so that the composite yield goes up and our customers have an economically viable advanced packaging process. All right. Before I hand it over to the team, I wanna kinda map out for you how we're now organized, I said this was a major change for us, and give you some background on what these people own in our organization. You can see clearly from the last 3+ quarters of financials that something has changed at FormFactor. It really starts with these people and the teams they've recruited in, and the way they're coming together to drive FormFactor towards 2030 and the doubling of revenue.
The most recent change is promoting Aric to CFO, right? That happened a little less than a year ago. What he's done in partnering with me and the rest of the team is make sure there's absolute consistency between our strategic priorities, where we're spending money, and how we're executing to make sure we're delivering a financial return on every dollar spent and invested. You can imagine that's critical with us spending the amount we are on Farmers Branch. If we move to the center, this describes at a high level how we engage with the customers you've seen before. Aasutosh, our Chief Commercial Officer, will walk you through some of the details of that engagement. The details, we use the word customer intimacy, because the idea that we're building customized design-specific products is really, really important for how we interact with these customers.
Jens and Sudhakar, one of the things we did as part of this reorganization was get our business units and our business unit leaders really focused on developing and defining world-class leading products. They're gonna walk you through some of the technical advantages that we have and where their focus is on engineering, product development, and driving these new products that we co-develop with customers into strong market share positions. Missy, who joined us about a year and a half ago, has completely revamped our operations. We decided we were going to create a global operations team, getting leverage and benefits from the scale we have across all these different revenue streams. You can see in recent results how that's benefited us, especially on the gross margin line. Finally, Aric will bring us home with the details of our new target model.
Before I hand it over to Sudhakar, I want to frame what we view as FormFactor's durable competitive advantages. I talked about probe cards being customized to each customer chip design, and you can see some examples of how complex these things are over at the table there. To be able to do that on lead times well within a quarter and do it first time right, do it perfectly, so that the probe card and probe cards, 'cause we're now doing this in parallel, get there on time, but work out of the box. This is absolutely critical for success in the industry, 'cause our customers are depending on us to be able to ramp on their schedules. You know, as you've heard from companies like, you know, Nvidia, TSMC, things are only getting faster, right?
This idea of speed of light in a one-year high-performance computing cadence has fundamental implications for a company that has to design and ramp things quickly in scale. Another piece that's critical to our competitive advantage in probe cards is the MEMS technology, and Sudhakar will walk through that, but it's a foundational proprietary technology that we keep in-house. We've added to that, if I walk you back to our M&A strategy, Keystone Photonics offers an optical probe, which is the analog of these MEMS electrical probes as we move into the era of co-packaged optics and photonics taking a much more prominent role in compute.
Finally, as we're positioned, right, at this intersection of high-performance computing and advanced packaging, and really redefining what's possible for FormFactor, we need to be able to build all these compelling products at scale and drive very fast ramps in the overall industry. Those are what I think we think FormFactor's key competitive advantages are, again, as we redefine what's possible for FormFactor. Sudhakar?
Good morning, everyone. First of all, thank you for taking the time to spend with us this morning and early part of the afternoon. My name is Sudhakar Raman. I've been with FormFactor for about five years, and I've been in the industry for about 25 years, working at various companies like Onto Innovation, Veeco, and MKS, which used to be ESI in the laser world. As Mike mentioned, FormFactor is strategically positioned at the intersection of high-performance compute and advanced packaging. What this means is, with the demand for higher and higher computational power, we're seeing increasing demands on power and thermal density increasing, while advancements, essentially in the networking space, is driving the demand for higher speed performance.
By addressing the thermal and the high-speed challenges that we have in front of us, we plan to double the probes business revenue by 2030 by innovating our differentiated advanced MEMS probe technology solutions and delivering differentiated products at scale to our customers. I want to take a minute here to kind of point out how, what's the major inflections happening in the marketplace. To the right of the slide that you see here, because of the high-performance computing and advanced packaging trends that's happening, we see a pretty strong robust growth in the advanced probe cards market. We're seeing a growth of about $2.8 billion-$4 billion by 2030, primarily driven by all the major inflections that's happening across compute.
When I say compute, it is GPU and custom ASICs, SLS inflections happening in networking, and HBM and DRAM. If I take a moment and focus your attention to the bars, the kind of the size of the bars that you see, the width of the bars reflects the size of the market that we see today and continuing to grow in the very high single digits or double digits into the future. Our plan is to gain market share in all these major growth segments that we are highlighting here. As you know, we have a pretty good position in CPU, and we will continue to see the growth that we are gonna expect in the GPU and the networking sector. The other interesting trend is, as you see at the top of these bar graphs, what you see is all the different technology trends that is happening, right?
Whether thermal challenges or speed, or whether it's higher parallelism or reduced pitches. What's unique is you can see there is multiple trends that are common across multiple segments. Being the leading provider of the probe card solutions, FormFactor has a very unique competitive advantage having participating in all these major growth vectors. What do I mean by that? We're able to see the trends across different market vectors and able to apply our innovative solutions strategically across all these areas with speed and efficiency. As an example, as you can see on RF all the way to the right, we have decades-long market leadership in RF. RF, as we all know, kind of really lends itself in terms of developing high-speed probing technologies, which we're leveraging and migrating onto the high-speed networking space.
That's where we are starting to continue to see growth and continuing to see our growing market share using differentiated solutions here. Now, I want to take a moment and just kind of walk the audience through what the probe card is. It's very hard to see what exactly does a probe card do because it's kind of the design-specific interface that kind of integrates into an automated test equipment, ATE, and that efficiently flows electrons and photons to the wafer. Think of it as essentially as a massive parallel, high parallelism of lanes on a highway, right? That allows electrons and photons to travel efficiently from the test equipment onto the wafer. Because each probe card is, as Mike said, it's highly engineered consumable, it's really designed, custom designed and manufactured to the specific needs of the customer device.
What does this mean? Mates to a standard automated test equipment. What this really means is, as FormFactor as a company, we are highly motivated, and we encourage and maintain an open ecosystem. That is very important for us because we are able to serve the needs of the industry and the customer's roadmap very efficiently. If you see here at the top, it shows, it is also on the table on the right side, is a foundry and logic probe card, and in the middle is a DRAM or an HBM probe card with very high parallelism. Essentially, this kind of demonstrates the ability that we have to have extremely high parallelism, which is about 2,000 or even 3,000 devices under test, and with extreme high density of pin counts.
When I say extreme high density of pin counts, we're talking about tens of thousands to even 150,000 to 200,000 pin counts that we can develop in a single probe card. That is, again, design specific to the customer's device. On top of it, we manufacture these things at tight pitches. Just to illustrate as an example, just to give you some compare and contrast, a probe card, you can think of it as a single basketball sometimes or even the size of two basketballs. If you really get it down, what is the actual probing area where all the needles, as some people call it, or we call it probes, it's kind of the tip of a pencil. That's where you are seeing sometimes tens of thousands of probes being manufactured, assembled with very high accuracy and repeatability.
The probe itself is actually the size of a human hair. Majority of our hair in this room probably is somewhere between 50 to 75 microns in diameter, right? Imagine the level of complexity that is we're manufacturing here. As these power requirements and thermal requirements are continuing to increase, we have to design these probes using advanced MEMS probe technologies. Mike talked about the proprietary mechanism. These are manufactured using standard semiconductor, production-proven semiconductor techniques like lithography, CMP, chemical mechanical polishing, and advanced plating technologies. What this allows us to do is design it in such a way it has the ability to carry multiple amps of current. The, a good corollary I can give you is think of your iPhone. iPhone take a few amps to charge, an entire iPhone.
We send that amount of current with a single probe across tens of thousands of probes, and it sees somewhere between 100 to 200 degrees C of temperature at the same time. The ability to withstand these kind of things is where the innovation comes in. The point being, the technical barrier to entry is very, very high. In the subsequent section, I'll talk more about these technology innovations as well as architecture improvements that we're making to address the demanding needs of our HPC ecosystem here. This slide will now dive into the foundry and logic space. We'll have two sections, one on foundry and logic, another one we'll get into the DRAM and the memory space. Essentially, you can see here, when Mike presented the entire HPC ecosystem, we talked about all the different devices.
What we'll focus in this section is gonna be all around the GPU, custom ASICs, the networking, and the co-packaged optics. As you can see, there are four major trends that are driving essentially the technology inflection in this area. Number one is essentially increasing power density. What does that mean? The devices are starting to see, particularly the GPUs, somewhere between 2,000 W all the way in the future, up to 10,000 W in the device. What this means is that thermal density is continuing to increase. At the per centimeter square, you're gonna see 200 W, which means you have to dissipate this heat as the customers are evolving the needs of the HPC compute.
That's where it puts the onus on suppliers and companies like us to continue to innovate. Number two, several of the high-speed test content from final test and system-level testing is kind of shifting left. They want to test every single die as a known good die to mitigate yield loss, which Mike talked about through the HBM example. It applies the same exact logic, and composite yield math applies here as well as in foundry and logic. Number three, you see those red bars that's connecting different chips. We call it D2D here, die-to-die. Essentially, they are basically silicon bridge, some people call it. Intel calls it EMIB. That's basically what that is. It's really starting to have significant adoption in the, both in the CoWoS as well as non-CoWoS architectures. Four is co-packaged optics.
With higher networking speed going from 112 to 224 to 448, it's not only demand, driving the demand for electrical probing, but it's also driving the demand for optical probing. Jens will talk more detailed about the CPO and the advances we're making there. At the end of the day, what does this mean to FormFactor? For the past several years, we've been stating that there are two major vectors that is starting to drive and is starting to play out, which is test intensity and test complexity. It's really starting to play out with HPC and advanced packaging intersection.
As you can see here, with the number of devices that needs to get tested, that needs to be known good dies, it's increasing the test intensity, which means more number of electrical probe cards and more number of electrical and optical combined probe cards, electro-optical probe cards, as we call it. Second, with the test complexity increasing, it really plays well into our advanced proprietary MEMS manufacturing and technology development, where you need to manage the thermal envelopes as well as the high speed with high pin count and high parallelism. All of them are converging together, which is driving the advancements of these innovations even further.
I'm gonna take a moment here and kind of dive a little bit deeper into kind of what are we doing as FormFactor to really future-proof the foundry and logic requirements and the technology inflections that's coming along as part of this whole high-performance compute. On your bottom left graph that you're seeing here, we picked the two major vectors that is driving the technology and the technical requirements forward, which is power, which translates to thermal, and speed. As you can see, the GPU, the HPC compute ecosystem is continuing to drive the advanced MEMS technology development. I wanna kind of take a step back and address the custom ASIC piece of it. Historically, in the past, custom ASIC has been about 2-3 nodes, technology nodes behind. What does that mean?
It really means the demand for power and speed hasn't been there, which means you can actually cater to those needs of those custom ASICs with essentially what I would call as a medium, mid-end, mid-tier type of MEMS probes. What we're seeing is the custom ASIC is now starting to be adopted with more of advanced technology nodes with like 3-nm type nodes, increasing number of transistor cores and density increasing. The power density is increasing, thermal requirements are increasing, pushing the custom ASIC as well into the advanced MEMS category. I shift to the right side of the slide here, essentially what you're seeing here on the top right of the slide is a, what we call the Apollo probe card. On the table, it's that big rectangle one that you see.
That's our foundry and logic probe card. You see there's vertical lines that are the advanced MEMS probes that gets integrated into the ceramic plates, which also is a highly engineered with special proprietary coatings that we develop. That gets integrated onto a metal multilayer organic. It's essentially an interposer which can spread the electrical signals onto a PCB, which gets integrated onto the automated test equipment platform. On your right, you're seeing is an advanced MEMS probes, examples of a scanning electron microscope, SEM, pictures of actual how the probe tip looks. That's the size of the human hair I'm talking about. It's actually the tip that actually probes at the wafer pad or the wafer bump onto the customer device, right? Again, these are manufactured using advanced semiconductor techniques like lithography, plating techniques that I talked about.
I wanna kind of drag the attention to the two areas of innovation. One is in the power and thermal requirements. Why advanced MEMS is important? Because it gives us the ability, by using these advanced semiconductor techniques, it gives us the ability to kind of marry or mingle different metals and alloys and come up with composite metals and composite materials, essentially doing metallurgy engineering that is specifically designed to the customer's device needs. This is very unique because only through semiconductor manufacturing techniques you can achieve this. You can see there on the middle portion of the MEMS probes, there are multi-layers that are being developed. That helps your ability to carry much higher currents and at much higher speeds at the same time.
You're designing in multiple requirements with a single probe or multiple types of probes that are very, very application-specific needs. The second one is, I want to highlight the attention to the bottom right, is on the probe card itself as we go beyond the probe tips into the ceramic plates, where we're actually engineering new thermal dissipation materials that gets manufactured and coated onto these ceramic plates. That has the ability to dissipate the heat that's coming through the probes so that you increase the uptime and predictability of the yield binning and the performance for the customer. The other axis that you see here is the speed. Again, with speed, what you need is extremely high signal integrity, meaning very, very good, clean signal to very low signal-to-noise ratio with very high signal integrity needs to happen.
When we do these composite metals and metallurgy engineering, we ensure that the way we design these probes, develop and manufacture it has very low impedance, very low resistance path to go through. Again, it's at the tight pitches that the customer needs, whether it's EMIB tight pitches or I/O that are much tighter pitches. Those are the two major differentiating factors that we continue to evolve as we go from our advanced MEMS to even future advanced MEMS. We're pretty confident as we look at the requirements, our MEMS can confidently approach the requirements and even beyond. That's what we mean by future-proof.
Switching gears onto HBM, as Mike talked about the multiple different stacks from 2 to 4E, what I wanna really highlight here is there are two things again, test intensity and test complexity playing out, as we've been talking about. When you look at HBM3 to HBM4 to HBM4E, we're seeing the number of stack going up, which increases the number of probe card needs to really test these stacks. As an example if I give you can think of test times increasing and the test insertions increasing, what does it mean to the overall, the growth of the market? It increases the test intensity by about 25%-30% when you go from 3, 3E to 4, 4E. We're expecting to continue to see the same as it migrates forward.
The other interesting piece is test complexity, right? There are two areas within the test complexity that's critical to pay attention to. One is our differentiated SmartMatrix architecture provides the need to test at very high speed at very high parallelism. This is the industry's only production-proven platform that is able to test and give the customer the ability to do high parallelism productivity at speed and enables a customer to test hundreds of HBM stacks simultaneously. That's number one from a complexity standpoint. Number two, as you can see, the base die migration from 3E to 4 to 4E to 5, we expect to see we're starting to see a lot of the hyperscalers and the end customers working with DRAM and HBM makers to integrate more of the logic functionality into the base die.
That's what they're calling it as custom HBM. The reason this is happening is really to optimize the AI performance, the power consumption, and having the ability to have more space on either a GPU or XPU to have more compute power there. They're shifting more into the base die of the HBM. What does that do to us? It means we need to start testing logic IP on a memory base die along with the memory IP. This provides a very unique opportunity to differentiate for us because FormFactor is the only company that has production-proven platforms on both foundry and logic using our Apollo architecture, as well as in the DRAM HBM space using SmartMatrix architecture. With that, as we migrate into the, hey, what does advanced MEMS do for this industry from a HBM DRAM perspective?
Again, here we're picked up a parallelism and speed as two major areas. The same set of advantages plays right here into the advanced MEMS techniques that we talked about. More importantly, we can tailor the need and then customize this metallurgy and the different metals that we use to address the needs of memory probing. It's memory as in DRAM and HBM probing, because for logic, they test different type of functionality test, whereas in memory in DRAM and HBM, it's very high parallelism, but its speed requirements has to be tested at the same time. When you see the SmartMatrix architecture on the right side, we have integrated some unique thermal management mechanisms to take care of the needs of the customer's different temperature that they wanna test at in a single sometimes in a single test insertion.
Essentially, on top of the advanced MEMS, we have also designed in some new thermal dissipation and thermal management into the architecture that gives us a very unique ability to say customers can test it from either from room temperature to sometimes as high as 175 degrees Celsius without changing the probe card for that test insertion. Again, the third one is, we talked about the custom HBM migration with the test complexity increasing. What you're seeing on the right side of the 3D MEMS HBM probes is a 2D vertical MEMS probes. It's pretty similar to what I showed you a couple of slides ago. This is where we are starting to see the fusion of 2D MEMS, advanced MEMS technology being integrated into a very high parallelism, high-speed testing into the DRAM HBM-based eye architecture.
This gives us a very unique opportunity again, to differentiate and continue to grow our market share and grow our revenue. Kind of in summary, we talked about the test intensity increasing, driven essentially by the HPC and advanced packaging intersection, expanding this addressable market, SAM, as we call it, with the test complexity both in foundry and logic, where the thermal and power are driving the needs for differentiated solutions, and in HBM, where there's a fusion of both with the custom HBM, logic and memory IP are being fused together to test. It gives us a unique opportunity for FormFactor. Lastly, we also are seeing a new vector of growth, which is the convergence of electrons and photons and co-packaged optics.
This is gonna be driving the need, which Jens Klattenhoff will be talking about, of how we're gonna continue to evolve that market across multiple test insertions. Kind of in closing, it's increasing test intensity and complexity in the HPC architecture. We believe our continued focus on developing and manufacturing differentiated, innovative solutions at scale will lay down the foundation to double our probes revenue by 2030. Now we will be in a position to kind of continue to further along on the technology evolution. With that, I would like to thank you. I would like to introduce Jens Klattenhoff, who's our General Manager for Systems Business Unit, who will talk about the CPO.
Good morning. Thanks for joining us today. As Sudhakar already said, my name is Jens Klattenhoff. I'm with FormFactor since 2016, and since 2021, I'm running the Systems BU as the Senior Vice President and General Manager. My background is in laser technology, which is a pretty good fit for co-packaged optics and silicon photonics. In my career, I worked for several optical measurement companies like Polytec and Baumer before. We are redefining what's possible. FormFactor is the global leader for wafer-level test systems, and we are extending this leadership to a new era of computing. Over three decades, the Systems Business, formerly known as Cascade Microtech, which was acquired by FormFactor in 2016, builds its position in foundational markets like RF testing for mobile and connectivity devices, DC testing for logical and mixed-signal devices, and high-power testing for automotive semiconductors. Today, this industry is shifting.
Advanced networking and high-performance computing are driving new demands for higher bandwidths and lower power consumption, as well as rapid scaling for artificial intelligence and next-generation data centers. This shift is a structural growth opportunity for FormFactor when silicon photonic is moving into the high-volume manufacturing for the first time. Our technology leadership, the global scale, and the customer relationship we have with all leading co-package optic customers positions us to set the standards for optical testing in this area. In the next couple of slides, I would like to show you how co-package optic will change our systems business, how FormFactor is well-positioned in this market, and how our strategic partnerships will help us to make this shift happen. By 2030, we are actually planning to double our revenue within the system segment. I like to show you how.
Co-packaged optics will be the engine of our growth. Traditional interconnects, electrical interconnects, have fundamental limits. As data rates scale, couplings can't provide the required bandwidths and the power efficiency which is required for hyperscaling computing and AI. Co-packaged optics will solve this. New architectures like TSMC's CoWoS technology are actually integrating optical I/Os into the package close to the compute die. Leading platforms like NVIDIA's Quantum-X800 switches are already using this technology and switching to co-packaged optics designs. This creates a complete new opportunity for FormFactor, as I said, when co-packaged optics is moving from design to high volume manufacturing. As Mike and Sudhakar already talked about advanced packaging, advanced packaging is, it's critical in advanced packaging that we are testing early because we are combining very expensive silicons, substrates, and optical components in one device.
Early wafer level test is essential, and that's exactly where FormFactor leads. We see silicon photonics and co-packaged optics to be a multi-hundred million dollar opportunity by 2030 for us, which is a significant faster growth than our traditional markets. We are not just entering the silicon photonics testing market. We created it from lab-to-fab. In 2017, we deployed our first 300 mm semi-automatic probe station to an institute in Germany. Since then, we have installed over 160 systems globally at startups, universities, leading co-packaged optics customers, foundries, and other companies. The silicon photonics test system is known as the industry standard for optical testing, which is the foundation of co-packaged optics. In 2025, we acquired Keystone Photonics, the leading company for advanced probing technology. Keystone Photonics brings two assets to the company.
First, their scalable optical probing technology, and second, their deep developed partnership with leading co-packaged optics customers. This strengthens both our technology leadership as well as our market position at the right time, because 2026 is an inflection year when co-packaged optics is moving from design and niche production into high volume manufacturing. With our Triton system, we are providing a fully qualified production-ready test cell for co-packaged optics testing in high volume manufacturing. We already deployed several systems, as you heard from Mike, and we saw first revenue impacts with qualification and niche production in 25, and expect strong growth in 2026 and beyond when high volume manufacturing starts to increase for co-packaged optics. Like any other advanced packaging technology, the economics of co-packaged optics favors early wafer-level measurements. Let me talk a little bit about this test strategy.
In semiconductor manufacturing, we are testing in different production points or insertions, as you heard from Sudhakar for HBM, for example. The known good die strategy, testing very early in the process, is very critical for advanced packaging. As I said, you're packaging expensive silicons, substrates, optical components. At insertion one, testing the photonic IC on a 300 mm wafer is where FormFactor operates and leads. We are catching defects very early in the process before high value components are getting added to the package. Testing here at this point reduces cost of good die by up to 5 times compared to other insertions. We also accelerate the yield learnings and can isolate photonic issues from assembly and production effects. Later insertion exists, they do not scale, at least currently. High complex test systems, long cycle times, low throughput are not economically yet.
FormFactor focus on insertion one today. Insertion one actually represents the foundational optical technology which is required for all other insertions as well. During the evaluation and during when the high volume manufacturing for co-package optic gets more matured, we are able to deploy the technology which we have developed for insertion one into other insertions later on. Our co-package optic leadership with Triton is built on deep partnerships from R&D to high volume manufacturing or lab-to-fab. We engaged early with a leading high-performance computing company directly at the development stage. Engaging that early helps us and the customer to shape testability into their devices. And to design scalable co-package optic architectures for the future. As a result, FormFactor is providing the whole life cycle and supporting the whole life cycle from lab-to-fab.
I showed you how co-packaged optics will change our market, how we are ideally positioned in this market, and how our strategic partnerships will help us to make this shift. Along with Quantum, this positions the system segment for durable and high-value growth. Now I'm excited to show you a video of our production-ready test cell for co-packaged optics in high-volume manufacturing called Triton.
Introducing Triton, the industry's first production-ready test cell platform optimized and engineered for high-volume silicon photonics and co-packaged optics. Designed to meet the demands of modern photonics manufacturing, Triton brings a decade of CM300 silicon photonics technology from advanced R&D into scalable production. Built through a strategic partnership between FormFactor, Advantest, and TEL, Triton combines proven test expertise with manufacturing-ready automation. The result is a single integrated platform delivering precision, repeatability, and automation for high-volume manufacturing scalability. At its core, Triton delivers the world's most proven and accurate automated photonic alignment with nine-axis nano-precision. FormFactor's Pharos optical probe technology achieves industry-leading ultra-low coupling loss from fiber to device. Triton supports low-loss edge and surface coupling for both grating coupled and edge-coupled devices within a single system. Single-click automated calibration, optical scanning, and measurement dramatically reduce setup time and accelerate production readiness.
Fully integrated with Advantest V93000, Triton enables fully automated wafer-level optical and electrical co-measurements in one unified test flow. Compatible with FormFactor's MF80 and MF130 high-density probe cards and optical fiber array units, Triton maximizes test density and throughput. Exceptional stability delivers coupled power repeatability within 0.3 dB across long production runs. Compatibility with overhead transport systems and SECS/GEM factory automation enables rapid deployment into manufacturing environments. Fully automated wafer handling, calibration, and measurement reduce operator involvement and increase throughput. SiPh-Tools manages wafer-level optical workflows, calibration, and high-density test execution. TritonConnect software simplifies probe station control for manufacturing operators. Together, these capabilities shorten test time while delivering the accuracy required for AI, co-packaged optics, and data centers. Triton, production-ready silicon photonics testing engineered to scale.
I forgot that part.
Well, thank you, Jens, for your presentation. I think we're at the point where we're ready to take about five minutes break. Help yourself with coffee, pastries, et cetera, and we will kick it off with Aasutosh and the commercial part of our business. I will see you at 11:05 A.M.
[Break]
Which one? Hello, hello. Okay, can you guys hear me? I guess nobody can hear me
Can you hear me? Yeah.
It's back to Mary.
Excuse me. Can we all try to get seated? We're going to proceed with the second half of our presentation. Just give it a minute. Please take your seats, grab your coffee, grab your water, and we will start momentarily. One second. All right. Aasutosh, the floor is yours.
Check one, two, three. Good morning, welcome back. My name is Aasutosh Dave, and I'm Chief Commercial Officer at FormFactor. Been with FormFactor two years now. Prior to that, I used to work at ASML, AMD, Mentor Graphics, Siemens, and I've covered the entire semiconductor design to silicon life cycle, and I'm super excited to be part of the FormFactor team. The question always comes up to say, what's exciting about FormFactor? It's all the advanced technology that's happening with advanced packaging, hyperscalers, and this is where test complexity that Sudhakar was mentioning, co-packaged optics, the industry is headed in a completely different era now. Today, it's about one core idea, and Mike mentioned about this.
Our strategy starts with the customer, and it's built to win in places where semiconductor industry is accelerating the fastest. When we talk about FormFactor strategy, it's a simple strategy, right? We are talking not just about redefining what's possible into the product areas. Sudhakar and Jens talked about products that we have in, into Probes Business Unit, into test systems, co-packaged optics. Our strategy is taking that beyond what we have from the product point of view, but extending that to customer value. How do we gain that, right? It's just extending not just market leadership, but how do we grow our market share too? Looking at this, we focus on global support, talking about customer intimacy plan and market leadership.
At the end of the day, we wanna make sure that we are driving all this value to the customer, and then that helps us continue the leadership position that we have. What differentiates us and what keeps us ahead than competition is our support. When you look at this global scale, it's not just the scale for the sake of it that we are spread across the region. I think it is more about we are at locations where it matters the most for the customers. This is a place where customers are ramping fast. This is a place where customers are ramping production cycles, and we are at the forefront of it.
Any execution misses are super costly for the customer, so it makes extremely important for us to be there with the customers, where customers are ramping fast and simultaneously at multiple locations, and it becomes important for us to be there. If you look at what's important from the customer point of view, all these geographical locations that we have, customers want to ramp faster. We are talking about how these customers are advancing into the next technology nodes. Making sure that we have our design team, we have our applications team, we have our service and support organization into all these different regions helps enable these ramps along with the customer. We have boots on the ground to work closely with the customer. It helps with the faster response time.
This is also what is mandated by the customer, but we are at the forefront of it. Global scale and support infrastructure is important. It helps us basically ramp faster. It helps us penetrate with the customers where they really need the support, and it also helps lock in with these customers where it becomes more, it brings a stickiness onto the solution that we are providing along with the customers. Switching gears a little bit. We talk about global support. Now, this question always comes up to say, what does a customer intimacy map look like? What is changing, right? Breadth is important, but what really differentiates us and how we engage, how FormFactor engages with a customer is we are not engaging with a customer at one single point in time. We are engaging with a customer for the entire semiconductor life cycle.
Using this example, if you look on the left side, we are engaging from early R&D to high-volume. Early R&D engagement helps us work closely with the customer to define the test architectures, which are changing significantly at a faster pace than we were talking about, that Sudhakar was talking about earlier. Staying involved with the qualification and driving all the way to the high-volume manufacturing, that continuity is super important and super critical. The early engagement, if you look on the right side, we engage with the customers on the R&D side, but we stay engaged with the customer all throughout the ecosystem and all throughout the life cycle of the product. You'll see that we are deeply engaged with multiple teams at a customer, whether it's R&D, yield management, quality, supply chain, and high-volume manufacturing.
We work with all of these customers together very closely, and we are embedded with all these customers. If you look at this multi-threaded engagement, it's strategic, right? It becomes super important for us to make sure that we are co-developing not just one thing at a time, but we are co-developing as a solution and the value that we provide to the customers. This co-development drives a pull-through, it increases lifetime value, and materially strengthens us for any competitive barrier. This is also where the dynamics are changing that we were talking about previously, is to say this entire life cycle by which whether you're taking high bandwidth memory applications, whether you're taking HPC high compute applications and advanced packaging applications, having this life cycle and ecosystem is critical.
What customers talk to us is that we are no longer just a supplier, we are a co-development partner for all these customers. I'm gonna change gears. We talked about global support, we talked about customer intimacy map and how we are embedded with the customer. This slide illustrates something more important. We are present not just through the semiconductor value chain, but we are trusted partners with these customers, and we are recognized by these customers. From early-stage R&D to high-volume manufacturing, we are there. These customers are the one where any execution misses are not allowed. This is the recognition is not just from the trophies point of view, but it's a value that we are providing to the customer. It's a testament from this customer that our operating model works, that our execution model works.
When customers are looking at us, it's not just one particular team. They're not looking at us as a product or a technology. They're looking all throughout the ecosystem that we are talking about, as the technology has to work perfect, quality, operations, service, support, all has to work perfect for us to get recognition from customers like this. Another thing that you will notice is, there are some repeat customers. These repeat recognitions, it speaks more about consistency that we are delivering and the value to these customers. It's not like one-off awards that we are getting, but it is repeated recognition that kinda testifies again to say we have to keep repeating it. We have to make sure that we are providing the value not one time, but consistently.
When you look from portfolio point of view, Mike had mentioned this in one of his slides. When you look at this perspective, right? We are in the semiconductor ecosystem. We are at all these semiconductor leaders in this industry, whether you're taking high-performance computing, AI applications, memory applications. We are working closely with these customers. What also is critical is, as these customers grow, as these customers are driving the market, we grow along with them. I wanted to make sure that again, these awards recognition and the customers that we are engaging it's not just simple awards, but it's recognition from these customers on the value that we are providing to them. We begin talking about our strategy starts with the customer, and we are leading at the forefront of these customers.
We are number one over there, and I'll end over there. We begin talking about leadership. I'll close that as well. As longstanding number one leader in this industry, we continue to grow our market share. We continue to be number one. Everything that we are talking from the product technology development point of view, from the products, making sure that we are supporting the customer, we are bringing that value to the customer. That happens with differentiated customer value, lifecycle partnership, global support. Last but not least, is reliable scale. What I mean by reliable scale is we talk about capacity, we talk about operations. With that, what I would like to do is I would like to hand it over to Missy Figueroa, Global Operations Leader. She'll talk a little bit more about global operations.
Hello, everyone. I'm Missy Figueroa. I'm the Senior Vice President of Global Operations, and I've spent the last 28 years building and scaling semiconductor manufacturing. 24 years at Texas Instruments, three years at Wolfspeed, and the last 18 months strengthening and unifying global operations here at FormFactor. Over the last year, we've been redefining what's possible. We've fused test and measurement innovation with the rigor and strength of semiconductor manufacturing, and we've done that by building a unified global operations team that didn't exist in this form before. We started by bringing together teams that had historically been distributed and aligning them under one operating model, one set of standards, and one shared vision. We strengthened that foundation by bringing in new leadership with deep semiconductor and manufacturing expertise, and we integrated them with the strong legacy talent already here.
This combination of fresh perspective and institutionalized knowledge has been a powerful catalyst. Our new unified operating model is already delivering results. Everything you'll see on the next few slides, the performance gains, the operational discipline, as well as the impact of our new site, all stem from this transformation. Over the last three quarters, we delivered meaningful, measurable improvements across our operations. Those gains did not happen by accident. They came from a disciplined approach to yield, execution, and efficiency. First, yield. By reducing defectivity, scrap, and operational friction, yields have steadily been improving. As you can see on the graph, we'll continue to improve them through process improvements. These improvements have translated directly into better margins, more predictable output, and faster startup learning. The second is disciplined execution.
By tightening cycle times and improving delivery predictability, we've not only strengthened our customer confidence, but we have also reduced our working capital needs. What do I mean by reducing our working capital needs? Well, we've been able to get more out of our installed tool base, otherwise boosting the asset utilization that we have today. We've further done that by driving repeatable processes across each of our shifts, lines, and sites. Our new operating momentum, or our operating execution, is already delivering results. The third driver is around efficiency, both as a cost-cutting exercise and a scaling enabler. Our goal here is to grow our output without growing the complexity of our operations. To do that, we must operate in a fundamentally smarter way. What does that mean?
That means that we focused on things like reducing non-value-add activity, simplifying work streams, harmonizing processes, and also automating manual tasks. Our job here, or the goal, is really to make execution easier, and this allows our people to focus truly on the work that moves our business forward. Again, these are the three strengths that have accelerated the performance you've seen over the last three quarters, and they're the same durable, repeatable capabilities that we plan to scale into our new factory. Farmers Branch. Farmers Branch is built for performance and scale. It's not just a building, it's a next-generation operations hub designed to maximize flow, increase throughput, and also provide sufficient growth for our future. We got a strategic head start by selecting this location. It's a brownfield. Why did we select a brownfield?
Because by doing that, we were able to direct capital into capabilities rather than building infrastructure. It's positioned in a fast-growing semiconductor corridor in North Texas with deep access to talent, as well as strong local and state incentives that have further strengthened our project economies. Since purchasing this site, we've really been intentionally laying the groundwork for the target model that you're gonna see today. When we purchased the site, we talked about the MEMS fab expansion, right? We're gonna expand our MEMS fab from Livermore into Farmers Branch. This frees up some critical capacity for us in Livermore that allows us to continue to develop our SmartMatrix roadmap that Sudhakar talked about earlier today, and it positions us for continued growth. Earlier this year, we also announced the site closure of our Carlsbad A&T.
We plan to consolidate our A&T line into Farmers Branch. This drives operational effectiveness and also better economies of scale for this site. I'm gonna spend the next few minutes talking about how we plan to ramp, what approach we're gonna take, and then we'll talk a little bit more about kind of the life cycle of a factory and how that lays out and what our milestones are. First, let's talk about how we plan to ramp. We plan to ramp in a modular fashion. This keeps execution risk low, as well as deploys capital only as it aligns with demand. Every phase is directly tied to revenue. There'll be no speculative capacity and no excess volume. The second is around bringing in new tools and processes, right? As you bring in new processes, we are planning to optimize those for scale.
As you ramp a factory or as you transfer technology, there's really three approaches that you can take. The first one that I like to think about is just a whole new redesign. You typically see a new factory with a new redesign when you're transferring from diameter. So think 150- 200. This means that they're bringing in all new tooling. They're having to reintegrate on that tooling. It's extremely high risk, and typically the timeline is much longer because it takes a while for the performance and the yield gains to get there. The second is a copy exact. This is probably what most of you are more familiar with. You hear that term a lot, right?
Copy exact means you pick up the site, basically, and you put the same tools and the same processes in the new site. It's very low risk. When you do that, you actually inherit all your legacy problems into that factory. You can take a third approach, which is the approach we're taking, and it's a really balanced approach. It's actually kind of a merging of the first two that I talked about, and it's a copy smart approach. We're taking a copy smart approach because we already know that we have proven technologies and capabilities. We're gonna bring those over, and we're gonna think about bringing in some best industry practices.
By bringing in more advanced tool sets, our current technology, along with some industry best practices and process development, and our learnings, we can optimize this site for cost, scale, and long-term reliability, which really just means that we have a faster startup, more predictable output, higher quality and better performance from day one. Now direct your attention to the bottom of the slide. If you look at the bottom of the slide, this is industry best practices for phasing of a factory life cycle. Phase I, planning and pre-startup. Everyone goes through this, right? This is the design, the layout of the tools. Phase II, pilot and qualification. As you bring in your first initial line, you go through a pilot and qualification of every single tool and every single process, and then all combined. Phase III is your initial production ramp.
Once you achieve that initial production ramp, now you really just move into the whole life cycle of a fab, which is ramping to target capacity and continuous improvement. Those cycles go on and on, which is why fabs can exist for 30- 50 years. Today in Farmers Branch, we're actually in both phase I and phase II. We're in phase I because as we announced, this site had a clean room, but we needed to further expand our clean room. That construction is already underway. We're in the middle of completing both the design and layout, as well as having vendors come in to do the construction. The second piece is really pilot and qualification. We're also in this phase. Why? Because when we announced the site, we were very specific.
We purchased this site because it already had existing clean room, over 50,000 sq ft. We've already been installing tools, and as those tools are getting installed, the pilot and qualification is underway. This overlap was intentional for us. We chose this specifically because it meant a faster ramp and reduced readiness risk. Our milestones are set. You can see here below, we're set to have both our MEMS fab line and our A&T site qualified by the end of this year in December, with initial production ramp starting in 2027. Once we complete those qualifications, we'll then move into a phased ramping, which means we'll start slowly increasing our output. We'll be bringing in additional tool sets, from there we'll continue with, you know, operational improvements, things like tightening process windows, bringing in new innovation, so on. Okay.
Now we've talked about Farmers Branch. Let's talk about how we're gonna engineer it to get us to the next level. Earlier I said Farmers Branch was built for performance and scale. To further expand upon this, I wanna spend a few minutes talking about what we're doing at the site. First, we're equipping this site with advanced tool sets. Advanced tool sets actually allow higher throughput in a smaller square footage, which means we're gonna be generating more revenue per square foot in a clean room. You know, clean rooms are pretty expensive, so the goal here is to drive lower COGS. We also have a flexible, scalable layout. This means higher output densities, and it also reduces waste around non manufacturing production space. This creates a complex high output environment that scales better than traditional setups.
We have shorter travel paths and easier handoffs. The second is really around embedding digital tools and AI from day one. We're gonna have real-time data, analytics, automation, and robotics driving decisions across our floor. Our quality control is also automated. We can detect and inspect millions of springs with speed and precision. Not only is our AI detect defects, it also classifies them, meaning faster rapid response, root cause analysis, and further yield improvements. Third, we're ramping this in a capital efficient way. What do I mean by that? With these advanced tool sets, the processing capabilities and the batching actually means that we can ramp in the phased way that I talked about. This means that we're able to keep our investments disciplined and the return rates high.
Farmers Branch and our new operating model is unlocks the next wave for FormFactor, and it unlocks what matters most. Profitable scale, high yield, automated cost-efficient production at a global volume capability that expands our margins with harmonized processes and tooling. Resilient growth, a diversified footprint with a secure supply chain. We have fungible manufacturing lines through our process improvements. This also allows for on-time delivery in any volatile market. Trusted execution, unbeatable quality, and a digital backbone that delivers identical data, metrics, and outcomes, instilling our customers' confidence. Finally, looking forward, Farmers Branch, a fully fungible, digitally integrated site that comes online with the same performance DNA that you've come to expect from FormFactor. It positions us to scale with demand and allows us to deepen our customer partnerships. Profitable scale, resilient growth, and trusted execution. That's what our operations makes possible.
With that, I’d like to introduce our Chief Financial Officer, Aric McKinnis, to share our new target model.
Thank you, Missy, and good morning, everyone. Thank you for joining us today again. I'm FormFactor's CFO. Been in the role for a relatively short period of time, going, getting close to one year now. I joined FormFactor actually back in 2019, so I've been with the company for a little while. Prior to that, I was with a company called Electro Scientific Industries that was subsequently acquired by MKS Instruments, and prior to that, spent about a decade in public accounting with Deloitte & Touche. The team did a pretty darn good job of laying out what our roadmap is going forward and how we're gonna get there. I'm gonna spend some time talking about what that looks like in terms of our financials, if you will.
Before I get into that, I just wanna say I'm really excited about FormFactor's position right now and our ability to execute on the plans that we're laying out here today. The reason why I'm excited about it is because, as you've heard from each of the people that presented, there's real fundamental substance here. There's a durable approach that we're taking. This is more than just a spreadsheet exercise. You've seen the renewed commitment that we've been demonstrating over the past three quarters or so around execution, both from an operational perspective and also from a financial perspective. You've heard from Sudhakar and Jens about our leadership position in the markets.
We have an enviable position right now of being the number one probe card supplier in the world, and we have a real ability to leverage the capability that we've developed over years, those intimate customer relationships that you heard Aasutosh talk about, and leverage those into these growth opportunities that we see in the intersection of high-performance computing and advanced packaging, two of the most powerful trends that we're seeing in the industry today. We have the breadth and depth of knowledge, and I think you've seen today, as people have spoken with you, the leadership change that we've also generated. I think we also have the leadership that's required to drive our plan forward. We've got a bold goal in front of us. We're gonna double our revenues by 2030. We're gonna drive more than double EPS in that same timeframe.
Let's dive into some of the details here, and let's start first by talking about where we've been. Our last target model was really established in 2020, and in that target model, we were targeting around $850 million in revenue, 47% gross margins, with operating expenses as a percent of revenue of about 25%, and a non-GAAP EPS of $2 per share. We've now demonstrated achievement of that target model. Our most recent quarterly results exceeded that target model performance. If you include the latest guidance that we provided for Q2 2026, you can see that actually on a trailing twelve-month basis, we've also demonstrated this model in a durable way.
As we look forward, as you heard in the prior presentations, we see significant growth in our markets. We're expecting the market to grow to $4.5 billion. In that environment, FormFactor expects to double its revenue to $1.6 billion. You've heard the leverage that we wanna drive in our operations from Missy. We believe that we can achieve 55% gross margins at these revenue levels. We're going to continue to drive operational leverage and discipline through our organization, generating further leverage on our fixed costs and on operating expenses as a percent of revenues of 23%. Ultimately, a non-GAAP EPS of $5 per share over this timeframe. You heard a lot about the market opportunities in front of us.
And you heard about them kind of by segment or business unit and specific applications. Let's pull that together for you here a little bit and talk about our total addressable market. Today, our total addressable market is around $3.1 billion. We expect that's gonna grow to four and a half billion dollars. You heard what's driving that, right? It's this increased intensity, this increased test complexity that we're seeing driven by high-performance computing, driven by advanced packaging, things like HBM, things like co-packaged optics. That's driving the expansion in our market, if you will. In addition, we expect to take market share. That's another element of that growth, and that multiplies the market growth percentage here.
What you, what you can see here is the market is expected to grow, or we expect it to grow, and those expectations for us are based on third-party research. We also take into account our knowledge of the markets. Based on those inputs, we believe the market's gonna grow at about 8% CAGR over this timeframe. If we look at what we're planning for FormFactor, we're planning to do double that growth rate in the same timeframe. We're gonna do that by the leverage that I mentioned on market growth, amplifying that with market share gains. You heard from Sudhakar and Jens in particular about some of the competitive positioning that we believe is gonna enable that. You also heard from Missy about how we're gonna drive operational excellence that's really important for our customers.
If we get into the different components from a market perspective, again, you heard this kind of in bits and pieces earlier, but this will give you a sense of scale. If you look at the growth trajectory from $3.1 billion today to a $4.5 billion in 2030, you can see the major what we view as being the major elements of that from a market perspective: co-packaged optics, high-bandwidth memory, GPUs, and custom ASICs. In addition, we have growth in our base business indicated here. The thing I want you to take away from this is how broad-based this is, right? We've got some buzzwords in here. There's some very popular topics. There's things that are very exciting and very fast-moving, but it's more than just that, right? We're growing fundamentally. We're changing how we do things.
We're changing how we approach the market. We're so well-positioned as a company. We have multiple horses in this race. I think we're uniquely positioned in that respect. Market growth, market share, working together to drive a multiplier effect on our revenues, allowing us to grow from the 2025 base level of around $800 million per year to double that, $1.6 billion. As we do that, we expect to generate leverage on that volume and also, again, leveraging the value of what we provide our customers to generate even higher gross margins, generating a 1,400 basis point increase in gross margins from our 2025 baseline. We're gonna do that on what we think is relatively similar mix.
The important element of that is that we aren't relying on luck, we aren't relying on mix to get this done. We're gonna drive this out of fundamental, again, market business value drivers. We have good line of sight to this growth rate, I believe. We have good line of sight for how we can drive value for our customers. Let me dive into gross margin roadmap here a little bit. As I mentioned, 1,400 basis point increase. That's a big number. I would divide that progression into three main pillars, those are volume, operational excellence, or you could also call this execution, innovation. You heard all of these elements. The team makes it really easy for me. You heard all of these elements in the prior discussions from the team, right?
We expect the relative weighting of these components to be like you see here on the screen. About 600 basis points from volume as we double revenues, about 500 basis points from operational excellence, and about 400 basis points from innovation. Farmers Branch is a major enabler of volume and also innovation and operational excellence. The contribution, we expect Farmers Branch to ultimately be accretive to our growth, and it's big part of this roadmap. You see it's a piece of both of the two middle and last pillars here that you see. Recent results demonstrate the credibility of this plan. Again, a very significant change in the trajectory of our gross margins. We believe that this is a balanced but achievable target. Our most recent quarterly results, we did 49% gross margins.
In our most recent guide, we guided 50 basis points up from that level. Some elements of those things were temporary in nature. Today, we've driven very substantial improvement over the past three quarters. We believe that we can make those things durable and long-lasting as we move forward. You can see the sort of path that we've been able to generate. Now, it's going to get harder as we move forward, right? We've been able to pick some low-hanging fruit and make rapid improvements to date. As we continue to move forward and focus on these areas of excellence and innovation, those things just naturally get harder. We believe that we have a good potential to move that needle forward and a good line of sight to the drivers for that value.
As we move further down the income statement, you can see we've got the growth. We're gonna generate significant dollars from gross profit by improving gross margins. Then we're gonna make sure that we generate leverage all the way down the P&L, including operating expenses. We're gonna do that by ensuring that we maintain discipline around our spending across OpEx lines, we're gonna do that while we still continue to invest dollars where we should, where we need to. R&D is very important for us as a company to maintain our competitive advantage. You heard about the product life cycle and about how our engagement with those customers from lab-to-fab is very important for our positioning, for our learnings, and for our customers as well, and driving their learnings.
It's very important for us to continue to spend money in R&D. By generating higher profit levels, we're able to really fund that sort of investment that's required for us to continue to be the number one supplier in our served markets. That leverage continues down the P&L to EPS, moving us from $1.27 in the baseline year of 2025 to $5 a share. That's about a 30% annual growth rate in EPS. You can see the sort of leverage that we're getting across the P&L here. The earnings power that we're generating, we're gonna invest that in driving the company forward. One of the primary ways that we're gonna do that is by focusing in the short term on the expansion of Farmers Branch. You heard how that's fundamentally driving these pillars of gross margin expansion.
You heard how our customers are really pushing us to move faster, and there's a lot of demand that we see today out there. In order to serve that, Farmers Branch is probably the single most important short-term priority for us to really generate value and accretion in the business. As we look to the next year and a half. It provides capacity we need today. It also provides a platform on which we can build capacity in the future as needed. As Missy said, making sure that we match the capacity that we have to the demand that's out there. In addition, Farmers Branch helps us on our gross margin roadmap. It makes us more efficient, drives down our unit costs. As you might imagine, in Texas, a little bit cheaper to manufacture there than in our existing footprint today in California.
As we expand into that site, that lowers our average cost to produce. We'll continue to invest in M&A to the extent that it makes sense, and that's always been our strategy. You saw recently we just did an acquisition of Keystone Photonics. Very strategic type of acquisition, enabling an important technology for us on our roadmap for co-packaged optics. That's the sort of acquisitions that you can expect from us as we move forward. We'll continue to engage in a share repurchase program. The main purpose of that program is to buy back shares and offset the dilution from our stock compensation programs. As you can see by the graphic on the right-hand side of this slide, we've been successful in doing that.
From the time of our last event to investor day that we had back in 2020 to today, we've driven down diluted share count over that same time period through this program. Long term, we expect to spend around 4%-5% of revenues in sustaining capital, you know, just maintenance-type activities. Our target model is ambitious. I think it's achievable, and I think importantly, it's self-funding. There's a lot of self-help elements in here. We're driving significant leverage across the business, and that's enabling us to do in things like Farmers Branch, to invest in R&D, to really sustain and move the business forward. I think it's structural, what we see in the markets. I think that we're driving organic growth that's really meaningful here and driving leverage across the business.
Thank you for joining me today as we share what I think is a very compelling vision of the future for FormFactor and the future for our customers. With that, I'll hand it off to Stan.
As always, we have some supplemental materials. I'm not gonna spend any time on these slides, but I want you to know we have it. We're gonna take about five minutes quick break, so we can set all of the chairs on the stage, and then we will move to Q&A. Take five, and let's reconvene at We'll do it at 12:00 P.M. Noon, I'll see you guys back.
[Break]
Nope. Okay, here it comes. All right, why don't we all get seated? We'll have our executive management team here on the stage getting ready for Q&A. We have eight. We have an hour. We will have Crystal and Kelly handling mics. Please raise your hand when you have a question. We can see you, Craig. You'll get your mic in a second. Wow, a lot of you. Okay.
Okay. Mm-hmm
Please keep your questions to two or so we can have broader participation. I will hand it off to Craig with Kelly giving the mic.
One more?
Are we on?
Oh.
Just logistically, I'll serve as emcee and traffic cop. Some of the questions I assume will be directed at me, but I can also disposition them to our team.
Okay. Yeah. Craig Ellis, B. Riley Securities team. Thanks for the very informative session and impressive target model. I had a near-term question and a long-term question. The near-term question is, we can see the demand is very, very strong, and it seems that we've been able to push up effective capacity recently with ops improvements. How confident are we we can continue to do that in the next year before Farmers Branch ramps up? Does it mean that some of those ops gains that you showed us are really kind of pulling in the near term versus more back-end loaded?
Yeah. Well, I'll hand this one to Missy, but I think it's important to note, remember on that timeline, we're ramping Farmers Branch in less than a year, right? We're making excellent progress on that. Missy, do you want to talk a little bit about the sustainable improvements we're making in our existing footprint?
Yeah. Everything I showed today, the charts on the second slide really stem from our Livermore operations, which is where our MEMS factory is. We continue to drive improvements there. Mike alluded to it in our last earnings call, you know, squeezing more out of the juice. Today we have several activities, so we're still focused on yield, cycle time reduction. As we're able to improve in those spaces, we can pull in more demand, right? As you improve your yield, you can start less wafers to satisfy one customer, which frees up wafers for another. I'm really confident in the projects that we have lined out and the continuous improvement activities that we have.
Great. The follow-up question is a long-term extension of that. As we look at the target model's 2030 timeframe, Mike, how do we think about the linearity of getting there with the contribution from probes versus systems? How below the top line do the COGS gains and the gross margin benefits pace on the way to 2030? Thanks.
I'll take the question around the top line linearity, if you will, and some of the mix pieces, and then hand off to Aric for as we work our way down the income statement. We expect continued growth in these segments. I don't expect it to be linear, right, on the way to 2030 over that timeframe. This has always been a cyclical industry. We operate with the principle that it always will be a cyclical industry. The overall demand drivers and where the industry is with a set of capacity constraints would seem to indicate we're gonna be in an undersupply condition for a while. There'll be a few ups and downs on the way to 2030.
That's why this idea of modular capacity is so important, so that we can preserve the gross margin and not have a whole bunch of underutilization and large fixed costs flowing through the P&L.
Yeah, I think Mike covered the majority. The only thing I'll add is that if you kind of reflect back on the slide, I had some pie charts on the side, I mentioned that mix is expected to be relatively consistent as we go through this, the next several years. What that implies that they're effectively growing over this timeframe at similar rates, right? Same proportion of business for our systems segment as well as our probes segment. Another element that I think I glossed over a little bit was the gross margin profile. We're starting to see convergence in gross margin profiles between our probes segment and our systems segment. In fact, if you look back at this last quarter, we exceeded 50% gross margin in the probe card business.
We expect that trajectory to, you know, continue in that direction as we drive these fundamental underlying cost improvements and drive scale, and we expect to see further convergence between the two segments.
Krish.
Yeah, hi. Thanks for taking my question. It's Krish Sankar from TD Cowen. Thanks for doing a very informative presentation. I have two of them. First one on the custom ASIC moving to advanced MEMS. Would that still be a kinda like how it is in GPU probe cards, two-person race? Do you think the pseudo MEMS competitor can actually also do custom ASIC advanced MEMS? I have a follow-up.
Yeah. Sudhakar, you wanna take that?
Yeah. Thanks, Krish, for the question. As I mentioned, what we are starting to see in the custom ASIC is it's starting to adopt more of the leading-edge transistor nodes. We're coming down to 3 nm and 2 nm, which is essentially with higher density in cores, it's driving test complexity first, which is translating to thermal. We strongly believe that in the long term, this is gonna be an advanced MEMS play with a two-supplier market in the long term for the custom ASIC as well.
Got it. Got it. Then a quick follow-up on the silicon photonics. The, the Triton platform, it seems very interesting. Do you think eventually, like, CPO would eventually go kind of the semis route where they unbundle probe cards away from testers? Do you think this is the path forward in CPO testing where it's an integrated solution? Thank you.
Why don't I take that one? 'Cause I think it's an industry supplier ecosystem question. Why are we partnered so closely with Advantest and Tokyo Electron on it? It's because of speed, right? You know, we've had some major customers over the past couple of years tell us they need systems, they need them to ramp in high volume. It's difficult to work with everybody in an open ecosystem now. You focus, right? Longer term, I expect CPO, and we expect CPO as a team, to broaden out and have a very similar disaggregated model where all ATE systems work with all interfaces, whether they're probe cards or something slightly different, work with all handlers. The time to market is so critical right now to meet this initial CPO ramp that it demands some focus. David?
Thanks for taking my question. I guess the first question is, you showed us a lots of great charts today. One thing I noticed is that you plan to double your revenue, but your SAM is only gonna grow by 50% between now and 2030. You showed us another chart where it had like five or six little vectors up. I'd like to know, or if you could help us, which one of those vectors are gonna provide the share gains that you're talking about to grow twice as fast as the market?
Yeah, I think I'll ask Aric to consolidate this one because there's different pieces, right, across the revenue landscape for us.
I think the share gains, where we're going to principally see share gains and where there's existing market today where we don't play as big as we think we should and we're entitled to, is places like CPU, some of the major CPU manufacturers out there where we have low to nonexistent share, we expect to penetrate there. GPUs, we didn't participate historically in GPU tests. We expect that to change. We've indicated the second half of this year, we expect to see demand ramping for GPUs for us with the major GPU producers in the world. Those are the main vectors that are going to drive market share. We've got new growing markets like CPO, that's really something that doesn't exist today that's going to drive growth as well.
Okay, as my follow-up, I think you said the CAGR of semiconductors or your market is like 8%. I'm kinda curious, you know, that's kinda like the long-term growth rate of semiconductor unit volumes. You know, at least semi revenue grew like 25% last year. I think it's growing like over 60% this year. I'm guessing it's gonna grow by more than 20% next year. I know there's a lot of increase in pricing in memory and GPUs, but units are definitely growing faster than 8%, so I'm kinda wondering what the disconnect is.
Yeah, the market figures that I share, we in great part are based on third-party analyst reports, and remember that they are focused on the advanced probe card market. It's not entire semiconductor spend, it's not capital spending. It's specific to the markets that we serve. The growth rates on average in that market in those industry reports is at an 8%. We expect to be participating in elements of that that are growing much faster, right? At a rate that's double that.
Okay. Charles?
Thanks, Mike. Charles Shi, Needham. Mike, I'm wondering if you can address what's the expected TAM for CPO testing insertion one, two, three, four. You showed the four re-insertions. Tell us a little bit more, expand a little bit more, 'cause I think you said one sentence why you wanna focus on insertion one. A related question, what's the expected ramp timing? I think on the earnings call you were talking more like 2028. I know it's hard to get an update like just two weeks later.
Yeah.
What do you think about the ramp timing?
Let me connect those dots, and then I'm gonna hand it over to Jens for some of the details on the insertions, and the relative size and how we expect that to potentially evolve. We updated you on the call a couple of weeks ago. We'd initially set expectations for our CPO revenue in this year, 2026, to be between $10 million-$20 million. On the last call, we updated that to the high end of that range. And that's indicative of the momentum we're seeing and, depending on how you wanna classify production, the beginning of that production ramp. We've also updated you that we've now installed multiple Triton systems, right, in one advanced foundry. That, for me, is always the indication that things are starting to tip towards production.
The narrative from our customers continues to be very strong and even accelerating.
Yeah, we are expecting a SAM by 2030 around about $300 million-$400 million from our view today for co-packaged optics. Why we are on insertion one, I think there are two things. First, I explained how the known good die strategy favors very early testing, and we are very early in co-packaged optics, especially going in high-volume manufacturing. Yes, there are other insertions, but insertion one is the most important one today to make this ramp happen, and we are involved in this insertion since over nine years. We developed the right strategy and technology there. As I said, this deployed technology, which we have in insertion one, have the capability then later on to address also other insertions when high-volume manufacturing for co-packaged optics gets more mature.
The $300 million, $400 million, that is all insertions, including equipment and test consumables? I just want to clarify what's in that.
That's the wafer-level test SAM we are addressing.
Which means insertion one too?
Right.
Okay, thank you. maybe a second question on gross margin. I know this is a question probably for Aric, maybe for Missy as well. you know, going back to the history, gross margin in the fifties for FormFactor, we've never seen this kind of gross margin. this is a very aspirational target. I know you laid out how you wanna get there.
We noticed that the part of the gross margin gain actually comes from productivity improvement, cost discipline, efficiency. Historically, this is a semiconductor industry. Very often suppliers getting those efficiency improvement, very often those value got given away to the customers. How do you make sure you don't give away the value created by those activities to customers, so you can keep the margin to yourself and maybe get to the 55% gross margin by the end of this decade? Thank you.
I think that's a really good question. It boils down to discipline and really delivering value to our customers. Where we deliver value to our customers and solve important problems for them, like how are they gonna yield in these fast ramps, the getting the production slots that they need, that's where we have the ability to really demonstrate value and differentiate ourselves. That's what's gonna drive our ability to really protect our gross margins, if you will. It's why it's so important why you see us talking here about what are we doing from a technology perspective. Why are we continuing to invest in R&D? Over the long term, if you don't do those things, it's going to be very difficult to preserve your gross margins.
Matt?
Hey, guys. Matt Prisco from Cantor. First up, just wanna hit kind of the target model again, that 2030 number that you put out there. Is that anchored in any type of semiconductor revenue or anything like that? 'Cause I know in the past you guys have talked about having that intensity.
Right.
You talked today about intensity going higher. I ask about that anchor because, you know, every week it seems like, you know, forecasts are going higher and higher with AI and everything.
Yeah.
As we start to think about that number shifting higher, how does that translate to revenues for you?
I'll hand this one to you in a second, but it came up in an earlier question, I think maybe from David. One of the reasons the overall industry is growing so rapidly from a dollar basis is ASP growth, right? If you looked at two of our major DRAM customers' recent earnings, they had in DRAM 0% bit growth and north of 50% ASP growth. Our business is really unit driven, units on new designs. There are opportunities for us to get some compensation for pulling things in in this environment, securing capacity slots. I'm sure you're gonna see a decoupling between the classic probe card intensity on a dollar basis just because of these inflated ASPs.
Yeah, maybe we'll add to this. I think it's fair to, for us to characterize ourselves as a management team as being pragmatic. There's a lot of, you know, bubbliness out there in the market right now, a lot of excitement around some of these trends. I think we're much more focused on what's actually happening and making sure that we bring up capacity when we need it, that's aligned to the demand that's really out there and really keying in on what are those drivers, what are those indicators, and making sure that we've got a flexible operating model that enables that. I think Farmers Branch is a really good example of how we're doing that.
It serves as a platform that not only addresses the immediate capacity needs that we have today, actually serves as an ability to drive the ability to drive additional capacity in the future.
That's actually a perfect segue. My next question is on Farmers Branch. I just love to know how you guys are thinking about that modular build-out today. Is that gonna be kind of taking deposits from customers, getting capacity corridors? At what point do you guys feel comfortable building out that capacity? Maybe how long does it take you to build this capacity once you see the demand, and when do you think it's fully up and running? Thanks.
Yeah. I'll take that one. We plan to have the site qualified later this year with initial production ramp starting in January. You know, this industry is slightly different than typical device makers. We get from a planning process, a real look at what our customers want. Sudhakar spent some time today talking about custom, right? Each of our probe cards are custom to what the customer's ramp is. We have about a six-month window where we can really understand exactly which products they're going to want us to run. That also happens to coincide with around our longest lead times. Through our SIOP and planning process, we pay close attention. We also get early indicators from the close partnerships that we have with our customers.
As we get past the initial qualification and ramp, we're already looking forward to the next version, and we'll just do it in modular phases.
Yeah. I think one of the interesting things about the current situation is, although our lead times, the POs customers place for a specific design are still well within a quarter. We've got much more engaged conversations on what their six-month forecast is, right? They understand there's a capacity pinch all through their supply chains, not just in probe cards, right? You know, whether you talked about potentially capacity agreements, paying for capacity, all of those different things are under discussion. I would argue that we're getting much better engagement, if not visibility, on what customers think they need from us more than a quarter out now, in terms of raw capacity, you know, probes per week kind of thing. They still don't know what they want us to build, but they know they're gonna need some of it.
Maybe just to layer on one component, the ability to add capacity now is no longer bringing up a full site.
Yeah.
It's bringing on just additional equipment, additional tools. Our ability to do that with much shorter lead times in the future, if necessary, is what I'm referring to when I refer to it as a platform.
Okay.
Elizabeth?
Hi. Thanks for taking my question. Elizabeth from Citi. I'm wondering when you think about your R&D budgets, how do you think about, like, the priorities and, like, how do you allocate your R&D resources across all these areas?
I'll let Aric take that one.
It will go back to the discipline part of the equation. I mentioned driving efficiency in OpEx, continuing to invest in R&D. We recognize that R&D is one of the ways that we can drive significant sustained value with our customers and in our products. It's important that we make that investment. It needs to be made in the right places. Making sure that we have the business processes, the decision-making upfront as we are looking at those decisions, probably more importantly, in some cases, really narrowing the aperture a little bit and focusing.
Yeah.
You know, you try to do too many things at once, you're not gonna get anything done. We need to make sure that we maintain discipline and how many things we try to do at once. We've got the feelers out there with our customers to understand what really matters, and that's how we determine what we focus on.
Got it. Then on the CPO test part, I'm wondering if you could share your thoughts on the involvement of the four insertions, the mix between the four going forward as things move, you know, like, from initial, like, ramp to more mature stage of the CPO testing market, and also, like, the shape of the SAM growth into the next few years, because, like, one of your test peers talk about, you know, there's a fast ramp at beginning of the SAM, and then maybe it slows down because things get more mature.
Right
thoughts on that as well.
I'll hand it off to Jens in a minute, but I think it's really important we know this is gonna change, the dollars spent between the different insertions. Right now, for very good reasons that Jens shared, there's a lot of the overall test spending being spent on insertion one. We feel like that's a good place to focus because the foundational technology of optical probing is gonna need to be used in any of the other insertions, right? This is gonna change. It's going to involve. It's why this customer intimacy that Aasutosh and I talked about is so important, right? We're in there at the beginning. Customers are depending on us. They don't want to surprise us, right?
There's very active engagement on not just with customers, but other co-suppliers on making sure that we're gonna be ready whichever way this thing shifts as it gets to HVM.
Yeah, there's not really much to add.
Short. We're aligned in our strategic views.
No, no. I think as I explained, I think the insertion 1 is already pretty complex, it's getting even more complex if you're getting to the co-packaged optics technology, for example, as TSMC is providing it when you have double-siding tests. These systems are getting so complex that they are just not economically today. We will see in the future, as Mike said, how this will develop over the different insertions, which insertions will be really necessary in the end for high-volume manufacturing. Right now we're right at the beginning and right in place for insertion one.
Thank you.
Before we let Craig go again.
Vi. Oh, Vi, you've got one?
Yeah.
Okay.
The first question I had. Oh, sorry. Thanks for taking my question. This is Vi from Evercore. The first question I had is, in your SAM growth opportunities, I didn't see CPUs. Maybe some thoughts on what you're thinking, for the CPU process, probe card intensity.
Yeah, it's a really good question. It's in that base business bucket. I think what you're probably referring to is some of the news cycle out there around, you know, agentic AI driving additional CPU intensity. We are seeing some, you know, some acceleration there, some interest there. To the extent that occurs, we have relationships in that space that I think you're aware have been traditionally quite strong, and I think we're well-positioned to serve that market to the extent that we see that grow as a result.
Would that be further upside to your kind of revenue growth?
I think it remains to be seen. I would say that overall, as we've discussed, growing at 2x the market rates, I think we've got a balanced plan. We've got multiple growth vectors. I think what you're gonna see is that reality is gonna be different. You know, some of those vectors will be different as we move forward in time. This may be one of those areas. I think overall, where we're targeting in terms of total growth and that rate is right where we really should be in terms of balance.
If I may one, on gross margins, you've laid out clear path for cost discipline and the levers you can pull, but probe cards are getting more and more advanced. Does that imply an ASP increase with, kind of a cadence for it?
I kinda answered this a little bit earlier, you know, to reiterate, I think where we see ASP is kinda tricky, right? What I really care about is, you know, value and profitability, that for us is driven by the same things that drive value for our customers. To the extent that we deliver value to them, we help them solve problems that avoid the yield. You heard Mike talk about the yield losses that happen just because of packaging these things into advanced complex packages. Those have significant dollars tied to them. To the extent that we can solve those problems at a more efficient, in a more efficient way for our customers, that's gonna drive value in our products, will drive the associated margins and ASPs.
Yeah. David, did you have a question?
Yeah.
Before we go back to Craig?
David Silver, Freedom Capital Markets. I had a question maybe about the target model and the breakdown between logic and foundry and memory. When the 2020 model was created, you probably had a certain breakdown of where the revenues were going to come from. My guess is it changed a little bit over time, major share, major customers, et cetera. Could you just kind of, you know, single point us, but in 2030 from 2025, what is the share regrowth from the memory side versus foundry and logic? I don't know, is that even the right way to look at it since there's going to be hybrid products that incorporate both sides? We're going to have multiple sources, I guess.
Just some comments on how much you're however you look at it, revenue or operating income, how much from the memory side, how much from foundry and logic?
Yeah. From a top line perspective, we expect the mix to be pretty similar today. I think I showed a couple of pie charts on this slide. Just a little bit better growth in foundry and logic than, say, in DRAM. You're hitting on a bit of a sore point with our last target model that I just wanna spend a little bit of time on. When we set our target model in 2020, it was premised on a certain mix of foundry and logic and memory products, and at that time there was a pretty significant difference in the, in the margin profile between those two sub-businesses, if you will. We saw the mix shift, and it was not favorable to our gross margins, and that's really been a pain point in us really achieving that target model.
It's one of the reasons why we're so focused on driving fundamental underlying cost and efficiency improvements because if we do that right, if we drive the cycle time and yields and things that Missy was alluding to, it's going to benefit all of our products, right? It doesn't matter if we're talking about DRAM. We don't wanna be the victims of mix. We don't wanna be the victims of volume. We wanna be a company that produces competitive margins, competitive fall through, if you will, from an operating income perspective. Whatever the outside circumstances are, we'll control what we can control, and we're gonna do that as best as we can by exercising discipline and making sure we get the appropriate leverage.
Maybe just a question or maybe an observation you can comment on, but this would be regarding talent acquisition, okay? I always think where your major customers are located, your customers are national champions, and they probably may have an advantage in acquiring, you know, the intellectual talent that they want. You need to service them. You need to maintain a certain standard. Just maybe some comments about your confidence in kinda keeping up with the technical demands and the talent demands that your major customers are gonna require if you're gonna hit your next target model?
I think I'm gonna ask Missy to comment on how recruiting's going in Farmers Branch 'cause it's basically a new area for us. Broadening out globally, are there challenges with talent? Sure, right. You know, not a lot of people going into hardware engineering, material science, physics, the kind of things that underpin this business. So far we've been reasonably successful in recruiting the kind of people we need 'cause they like the idea of If you look at the core constituents of the products we make, they're a really unique blend of mechanical engineering, electrical engineering, material science, physics. You don't get that in a lot of different places in the industry. It's one of the reasons I love working here, right. It, you learn something new technically every day.
I think the people we've been able to recruit have that same passion for continued learning and learning in areas that are maybe a little bit different than mainstream semiconductor, while also having the impact on a fast-growing smaller company. They can see their daily impact on what we do.
Yeah. With regards to Farmers Branch, you know, it's located in North Texas with access to multiple universities right there in the area, as well as several different community colleges. We're working with the North Texas Semiconductor Institute as well, along with some of the other major semiconductor companies within the ecosystem. To date, we're on track with our hiring plans. We've had no issues recruiting great talent. We are expanding that out. There's some high schools that we're considering working with. North Texas has really spent a lot of time over the last three years focusing on how do we bring in technicians. We all know that that's an area in our industry where there's gaps right now, is getting, you know, two-year degrees. There's a real push there in North Texas to go drive it.
I'm really excited. That was one of the reasons why we selected this site specifically, because we felt confident we could get the talent.
Other questions before we let Craig have a second swing?
Jason Ursaner at Bumbershoot Holdings. I thought you guys did an amazing job of kind of explaining, you know, your position with your customers in memory, DRAM, why everything is, I guess just set to, you know, have a great outlook for a long time. Could you maybe explain, you know, the memory customers in the LLMs, what's happening with some of the compression algorithms and relational database? It's beyond my understanding, but is there focus to kinda use.
Fewer DRAM over time, use less memory and kinda the whole cache system of AI, and how does that kinda impact some of this? Is it real? Is it not? You know.
Yeah. It's also an area that I don't, and I don't think any of us know much about. What I will say is it's quite clear, given the dynamics, and I'm gonna restrict things down to DRAM. Most of you know that we're not a relevant player in flash, and we can get into that if you want. It's really focusing on the opportunities where we can drive real value and differentiation. They're available in DRAM and HBM, of course, not in flash. I'm gonna restrict comments to that.
When you look at, even with some of these compression algorithms and different things, the growth rates for compute and different elements of memory, whether they're HBM or the explosion we've seen in DDR5 for these applications, it's hard to invent a scenario where some sort of compression algorithm or compute innovation takes that off the trajectory of us being essentially under-supplying the market. Us as an industry, right? Under-supplying the market. That's consistent with the narratives we hear from our customers. All the different pieces of that high-performance computing module we used as a consistent thread through the story, it's more memory, more memory, more memory. I think over the years, you know, the insatiable demand for compute, it's tough to figure how some sort of algorithm development is gonna flip that script. Okay. Craig Ellis.
All right, thank you. I wanted to go back to the point we made about the ambition to capture 50% incremental SAM share and the points we made about future-proofing our DRAM and foundry logic probe card products. The question is, what do we need to do between now and 2030 with each of those products to ensure that we've got product leadership and are in a position to capture that 50% incremental share?
As I alluded to in both segments, both foundry and logic, as well as DRAM, HBM, the major vectors driving us is test intensity and complexity. I'll stay on the complexity side for now to really answer your question. With the increasing thermal demands that's happening, we're starting to see that the current set of products are not able to keep up with the next generation of GPU that's coming out next. We've already closely working, as Aasutosh talked about, co-development. We're already working closely with our customers, leading-edge foundries and IDMs on how do we enable them to thermally manage and dissipate these heat that is getting generated because of the power. That's already underway, so to speak.
That's part of the things is baked into the overall target model is getting those kind of design wins into the future and how that is growing is already baked in, particularly on the foundry and logic. On the HBM and DRAM side, a similar set of trend is going on on the power side, but also this custom HBM-based die that I talked about is starting to drive new architecture development that may be coming to fruition more in the 2028 and beyond timeframe. That's already baked into it, and we're starting to work with all major customers in that space as well.
Got it. Mike, because in talking about capital allocation, the company left the door open to M&A, and my sense was that was more tuck-in than something big. Can you help us understand where that type of thing might make sense for FormFactor that you see between here and 2030? Thanks.
Yeah. Well, M&A, as you saw, right, has been a fundamental tool we've used to build the current FormFactor, and we remain very active in prospecting for different things. It is more stilted towards the tuck-in. Keystone's a pretty good example of where we're putting our focus on priority. A key enabling subcomponent that drives the themes you saw today around high-performance computing and advanced packaging forward. Over this timeframe, that's where we expect to focus. I'll give you my standard speech. I still believe this part of the overall semiconductor equipment and consumable space needs to consolidate as a supply chain, right? If you look at what we're taking on as a leader in a small subpiece of the industry, there are challenges, right?
If you look at what happened to the front end two decades ago, right, there was consolidation, and now you have arguably five very capable large suppliers that drive the front-end wafer fab forward. I really think, right, the same thing needs to happen longer term in the back end, in test and assembly, because of the technical requirements associated with advanced packaging and the customers we're engaged with and the speed they want to move.
Thank you.
Hey, guys, Matthew Prisco again. Just thinking about the share side again, you talked about the opportunity in CPU, GPU, then kind of that end driving the share gains. As we think about memory and DRAM, HBM in particular, I guess, what gets you most excited across the three insertions, whether you're talking that core die, the base die, or the final test? As we think about mix of that market changing, as more suppliers come to the market and maybe your primary customers', you know, dominance, it's more spread out over time, are your gain opportunities in the other areas enough to kind of maybe offset that mix dynamic so you can gain share within the memory market?
Can you talk to this?
When it comes to the HBM and DRAM, again, going back, I feel like I'm going to repeat myself a little bit to answer your question as well. Going back to the test co-complexity and intensity piece of it's really in the longer term what's exciting about this space is the evolution of not only the stack height, as we described about increasing number of probe cards needed to measure the increasing number of stack heights, but kind of on HBM5 onwards, where we think this hybrid bonding is also starting to happen and we're starting to watch. We don't know exactly how that's gonna create new test insertions for us.
What we do know is with the custom logic that's coming in from the right and shifting into the base die, we are starting to see we need to develop new architectures, which is kind of exciting because that's another new test insertions that's evolving, that we are seeing that, hey, we will start to gain share and with new differentiated offerings at the same time. As I mentioned, we're the only probe card supplier with production-proven foundry and logic probe architecture, as well as the DRAM HBM probe architecture. We'll be able to kind of fuse those things together as we understand the requirements quite well.
David?
Yep. Thanks for taking my second question. The first one is, in Q1 of next year, what do you think the gross margin impact will be from bringing on Texas? What I mean by that is you're gonna obviously hit the depreciation button at one point or another. I imagine Q1 would be the quarter where you see the most increase in depreciation, but maybe I'm wrong. That's one, the first part of the question or first question. Second question is, being somewhat capacity constrained in the second half of 2026, do you think you've lost market share or are the competitors in the same situation that you are?
Yeah, I'll answer the second one first. Everybody's in the same constrained situation. That's one of the reasons why, one, as I mentioned, we're getting better. I'll call it, you know, forecast-y guidance from our customers, is the overall probe card supply market is constrained, like many other places in the industry are. This is one of the reasons why we're executing with urgency, as you, I'm sure felt from, Missy's presentation on getting Farmers Branch up and running, so that we're not in a share loss situation, but even arguably could be on our front foot here and able to gain share based on availability. You wanna then touch on the first one?
Maybe, just to tag on to that a little bit. You know, the levers that we're using today to drive really output up are the same levers that we're using to drive gross margin improvement in the short term and also over the long term. Cycle times and yields, our ability to produce more good product with the same amount of inputs and do it faster means that we're generating more revenue intensity and more profit intensity in the same unit time, if you will. That's what's gonna enable us to continue and we're gonna continue to push on those levers through the back half of the year.
As it relates to gross margins for Farmers Branch ramp, we believe that during the ramp, there are some headwinds that will, but we're doing a fast ramp. Ultimately, that site's gonna be accretive to gross margins. It's part of our gross margin roadmap, as I articulated. As you look at those operational excellence and the innovation column, Farmers Branch is a key enabler of those two things. In that short period of ramp, about a year that we're talking about, we expect any headwinds associated with gross margins to be offset by the efficiencies that we're driving over the next several quarters.
Any more? Elizabeth again? Oh, Vi again.
Thanks for taking my question again. The one on gross margin improvements, in historically, you had a huge variation in gross margin and you kind of talked about it. It was product mix and volumes kind of both driving it hand in hand. As you kind of stress test this 2030 model, what kind of variation do you see now by product volume? Will it be 5 percentage points, or can it be 2 percentage points down, up?
I think in a normal we do stress test our models, and we, you know, I think as Mike mentioned, we're under no illusion that we continue to operate in a market and a industry that remains cyclical, despite all the optimism that we see. We test our models. I think the way mix and volume will always impact gross margin. To the extent that we drive convergence between the gross margin profiles between our different products and different segments, which we see occurring, the importance of mix it's gonna be less important, exactly what our mix is. We do expect some variation from period to period around mix, and you know, plus or minus a couple 100 basis points would be normal, right?
To the extent we end up with a more extreme scenario, we'd take appropriate measures as required.
Elizabeth?
Sorry, I also have a gross margin question. You laid out from 41% last year to 55% target model. It's a mix of 40% volume, 35% execution, and 25% innovation. Now we are sitting close to 50%, so I'm just wondering from now to 55% in 2030, is there any difference between the mix of these three, and what other things have been done? What other things you can do more?
Yeah. Good question. I think that you can think of it. Those three pillars that I described in the roadmap from 41%, which as a reminder for everyone was 2025 baseline average gross margins for the full year, right? Obviously, we've generated a very different trajectory to date, but elements of that are temporary in nature. We need to make those things permanent. That's one element. If you think about the distribution of the continued improvement from where we are today, if you will, to the 55%, it's similar to what I showed on that table, just, you know, scale is a little bit different.
What we've generated, the 1,100 basis point improvement that we've generated from Q2 of last year to today, if you look at our most recent guide, is also across all three of those pillars.
Right.
Hi, this is Daniella. I'm from Stifel. I wanted to focus on Texas and Farmers Branch. When you begin production there, what is the determining factor for if it'll be more focused on logic or more focused on DRAM HBM?
Sure, I'll take that one. As I mentioned in my presentation earlier, we now have fully fungible manufacturing lines, which really just means that I can run two different technologies through the same tool sets. As we see, again, kind of going back to the previous question around the shift in mix and demand, part of this reasoning was so that as we see that mix move, we don't have as much variation into our margins in turn. For us, we're starting with the foundry and logic, and that's what we'll be ramping first. That frees up some additional capacity in Livermore to also still have foundry and logic as well as the DRAM.
Great. Thank you. To follow up, if you could give sort of a number or a guide for that maximum incremental revenue that you could see from Farmers Branch if we look at it from a 12-month window.
I'll let you take that one, Aric.
I think what's key to understand here, our target model, doubling revenues between now and 2030. The investments that are included in these financials that we've shared, they're all inclusive, right? What you see and what we've presented, it includes the required investment. What we're doing in Farmers Branch will enable that revenue level. It also enables, like I've said, a platform for additional capacity to the extent it's required in the future. It allows us, you know, shorter time to get to that because it's really about equipment sets.
Thank you, and great presentation.
Thank you.
Anyone else? All right. With that, I'll wrap us up then. Stan, you wanna. I've talked a lot. We'll let Stan handle this.
All right. Well, thank you all for joining us today, and for people online, thank you again. I wanna thank our executive management team and our CEO for putting this presentation together. Hopefully you found it informational and educational. We will see you in the upcoming weeks, months, and please follow our story. Thank you all.
Thank you.