Thank you and welcome everyone to FormFactor's Third Quarter 2021 Earnings Conference Call. On today's call are Chief Executive Officer, Mike Slessor, and Chief Financial Officer, Shai Shahar. Before we begin, Jason Cohen, the company's General Counsel, will remind you of some important information.
Thank you. Today, the company will be discussing GAAP P&L results and some important non-GAAP results intended to supplement your understanding of the company's financials. Reconciliations of GAAP to non-GAAP measures and other financial information are available in the press release issued today by the company and on the investor relations section of our website. Today's discussion contains forward-looking statements within the meaning of the federal securities laws. Examples of such forward-looking statements include those with respect to the projections of financial and business performance, future macroeconomic conditions, the benefits of acquisitions and investments in capacity and new technologies, the impacts of the COVID-19 pandemic, our supply chain, the impacts of regulatory changes, the anticipated demand for products, our ability to develop, produce, and sell products, and the assumptions upon which such statements are based.
These statements are subject to known and unknown risks and uncertainties that could cause actual results to differ materially from those expressed during this call. Information on risk factors and uncertainties is contained in our most recent filing on Form 10-K with the SEC for the fiscal year ended 2020 and in our other SEC filings, which are available on the SEC's website at www.sec.gov and in our press release issued today. Forward-looking statements are made as of today, October 27th, 2021, and we assume no obligation to update them. With that, we will now turn the call over to FormFactor CEO, Mike Slesser.
Thanks, Jason, and thank you everyone for joining us today. FormFactor delivered solid results in the third quarter, again recording the second-highest revenue in company history, with gross margins at the high end of our outlook range. Together with sustained operating expense control, these factors resulted in non-GAAP earnings per share above the high end of our outlook range. We continue to benefit from strong demand from FormFactor's diversified set of market-leading semiconductor test and measurement products. With our added production capacity now coming online, we're well-positioned in the current quarter to deliver sequential growth. Like many companies in the semiconductor industry, we're facing a variety of supply chain and labor challenges. Our long-term investments in automation and vertical integration have helped reduce the effect of these headwinds, and we've thus far successfully navigated these issues with minimal impact.
We do expect this situation to persist into next year, and our team remains focused on actively resolving these challenges. Minimizing impact to our supply chain is especially important as our capacity investments start coming online. Tool installations and qualifications are now underway in our new Livermore manufacturing center, which remains on track to deliver initial customer shipments in the current quarter. We're excited to be on the cusp of commissioning this new flagship facility, which will remove a growth constraint we've operated under for almost two years, helping us to reach the $850 million revenue of our target financial model. I'd like to provide some detail behind our sequential improvement in gross margins, as it provides insight into the importance of the design-specific product mix within each of our served markets.
As you can see from the supplemental materials posted on our website, and as Shai will review later, third quarter revenue by segment and market was very similar to the second quarter, yet we delivered 160 basis point improvement in gross margins. The largest single factor in this improvement was a more favorable product mix within specific markets. As we've noted in the past, probe cards are a consumable that is specific to each customer chip design, and the custom nature of each probe card design drives a different manufacturing cost and selling price. For example, two DRAM probe cards configured to test two different customer chip designs can produce significantly different gross margins because of the details of the probe card and tester configuration, like the difference between a one-touchdown configuration versus a two-touchdown configuration.
Consequently, two quarters with similar revenues in each segment and market can, and often do, result in substantially different gross margins, depending on the specific probe cards shipped to produce that revenue. This variability is inherent to FormFactor's broadly diversified design-specific consumables business. Since we serve major applications at all the leading customers in the semiconductor industry, delivering probe cards to test hundreds of different chip designs each month, we expect to continue to see quarterly fluctuations of gross margin around a long-term trend line expanding to the 47% of our target financial model. Turning now to segment and market level details. Foundry and logic probe cards, our largest business, performed at levels in the third quarter comparable to the second quarter, with an expected seasonal reduction in mobile application processor demand offset by an increase in microprocessor demand and sustained strong growth in 5G RF applications.
With 2021's large foundry and logic wafer fab equipment investments now coming online at our customers, we're experiencing increasing demand for leading-edge foundry and logic probe cards. Along with their capacity additions, our customers are adding innovative advanced packaging architectures like EMIB, Foveros, and 3D Fabric to their roadmaps to help offset the slowing of front-end driven Moore's Law. As we've discussed in the past, these chiplet or tile-based integration schemes drive both higher test intensity, which expands the number of probe cards required per wafer out, and test complexity, which widens FormFactor's competitive advantage. With lead times of less than a quarter, short-term visibility remains challenging as always, but the combination of significant customer capacity increases paired with their adoption of advanced packaging creates long-term demand for FormFactor's probe card products.
Demand for DRAM probe cards in the third quarter sustained near the high levels of the second quarter, and we expect comparable strength in the current quarter. As in foundry logic, we experienced demand shifts between different customers and different chip designs across multiple DDR4 and DDR5 designs in both mobile and PC server applications. FormFactor's ability to absorb these short-term fluctuations in demand from any one customer is a result of our long-term initiative to be a diversified market leader supplying all major DRAM manufacturers and remains a key tenet of our operational strategy in all of our served markets. Our engineering systems business also delivered good results in the third quarter, with gross margins at 50% for the first time since early 2020.
A highlight of the quarter was the shipment and installation of multiple FRT high-performance multi-sensor 300-millimeter optical metrology tools to a single foundry customer for advanced packaging applications. Advanced packaging is a key driver for all of FormFactor's businesses as our customers begin to adopt chiplet-style integration strategies that generate new test and measurement requirements. The FRT tool's ability to quickly, accurately, and non-destructively measure and control critical dimensions, film thicknesses, and surface topologies on chiplet-to-chiplet interfaces is key to improving and maintaining advanced packaging assembly yields. Let me close by noting that with an accelerating demand outlook, increased capacity to meet that demand, and good execution on both short and long-term gross margin improvements, we're well on the path to the target financial model we unveiled last year that delivers $2 of non-GAAP earnings per share on $850 million of revenue.
Test and measurement is clearly gaining importance in the semiconductor industry, driven by powerful trends including 5G, advanced packaging, and memory content growth. Our leadership position in these attractive markets, paired with our differentiated strategy and disciplined execution, will drive continued growth and share gains as we progress towards our target model. Shai, over to you.
Thank you, Mike, and good afternoon. As you saw in our press release, and as Mike mentioned, third quarter revenues and non-GAAP gross margin were at the high end of our outlook range, with revenues again reaching the second-highest level in company history and non-GAAP EPS exceeded the high end of our outlook range. FormFactor's third quarter revenues were $190 million, a 1% sequential increase from Q2, and an increase of 6.7% year-over-year. Probe card segment revenues were $155 million in the third quarter, a small increase of $1.2 million or 0.8% from Q2. As Mike mentioned, third quarter revenue by segment and market was very similar to the second quarter.
System segment revenues was $35.1 million in Q3, an increase of $0.7 million or 2% from the second quarter, mainly as a result of higher sales of optical metrology and thermal systems, driven by advanced packaging and automotive applications. Within the probe card segment, foundry and logic revenues increased by $1 million from Q2 to $105 million in the third quarter, comprising 55% of total company revenues, same as in the second quarter. DRAM revenues were $40 million in Q3, a decrease of $2 million or 5.5% from the second quarter, and were 21% of total quarterly revenues as compared to 22% of revenue in the second quarter.
Flash revenues of $10 million in Q3 were $2.5 million higher than in the second quarter and were 6% of total revenues in Q3, up from 4% in Q2. As we have said in the past, we expect Flash revenues to be lumpy from quarter to quarter. GAAP gross margin for the third quarter was 42.2% of revenues as compared to 40.6% in Q2. Cost of revenues included $7.2 million of GAAP to non-GAAP reconciling items, which we outline in our press release issued today and in the reconciliation table available in the investor relations section of our website.
On a non-GAAP basis, gross margin for the third quarter was 46% at the high end of our outlook range and 160 basis points higher than the 44.4% non-GAAP gross margin in Q2. As mentioned, third quarter revenue by segment and market was very similar to the second quarter and most of the improvement in gross margins, approximately two-thirds of it, as a result of a more favorable product mix within the specific markets. Our probe card segment gross margin was 45.1% in the third quarter, an increase of 180 basis points compared to 43.3% in Q2. The increase is mainly due to the more favorable mix I just mentioned.
Our Q3 systems segment gross margin was 50%, 90 basis points higher than in the second quarter. As we have said previously, we expect our systems segment gross margin to range between the high 40s-low 50s. Gross margin remains an area of focus for us, and we are encouraged by our progress towards our target financial model gross margin of 47%. As a reminder, we expect that margins will fluctuate from quarter to quarter, mainly as a result of changes in product mix. Our GAAP operating expenses were $57 million for the third quarter, $1 million higher than in the second quarter. Non-GAAP operating expenses for the third quarter were $48.5 million, or 25.5% of revenues, essentially flat with $48.4 million or 25.7% of revenues in Q2.
Company non-cash expenses for the third quarter included $7.9 million for stock-based compensation, $2.5 million for the amortization of acquisition-related intangibles, and depreciation of $6.6 million. Amortization for acquisition-related intangibles decreased from $7.1 million in the second quarter as certain assets became fully amortized during Q2. The increase of depreciation from $6.6 million in Q2 reflects additional assets placed in service as part of our capacity expansions. As disclosed in a Form 8-K filed last month, we adopted restructuring plans during the third quarter to streamline and improve the efficiency and business effectiveness of our operations. We are consolidating certain manufacturing facilities, including moving operations into our new Livermore manufacturing center that will begin production in the current quarter. We expect these actions will be largely completed by the end of 2020.
We estimate that these actions will reduce our cost structure by approximately $3-$4 million on an annualized basis once the actions are fully implemented. Non-GAAP operating income for the third quarter was $39 million, $3.8 million higher than in the second quarter. GAAP net income for the third quarter was $20.5 million, or $0.26 per fully diluted share, compared to $17.9 million or $0.23 per fully diluted share in Q2. The non-GAAP effective tax rate for the third quarter was 18.9%, practically the same as the 18.7% in Q2, and within our anticipated non-GAAP effective tax rate for fiscal 2021 of 15%-20%.
As a reminder, our annual cash tax rate is expected to remain at 6%-8% of non-GAAP pre-tax income until we fully utilize our remaining U.S.-based R&D credits. Third quarter non-GAAP net income was $31.6 million, or $0.40 per fully diluted share, compared to $28.4 million or $0.36 per fully diluted share in Q2. Moving to the balance sheet and to cash flows. We generated $14 million of free cash flow in the third quarter compared to $16 million in Q2, and we have had total cash and investments of $268 million at the end of the quarter. The third quarter's $2 million sequential decrease in free cash flow reflects an increase in capital expenditures.
As of the end of the third quarter, we had two term loans remaining on our balance sheet, totaling $27 million. We invested $20 million in capital expenditures during the third quarter, compared to $18 million in Q2. This brings our year-to-date CapEx to $51 million, chiefly related to the capacity expansion in our Livermore manufacturing center. We continue to make significant investments in capacity in 2021, and with most of the year behind us, we are fine-tuning the range for forecasted cash CapEx for the full year to $70 million-$80 million from the $70 million-$90 million previously communicated. As a reminder, we expect CapEx to return to 3.5%-4% of revenues in our target financial model after we conclude these capacity expansions.
At quarter end, our total cash and investments balance exceeded the debt balance by $241 million, an increase of $11 million from Q2 quarter end. No significant share repurchases were made during the third quarter, and year to date we purchased 620,000 shares under our two-year $50 million share repurchase plan. This brings our year-to-date share repurchases total to about 0.8% of our outstanding shares. The main purpose of the plan is to offset dilution from stock-based compensation, and at quarter end, $26 million remained available for future repurchases. Turning to fourth quarter non-GAAP outlook. As Mike mentioned, we expect generally strong demand to continue, with sequentially higher foundry and logic and systems revenue.
These factors result in a Q4 revenue outlook in the range of $192 million-$204 million. non-GAAP gross margins for the fourth quarter is expected to be in the range of 44%-47%. At the midpoint of these outlook ranges, we expect Q4 operating expenses to be higher than Q3 by $1 million-$2 million, mainly due to increased investment in R&D and higher travel expenses as things start to get back to normal. Accordingly, non-GAAP earnings per fully diluted share for Q4 is expected to be between $0.37 and $0.45. A reconciliation of our GAAP to non-GAAP Q4 outlook is available on the investor relations section of our website and in our press release issued today. With that, let's open the call for questions. Operator?
Our first question comes from the line of Tom Diffely of D.A. Davidson. Your line is open.
Thank you. Good afternoon, and nice results. Mike, just a couple of quick questions about the industry and how it relates to probe cards. If we spend, you know, $85 billion-$90 billion on CapEx this year, what is the timing over the next year or over the next years that that's fully integrated into probe card business for you?
Yeah. It's an interesting question, Tom, and one that we've tried to highlight for people. Obviously, 2021's gonna be an extremely strong record for wafer fab equipment. And what we've seen over the years is, as those equipment gets delivered, installed, qualified, and then begins to produce designs, there's a two- to three-quarter lag between those equipment shipments and us seeing probe card demand. You can imagine it does take customers some degree of time to get the tools in, running, qualified, and be producing the specific designs that then require probe cards. Obviously, we continue to operate on very short lead times, well within a quarter. We are...
Part of the reason for our large capacity investments here in 2021, so that we're ready as all of this WFE comes online and begins producing leading edge designs, we're ready to meet that probe card demand in 2022 and beyond.
Great. As a follow-up to that, Mike, when you look at the move from, say, 5 nm-3 n or the move from FinFET to gate-all-around, what is the incremental opportunity for probe cards?
I don't know that there's a really big impact there, Tom. I mean, if we go back in history in foundry and logic, as we went from planar to FinFET, it drove up test intensity somewhat at the beginning of that transition, just because there's new defect modes that the customers have to understand and drive the yield down. They're gonna test more and drive more data through the test chain. That's gonna require more testers and more probe cards. You know, in the end, as it asymptotes out, they get pretty good at driving things back down to the standard test levels.
I think the more interesting thing that's happening is we're seeing a push towards more wafer test, probably associated with advanced packaging, where people are trying to get something closer to known good die before they stick the individual die into these advanced die stacking packages. We may be less associated with transistor architecture and more associated with the move towards advanced packaging. We do see a bit of a tailwind there.
Okay, great. That's helpful. Then, Shai, final question here. When you look at the labor market, how is that impacting you today, I guess both in light of expanding your facility in Livermore and also restructuring a few other facilities?
Yeah. It's true that, you know, we see some delays in hiring and vacancy is a little longer. You can see that by looking at our OpEx for Q3, right? We expect it to be a little higher than it came in. But we do. It's something that we keep an eye on, and we think we can manage. We see these labor shortages coming to better results as we get into Q4. We already have, you know, more people on board, you know, one month into the quarter, so that's encouraging, and we are on schedule to start production in our new manufacturing center this quarter.
Great. Thank you for your time today.
Thanks.
Thank you.
Thank you. Again, that's star one on your touchtone telephone, and we ask that you please restrict yourself to two questions before returning to the queue. Our next question comes from the line of Brian Chin of Stifel. Your line is open.
Hi there. Good afternoon. Nice results, and thanks for letting us ask a few questions. Yeah, also nice gross margin print here in 3Q, and comparable, I guess, guidance for 4Q. You know, I understand that there's costs and labor headwinds that might persist into next year, but also assuming that revenue maintains or progresses from 4Q levels, and maybe the mix kind of even enriches in terms of logic foundry, what would hamper you from continuing to operate in sort of the upper half of the 44%-47% gross margin range? Not on every quarter basis, but at least averaging out to that over a full year.
As we said in Q earnings call, the biggest factor, you know, that impacts our gross margin the most is product mix. We are running along this trend line, as Mike mentioned in the call, towards our 47% gross margin, the target model. As we make progress towards that goal, we can see things fluctuate along that line. We saw it, you know, Q3 was 46%. When we look at Q4 outlook, revenues, at least at the midpoint, are higher than Q3. Gross margin midpoint is lower than the actual gross margin for Q3, and because of that mix. That phenomenon, I think, will stay with us as we continue to make progress towards 47.
Okay. Thanks. Yeah, earlier this morning, Teradyne referenced I think a 25% increase in test intensity for heterogeneous chiplet style packaging relative to a comparable monolithic IC. Mike, I think that's within range, maybe even sort of the high side of what you've talked about previously. I was wondering how big of a tailwind could this activity represent next year? Do you see broader activity across the handful of key leaders in this space? And does that add any basis points of growth on top of the revenue targets you've talked about at last year's analyst meeting?
Maybe start with the second part of the question first. If you look back at some of the underlying assumptions that went into the target model that gets us to $850 million in about 2023. Advanced packaging and die stacking heterogeneous integration are a very prominent driver to get us to that model. You know, when we think about the growth rates that get us there in 2023, they largely are incorporated. But as you note, you know, we are seeing some pretty prominent leaders in the industry shift their roadmaps towards these heterogeneous integration techniques. Our largest customer has been very clear that they've got a major client processor coming out that's gonna feature this architecture.
you know, as that drives up test intensity, at least initially, you know, you are going to see some potential tailwinds. I think at the top level, we've contemplated these factors in our revenue growth assumptions that get us to $850 in 2023.
Okay, great. Good. That's helpful. And maybe just, I was wondering, could you take a stab maybe at, you know, if you double or even triple the transistor count on one of today's sort of leading SOCs, what kind of proportional or not increase in test times do you think we're talking about?
Yeah. I mean, test time doesn't scale as you might imagine, linearly with transistor count. There's all kinds of interesting test technologies that allow customers to make sure there's not a one-to-one correspondence to that. Going back maybe to Tom's question, I think the bigger thing, rather than transistor architecture or transistor count, is people trying to drive more and more test content to the wafer level and the die level, so that when they put the die in these advanced packages, they have a pretty high probability of making sure they're good. You can imagine a scenario in an eight die stack, where if one of those eight die is bad and you don't have the rework or redundancy available, that's a pretty big yield hit.
That's the fundamental economics that are driving up test intensity associated with heterogeneous integration and advanced packages.
Okay, great. Makes sense. Thanks, Mike.
Yep.
Thank you. Our next question comes from Craig Ellis of B. Riley Securities. Your line is open.
Great, congratulations on the nice execution and the nice outlook. Mike, I wanted to start just by getting some clarification on the Livermore facility. One, I believe the press release and your comments indicated that it would be ramping in the fourth quarter, but is it ramping for revenues, and is the revenue associated with that ramp the growth that we're seeing sequentially in the business or just a part of that growth? I guess it would be a part because you also said that systems is gonna be up quarter-over-quarter.
It's a pretty modest contributor to the overall midpoint revenue outlook, but I think it is significant that we are bringing it online on schedule, removing what's been a pretty significant constraint for us. You know, we think about this in three components, the footprint we have, the installed tool capacities that we have, and then finally, the labor in our factories. For those of you that have been following us for a while, you know we've been footprint constrained. Bringing this facility online, powering it up, getting tools installed as we're doing now, really relaxes that major long lead time constraint that we've been operating under. You know, there's other capacity expansions going on worldwide.
One of them, in our RF line, where we continue to see increasing business associated with the 5G ramp. There's other components too. You know, I'd say a modest, a little bit of the uplift in the sequential growth Q3 to Q4 is associated with the new facility, but the potential is a much more exciting opportunity for us.
Got it. And on that RF point, you mentioned that RF was particularly robust. Is that RF front end for 5G smartphones in that? Is that the move from more high-end smartphones over the first year one and year two of shipments to next year's move more into the mid-range? I know we've got a number of Android releases that are coming in the first quarter, and there's quite a bit of excitement about some new baseband product for those. Is that what you're saying, or are you seeing things related to 5G that might be Wi-Fi related customer premise equipment, that type of thing? You know, Wi-Fi 6, et cetera.
No, I think there is some Wi-Fi 6 elements to it, but I really think the strength is being driven by 5G handsets. You know, both the growth associated with 5G handsets, but probably more importantly from a test perspective, the content growth in an RF front end in a 5G handset, really pretty significant. If you look at the components and functionality in a 5G handset, the RF front end of a 5G handset, you've got a whole bunch of new components compared to a 4G handset. Of course, you're operating, in many cases, at significantly higher frequencies, up to millimeter wave frequencies, which drives up the test complexity. To answer your question, I don't think a lot of it's Wi-Fi 6. Some of it really is a handset dynamic that's driving our RF growth.
Last, for me, and at the risk of admonishment versus the two-question guideline, could you just talk a little bit more about the systems business side? You know, thought that $35 million a quarter was a very robust quarter for the business, but it looks like it's growing above that. As we look at the different pieces of that business, you mentioned FRT, but there are others. Can you talk a little bit more about what's happening there, Mike?
Yep. The systems segment, you can really think in our current state, it is broken down into three pieces. One is the legacy systems business that we acquired as part of the Cascade Microtech acquisition. That was a business that both under Cascade's umbrella and ours, kind of a low single digit grower serving R&D labs, characterization, failure analysis, things like that. We've added two of our recent acquisitions to the systems segment, one being FRT. I mentioned the robust adoption we're beginning to see, especially in the foundry ecosystem associated with the advanced packaging metrology applications that FRT serves. That's been a really nice growth driver for that segment. The other piece is our acquisition, almost exactly a year ago, of High Precision Devices, to get us into the quantum computing test space.
This is test and infrastructure for enabling all of the providers of quantum computers. There's a long list of companies beginning to commercialize useful quantum computers for some really interesting applications. We're helping them with the test and measurement applications associated with that as that industry begins to grow. As we look at the Systems segment, which a couple of years ago was $25 million a quarter, adding on these two other components, FRT and HPD, which are both growing faster but also adding the revenue to it, is the primary reason for us getting up to the mid-$30s and hopefully the high $30s in the near term.
Makes sense. Thanks, Mike.
Yep. Thanks.
Thank you. Our next question comes from David Duley of Steelhead. Your question, please.
Yeah, thanks for taking my questions. You mentioned that you had a nice improvement sequentially in growth margins. I was just curious. I'm sure there was still some supply constraint margin impact and higher freight costs. Could you perhaps help us understand how much is still dragging there with freight and supply issues?
Actually, the supplies
Yeah
Sorry, Mike.
No, go ahead, Shai.
Actually, we didn't see a big impact from increase in shipments, cost and things like that for gross margin. It's just by the nature of our products, the way we ship them, the terms with the customers; this doesn't have a big impact. It didn't have a big impact on our gross margin in Q3, and we don't think it's gonna be significant in Q4 as well.
I know it wasn't, probably wasn't very meaningful, but was there some revenue impact?
No, not at all. Pretty small. You know, sometimes you can have a big probe guard, probe card waiting for to be shipped because it's missing a small component, but it was really minor so far.
Okay.
Really great job by our operations team.
Yeah. Then as far as the FRT business, could you help us understand how big a segment of business or how big this market is? What exactly are you measuring in the advanced packages? You gave some description. If you could just dig in a little bit deeper, that would be great.
Why don't we start with the application space? I mean, there's a variety of new measurements that need to be made on these chiplets, primarily at the interfaces between the adjacent chiplets. You can imagine if you're taking two of these chips and stacking them together, you wanna make sure that the connections, typically solder bumps or copper pillars, are reasonably planar, flat, that there's no defects, things like that. A rapid measurement of planarity of those surfaces, and things like bow and warp, very important for maintaining assembly yields in these brand new chiplet applications. A lot of our measurements were associated with that, the interfaces between chiplets before they're assembled together. Now, because this is an emerging application, it's difficult to size the market.
I think, you know, we've looked at it as overall, probably from a TAM perspective, a total available market of a couple of 100 million. That's a fairly wide swath across all possible advanced packaging apps. You know, there's a variety of different competitors and alternatives that are probably gonna divide up the different sub-segments and sub-applications as, you know, as standard semiconductor applications have been. I think, you know, 150, 200 million, probably a reasonable placeholder for a total available market. It'll obviously, from these levels, take us a while to grow into that, but we're really excited about the progress from the FRT team so far.
Thank you.
Thank you. Our next question comes from Krish Sankar of Cowen and Company. Please go ahead.
Hi. Thanks for taking my questions. This is Steven calling on behalf of Krish. First question, maybe for Shai. If I could just dig in a little bit more on the gross margin during the quarter and as part of your guidance. Your probe cards revenues were flat compared to prior quarter. You guys had pretty nice margin expansion there. Was that more of a function of a better product mix within your foundry customer segment? Or did you also see some material margin expansion within the memory customer segment as well?
You were not very clear on just the quality of the line. Can you repeat that? At least part of it.
apologize for that. Can you hear me better now?
Yes, that's better.
Okay, great. Yes, the question was on your gross margin expansion within the probe card business in Q3, and for the implications about Q4 margin expansion as well. Your foundry and memory customer mix were relatively flat sequentially between Q3 and Q2. Just curious if most of the gross margin expansion within the probe card business in Q3, if that was primarily from foundry, or did you see a meaningful amount of margin expansion from memory customers as well?
Yeah. In Q3, it was, you know, both our segments had a better margin. If you look into Q4, actually the midpoint of the outlook range for gross margin in Q4 is below the 46%, right? It's 45.5%. The main contributor goes back to what we said in our prepared remarks about product mix within the specific markets. It's true that, traditionally, systems and then foundry and logic has higher margins than DRAM and than Flash, but first of all, there is overlap between them. You can have a high-end DRAM design that has a better margin than a foundry and logic design. You can certainly have, within the DRAM market, two designs with completely different margins.
As we look into Q4, what we see is we expect less favorable product mix within the markets, that's, you know, within foundry and logic, within DRAM, et cetera, and not necessarily between the markets.
Gotcha. Thank you for that. As for my follow-up question on your large Korean customer in the quarter. I imagine most of the business there is probe cards on the memory side. Can you talk about any potential opportunity longer term within their foundry business as well?
Yeah. They were, as you know, the 10% customer in Q3, and I believe they were a 10% customer in the second quarter as well. In Q3, there was a pretty strong contribution from their foundry business as well. That's probably the only customer worldwide where we have a significant DRAM business and a significant foundry and logic business, not to mention the systems business as well. You know, they've been a customer of us on the probe card side in foundry and logic for many years. There is, I think, as they grow their capability and capacity in competing for fabless customers, I do think there's an opportunity there.
Probably not much of a share gain opportunity because we feel like we do have reasonably high native share with that customer and a strong relationship across both the DRAM and foundry and logic businesses. You know, as we talked about some of the other customers making big capacity investments this year, they're certainly in that group. As they expand their foundry and logic business, we would expect to grow with that.
Great. Very helpful, Mike. Thank you.
Thanks.
Thank you. Our next question comes from Charles Shi of Needham & Company. Please go ahead.
Hi. Good afternoon, Mike and Shai. Thanks for taking my questions. First I wanna really congratulate the FormFactor team for strong execution on gross margin. Looks like not only your Q3 overall margins back up, but the probe card side is retaking the 45%. Really congratulations on that. So I wanna ask a little bit longer term questions. I wanna ask you about market share. Obviously, we know where you are strong. You're very strong with the leading IDM on the microprocessor side. You have a very strong position on the foundry side with a good exposure to the mobile SOC side. Maybe let me start with the microprocessor.
Obviously the leading IDM with the aggressive turnaround effort, it's becoming very interesting for us to really watch the competitive dynamics between them and the other microprocessor provider. Obviously, that's a fabless customer, but you may have some indirect exposure there. I want to really ask you in that space, with the market share gives and takes, what is your high level view on the potential tailwinds or either, or headwinds in the context of that competition between those two microprocessor providers? And how would the potentially changing market share could affect your microprocessor probe card business, either favorably or unfavorably? Then I have a follow-up.
Okay. Yeah, I think as you know, we obviously have strong share at the leading microprocessor manufacturer. The fabless microprocessor manufacturer has obviously posted some very impressive growth numbers over the last little while, which haven't gone unnoticed by us. Our native share is lower there. Our market share is lower there for a variety of reasons. Because it's lower and because they are gaining share, we've got a pretty strong initiative to go raise that share at the fabless microprocessor customer. Right now, you know, being candid with you, if the share were to shift one to one, that would be a net headwind for us. That would be a net loss for us. It's been an area where we've been focusing on and we got good traction.
I think consistent with our long-term strategy of being a leading supplier at all the leading customers around the world, I think we're in a reasonable position to rectify that. Some interesting subtleties with that business where they won't necessarily use the foundries test services but actually manage themselves. In that case, they're actually a fabless direct customer, which not to complicate things too much, is one of the interesting subtleties of this business.
Thanks for that great color, Mike. Maybe I want to turn to the mobile SoC processor side. It's been quite topical that one of the Chinese smartphone maker reportedly is working to bring the apps processor design back in-house, maybe starting from 3 nm. I think that this is the move kind of reflects the continued evolution of like all those system companies coming in, take designs in house, and maybe there is a headwind for one of your indirect foundry customer in the mobile space, a U.S. company there. I wonder what FormFactor thinks in terms of the potential landscape shift also in the mobile SoC space in particular.
I mean, when the system companies, they tend to favor one leading foundry over the other, where you may have a little bit stronger position there. Could you give us a little bit more color on that? Thank you.
Yeah. I mean, I think this trend of insourcing and more vertical integration with our customers is an interesting one. It’s not new, right? The hyperscalers have been taking design in-house. Obviously, the large handset manufacturer continues to take more in-house and almost becoming their own silicon company, their own fabless company. This again goes back to similar comments I made about the microprocessor situation, where it's one of the reasons why we wanna make sure we have strong shared positions and strong relationships at all of the leading foundries and with all the leading fabless customers. Because many of these fabless customers actually manage their design and test operations themselves and don't outsource them to the foundry. Creates a more complicated account management structure.
Again, I point back to really making sure that we as a company are engaged, and I think we're pretty well engaged with all of the leading fabless and foundry customers, so that these shifts in their manufacturing strategies and sourcing strategies end up being net neutral for us.
Thank you. That's all my questions. Thank you.
Thanks, Charles.
Thank you. Our next question comes from Christian Schwab of Craig-Hallum. Please go ahead.
Hey, this is Tyler on for Christian. Thanks for letting us ask a couple questions and congrats on the solid quarter. I wanted to ask about your DRAM probe card business. You know, including your kind of implied Q4 guidance of flat, 2021's gonna be a pretty strong growth year. So I'm just wondering, is there really any reason we should, you know, be suspect about this strength continue to 2022? You're serving all customers and, you know, there's gonna be continued design and no transitions next year. So I'm just wondering if there's, you know, any reason that the strength shouldn't continue.
Yeah. A good question. As you might imagine, a topic of pretty significant internal debate. I think, you know, one of the things to remember is history has taught us that DRAM probe cards, like all DRAM supply chain elements, are a cyclical business. You know, for the past couple of quarters and we expect here in the current fourth quarter, you know, kind of to be at the historical high end, $30 million-$40 million quarterly DRAM probe card revenue, which is above sort of the mid-$30s that we've talked about being the average normalized level. The fundamental drivers for that, as you note, are continued node transitions and design releases.
As long as those continue in the customer base, you know, I don't see any reason why at least the mid-30s assumption is not a reasonable one. We are a little bit surprised that we've got now two quarters in a row in the books and one quarter looking at the current one sort of hovering around that $40 million level, which, you know, represents the all-time record quarter, I think, at $42.5 million or $43 million. We're operating pretty close to all-time record levels in DRAM. Given the historic cyclicality, that might be a little surprising, but, you know, the fundamental constituents who drive that demand are still there, and I think we're executing pretty well in DRAM. I think we've probably gained share here in 2021.
That's great. Appreciate that color. Second question, kind of on the supply chain, it sounds like you guys are executing well and not seeing any material impacts yourself, but I think you made the comment that you do expect the environment to persist into next year. Presumably, you know, that's coming from conversations you're having with customers. I'm just wondering if, you know, there might actually be a benefit to this environment, you know, further conversations with your customers, at least benefiting you from a visibility standpoint. Any comments there on the supply chain impacts would be great.
Yeah. I mean, the supply chain, you know, issues have been really kind of interesting. Ironically, many of them have their root cause back in the semiconductor shortage. You know, where we've had challenges with subsystems or subcomponents that are not vertically integrated, they really have been associated with some isolated chip shortage. I don't think it's improving visibility all that much. I mean, we're still running lead times well within a quarter. We'd hope that, you know, as the capacity comes online associated with all of the different increases in capacity across the industry, especially at the leading edge node to satisfy this shortage, that again should be driving pretty solid demand across our different businesses.
You know, part of what we've benefited from, as I mentioned in the prepared remarks, is pretty vertically integrated for a tech manufacturing company in the semiconductor ecosystem. That was not a very fashionable move to make several years ago, but you know, we did it for a couple of good reasons, one of them being continuity, and it has paid off here a little bit.
That's great. Appreciate that. That's all for me. Thanks, guys.
Thanks.
Thank you. At this time, I'd like to turn the call back over to Mike Slesser for closing remarks.
Great. Thanks everyone for joining us again. We're hopeful to attend a couple of conferences at end of the year. I don't know whether they'll be in person or virtual, but hope to see you there, and thanks for your continued interest in FormFactor. Bye-bye.
This concludes today's conference call. Thank you for participating. You may now disconnect.