FormFactor Inc (FORM) 2022 Q4 法說會逐字稿

Thank you, and welcome, everyone, to FormFactor's Fourth Quarter 2022 Earnings Conference Call. on today's call are Chief Executive Officer, Mike Slessor; and Chief Financial Officer, Shai Shahar. Before we begin, Stan Finkelstein, the company's VP of Investor Relations, will remind you of some important information.

Thank you. Today, the company will be discussing GAAP P&L results and some important non-GAAP results intended to supplement your understanding of the company's financials. Reconciliations of GAAP to non-GAAP measures and other financial information are available in the press release issued today by the company and on the Investor Relations section of our website.

Today's discussion contains forward-looking statements within the meaning of the federal securities laws. Examples of such forward-looking statements include also with respect to the projections of financial and business performance, future macroeconomic and geopolitical conditions; the benefits of acquisitions and investments in capacity and the new technologies; the impacts of global regional and national health crisis, including the COVID-19 pandemic; anticipated industry trends; potential disruptions in our supply chain; the impacts of regulatory changes, including the recent U.S.-China trade restrictions; the anticipated demand for products are built 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 2021 and in our other SEC filings, which are available on the SEC's website at www.sec.gov and in our press release issued today. Forward-looking statements are made as of today, February 8, 2023, and we assume no obligation to update them.

With that, we will now turn the call over to FormFactor's CEO, Mike Slessor.

Thanks, everyone, for joining us today. As anticipated, FormFactor's fourth quarter revenue and profitability were down sequentially from the third quarter. That said, revenue exceeded the outlook range and non-GAAP earnings per share were above the midpoint. Gross margin was below the midpoint due to larger-than-expected end-of-year excess and obsolete inventory charges. The revenue upside compared to our October outlook was primarily driven by our ability to ship products to certain domestic China customers in compliance with new export controls, which also drove stronger-than-expected DRAM probe card and systems revenues in the quarter.

As we start the first quarter of 2023, we're experiencing similar overall demand for the fourth quarter, with moderately stronger demand for foundry and logic probe cards, offset by weaker demand for both DRAM and Flash memory probe cards. At the same time, our Systems business continues to run at record levels. We expect a significant increase in profitability in the first quarter on revenue levels similar to the fourth quarter from gross margin improvement driven by 2 factors: first, the full quarter benefit of our October restructuring; and second, a return to typical excess and obsolete inventory costs.

With probe card lead times typically less than a quarter, our visibility remains very limited, but we are encouraged by the first quarter stabilization of demand for our products as customers invest in innovations like new chip designs and advanced packaging. These customer investments are producing steady demand even in the face of industry-wide weakened high unit volume end markets, like mobile handsets and client PCs, along with export restrictions and serving domestic China semiconductor customers.

As we are a U.S.-based supplier with significant exposure to the leading-edge foundry and memory technologies and customers affected by recent U.S.-China export controls, these regulations are a headwind in all of our businesses. FormFactor continues to take all necessary steps to ensure full compliance with the new rules by holding shipments and support as required. As additional information has emerged regarding the scope of these regulations, our local China team has worked closely with customers and our trade compliance team to enable permitted shipments to certain customers.

I'd like to recognize this team for their outstanding work, which enabled us to ship more than we initially expected to domestic China customers in the fourth quarter. We expect to continue this level of domestic China shipment activity in the first quarter. Overall, we believe this U.S.-China trade headwind will persist over time, as we do not anticipate any relaxation of advanced semiconductor export controls. This, of course, provides a strong incentive for domestic China customers to onshore their supply base and deemphasize foreign suppliers like FormFactor to ensure their business continuity.

Even as we navigate through this trade issue and the industry's current cyclical weakness, we believe the core tenets of FormFactor's strategy remain firmly in place. These core strategic tenets are: first, sustained long-term semiconductor content growth in both consumer and enterprise applications; second, the industry's relentless investments in new technology and capacity; and third, the device-specific consumable nature of advanced probe cards, which when combined with our customers' innovation-driven investments in engineering systems have historically resulted in less volatile demand cycles than wafer fabrication equipment. Consequently, we remain committed to achieving our target financial model and continue to invest in both R&D for new product innovation and competitive differentiation, as well as in the long lead time facilities and equipment portions of our capacity increase plans. These investments are designed to produce market share gains in above-industry revenue and profit growth when we emerge from the current cyclical downturn, positioning FormFactor to achieve and even surpass the levels of our current target financial model.

Turning now to segment level details. In Foundry and Logic probe cards, our largest business, we see moderate strengthening in the first quarter, driven primarily by pilot production ramps for new mobile application processor designs together with stronger demand in our microprocessor business. Although customers are still burning off the excess inventory of processors, modems and RF chips in the channel, they're also investing in early production of innovative new chip designs that will ramp in volume once this inventory is consumed.

This provides insight into the unique characteristics of probe card demand. Since probe cards are a device-specific consumable that is specific to each individual customer chip design, this early production activity for these new chip designs is driving demand for new probe cards, albeit at lower overall levels than we'd expect to see in a cyclical upturn.

As an example, we continue to ship probe cards to support pilot production of a major chiplet-based client CPU product despite the well-documented weakness in end market client PC demand because customers continue to innovate to differentiate their future product road maps. As we noted in the past, chiplet-based advanced packaging processes like Foveros and 3D Fabric are an exciting opportunity for FormFactor. These integration schemes drive both higher test intensity, which expands the number of probe cards required per wafer out; and higher test complexity, which raises the performance requirements for the probe card.

Advanced probe card architectures like FormFactor's MEMS technology are essential to meet these challenging technical requirements at a compelling cost of ownership while also meeting the short delivery lead times needed to support our customers' rapid and dynamic production ramps.

Turning to memory probe cards, we expect first quarter sequential weakness in both DRAM and Flash probe card demand driven by extremely weakened market conditions for both DRAM and Flash chips that are causing our customers to reduce the magnitude and speed of their new product ramps. Despite this memory end market softness, each of our customers continues to release and validate new chip designs, like high-density DDR5 DRAM to drive the road map forward. As in Foundry and Logic, this new design activity is driving demand for new probe cards specific to each one of these new memory chip designs, but at low levels aligned with the reduced output levels announced by each of our major memory customers.

In the Systems business, we expect the first quarter to sustain the strong momentum which produced record revenue in both the third and fourth quarters of 2022. This strength partially offset some of the downturn driven softness in the production probe card business, highlighting the financial benefits of our lab-to-fab diversification strategy. The strategic benefits of the Systems business are also significant, as we partner with leading customers in their R&D labs to advance the industry state-of-the-art with innovations like gate-all-around transistors, advanced packaging, silicon photonics and quantum computing.

We're also working with key customers on advancing high-power applications, like silicon carbide and gallium nitride, as the semiconductor industry enables the widespread electrification of the automotive industry.

Finally, Ray Link, one of our directors, has notified us that he will not stand for reelection at this year's Annual Meeting of Stockholders. Ray has been a valuable member of our Board and has provided many useful insights both to me personally and to our management team. I'd like to thank Ray for his nearly 7 years of service to FormFactor and wish him well in all his future endeavors.

I'd like to close by reiterating that in the short term, we're encouraged by the stabilization of demand for our products in the first quarter. In the longer term, we remain confident in the growth prospects for FormFactor in the industry overall, driven by the fundamental trends of semiconductor content growth and innovations like advanced packaging. These are trends where FormFactor is well positioned as an industry and technology leader, and we're confident that our resilience and commitment to invest in R&D and capacity will position FormFactor to emerge from the current downturn a stronger and leaner competitor, enabling us to achieve our target model that delivers $2 of non-GAAP earnings per share on $850 million of revenue.

Shai, over to you.

Thank you, Mike, and good afternoon. As you saw in our press release and as Mike mentioned, Q4 revenues exceeded the high end of our outlook range. Non-GAAP gross margin was at the low end of the range and non-GAAP EPS was at the high end of the range.

Fourth quarter revenues were $166 million, an 8.2% sequential decrease from our third quarter revenues and a year-over-year decrease of 19% from our Q4 '21 record revenues. As Mike mentioned, Q4 revenues were above the high end of our outlook range, mainly due to our ability to ship to certain domestic China customers.

Probe card segment revenues were $124.4 million in the fourth quarter, a decrease of $15 million or 10.8% from Q3. The decrease was driven mainly by lower Foundry and Logic and DRAM revenues. Systems segment revenues were a record $41.6 million in Q4, a $0.1 million increase from the record third quarter and comprised 25% of total company revenues, up from 23% in Q3. Within the probe card segment, Q4 Foundry and Logic revenues were $82.1 million, a 9.4% decrease from Q3. Foundry and Logic revenues comprised 50% of total company revenues, similar to the third quarter.

DRAM revenues were $27.3 million in Q4, $7.6 million or 21.8% lower than in the third quarter, and were 16% of total quarterly revenues as compared to 19% of revenues in the third quarter. Flash revenues of $15 million in Q4 were $1.1 million higher than in the third quarter and were 9% of total revenues in Q4, slightly higher than the 8% in Q3.

GAAP gross margin for the fourth quarter was 27.2% of revenues as compared to 34.4% in Q3. Cost of revenues included $7.5 million of GAAP to non-GAAP reconciling items, which we outlined in our press release issued today and in the reconciliation table available in the Investor Relations section of our website. Q4 gross margin reconciling items included a $4.1 million charge related to the restructuring we announced in October 2022.

On a non-GAAP basis, gross margin for the fourth quarter was 31.7%, 7.3 percentage points lower than the 39% non-GAAP gross margin in Q3 with lower gross margin in both the probe cards and the Systems segment. Our probe card segment gross margin was 25.5% in the fourth quarter, a decrease of 9.1 percentage points compared to 34.6% in Q3.

The decrease is mainly due to 2 factors: first, lower overall segment revenues and lower factory utilization, which accounted for 6 percentage points; and second, higher than usual excess and obsolete inventory reserves, which accounted for approximately 2 percentage points. As a reminder, our cost reduction measures were implemented at the end of October '22, so Q4 only partially benefited from the overall expected savings.

Our Q4 Systems segment gross margin was 50.4%, 330 basis points lower than the 53.7% gross margin in the third quarter, reflecting a less favorable product mix. Our GAAP operating expenses were $61 million for the fourth quarter, $3 million higher than in the third quarter. The increase is attributable to the restructuring plan costs we implemented in Q4, higher stock-based compensation as a result of one more week of -- sorry, one more working week in Q4 partially offset by the savings from the cost reduction measures we took during the quarter.

Non-GAAP operating expenses for the fourth quarter were $47.9 million or 28.8% of revenues, as compared with $49.5 million or 27.4% of revenues in Q3. The $1.7 million decrease relates mainly to the impact of the restructuring we implemented during Q4, lower performance-based compensation and higher PTO taken. Company noncash expenses for the fourth quarter included $9.5 million for stock-based compensation, $1.5 million higher than in the third quarter due to one additional week of vesting during Q4 and the timing of annual grants; $2.9 million for the amortization of acquisition-related intangibles, similar to the third quarter; and depreciation of $7.5 million, $0.4 million higher than in the third quarter.

GAAP operating loss was $16 million for Q4 compared with GAAP operating income of $4 million in Q3. Non-GAAP operating income for the fourth quarter was $4.8 million compared with $21 million in the third quarter. GAAP net loss for the fourth quarter was $13.7 million or $0.18 per fully diluted share, compared with a GAAP net income of $4.4 million or $0.06 per fully diluted share in the previous quarter.

The non-GAAP effective tax rate for the fourth quarter was 21.3%, 230 basis points higher than the 19% in Q3. The annual non-GAAP effective tax rate for fiscal '22 was 15.4%, and the low end of our estimated non-GAAP annual effective tax rate of 15% to 20%. For 2023, we expect the benefits of the new advanced manufacturing investment credits, or AMIC, to lower our non-GAAP effective tax rate to mid- to high-single digits, approximately 6% to 9%, similar to our annual cash tax rate.

Maintaining profitability at the current reduced demand levels is an important goal for us. The actions we took during Q4 to reduce our costs contributed to a fourth quarter non-GAAP net income of $4.1 million or $0.05 per fully diluted share, compared to $18.3 million or $0.24 per fully diluted share in Q3.

Moving to the balance sheet and cash flows. We had negative free cash flow of $5.4 million in the fourth quarter compared to positive free cash flow of $15.5 million in Q3. The $21 million difference between -- related to a $4 million decrease in net cash provided by operations mainly as a result of severance payments related to the restructuring made during Q4, and an increase of $17 million in capital expenditures. At quarter end, total cash and investments were $242 million.

As of the end of the fourth quarter, we had 1 term loan remaining with a balance totaling $15.5 million. We invested $26.2 million in capital expenditures during the fourth quarter compared to $8.9 million in Q3. With the core drivers underpinning our strategy still in place, we continue to execute on increasing our long lead time facilities and equipment portions of our capacity increase plan, a bit placing equipment in service at a slower rate to ensure capacity does not significantly outpace demand.

Capital expenditures in 2022 totaled $65.2 million at the midpoint of the estimated range we previously communicated. For 2023, we expect CapEx to range between $50 million and $60 million, $10 million lower than in 2022 at the midpoint of this range.

Regarding stock buyback. During the fourth quarter, we purchased approximately 365,000 shares under our $75 million 2-year buyback program for a total of $8.9 million. At Q4 quarter end, $18.6 million remain available for future repurchases.

Turning to the first quarter non-GAAP outlook. As Mike mentioned, we expect Q1 revenues to be comparable to Q4 with higher Foundry and Logic revenues offset by lower DRAM and Flash revenues, with Systems segment revenues similar to Q4. This demand results in a Q1 revenue outlook of $162 million, plus or minus $5 million. We have largely completed the restructuring that we announced in Q4 to reduce cost and improve the efficiency and effectiveness of our business, and Q1 outlook reflects the full impact of these savings. As a reminder, we estimate that these actions will reduce our cost structure by $25 million to $30 million on an annual basis, with approximately 2/3 of the savings benefiting cost of sales and 1/3 benefiting OpEx. In addition, we don't expect a similar excess and obsolete inventory reserve charge in Q1.

Accordingly, first quarter non-GAAP gross margin is expected to increase to 37%, plus or minus 150 basis points. At the midpoint of these outlook ranges, we expect Q1 operating expenses to be $50 million, plus or minus $1 million, with a $2 million increase as compared to Q4 mainly due to annual benefits reset and less PTO taken in Q1. Non-GAAP earnings per fully diluted share for Q1 is expected to be $0.13 plus or minus $0.04. Reconciliation of our GAAP to non-GAAP Q1 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?

(Operator Instructions) Our first question comes from the line of Brian Chin from Stifel.

Maybe Mike and the team, can you first size the piece of the China domestic revenue that you -- sounds like you initially excluded from guidance, but ultimately, something that you could ship in the quarter. Can you maybe put some parameters first on what that represented?

Yes. Brian, when we talked about it in the October call and set expectations for Q4, we talked about it being a $10 million to $15 million headwind in the quarter. We certainly didn't recover all of that, but recovered probably more than half of it. There were some other pieces of momentum in the business outside of the China domestic customers that helped get us above the high end of the revenue range.

But as I said in the prepared remarks, we did do a better job once we got some clarifications on these regulations in making sure we have the appropriate documentation, appropriate paperwork to get the shipments to the customers. So perhaps in a little bit more effective way than we originally thought we were going to be able to.

Got it. Okay, that's helpful. And Mike, I think just to state one more question on that topic. I think you've been pretty clear about, sort of, how you see that business potentially, some of the domestic China business transitioning over time. Do you think it would be sort of more of a gradual phase out?

And I guess there's nothing gradual about U.S. restrictions. So maybe put that to the side. But -- and just in terms of the activity you're seeing in engagement, do you feel there's any sort of pull-in happening right now from later in the year into now or even longer horizons based on just the uncertainties that maybe some of your -- the local customers have right now?

Yes. I think there's several facets to that question. We have tried to be as clear as we can with people about our long-term assumptions that we're going to be really challenged to serve the domestic China market as a non-China/U.S. supplier.

Over time, as you say, and there is a question of what that trajectory looks like. But over time, we would expect that business to go to zero, if not close to it. As we look at what's happening right now, I wouldn't say there's much pull in. Now that's probably because, again, the bulk of our business, although the Systems business is performing nicely, it's probe cards, which are a design-specific consumable. So unless customers have the mass sets fix, the design fix, it's really impossible for them to pull that demand in out in front of any other anticipated restrictions or headwinds. So we typically don't see pull-ins for structural kind of conditions like that, and we're not seeing one here.

Okay. That makes sense. In terms of your -- last question on sort of the commentary around seeing stabilization in Logic/Foundry, at least at this point. What's your sense in terms of Logic, foundries as well as memory, when we tie this into inventory reductions and obviously, there's some fab utilization reductions tied into that. Do you see a situation through your discussions where there's optimism that utilization rates come back up in the second half?

Maybe some of the inventory management has had effect here in the first half. And so any way of characterizing how you view your business into the second half, if those are some of the dynamics that are being discussed right now.

Yes. I think, first of all, the dynamics are very different in different segments. Obviously, anybody who covers the Memory segment, both DRAM and Flash, understands that there's significant inventory consumption that has to happen, and different customers are approaching that in different ways.

Foundry and Logic does appear to be stronger, but there's areas like client PC that still have a tremendous amount of inventory to burn through. Compounded on top of that, obviously with very short lead times, in the probe card business, we're going to be seeing when customers turn back on wafer starts in production in almost real time, with lead times well within a quarter to look into a second half utilization recovery is something that we just don't have the visibility, the length of lead time and the backlog, to really make any intelligent comments on.

(Operator Instructions) Our next question comes from the line of Christian Schwab from Craig-Hallum.

Congrats on good execution in a challenging environment. On your China exposure, can you just tell us the mix of the China revenue? I assume that it's not all 100% domestic Chinese manufacturers, but other people manufacturing inside of China. So when you talk about domestic China over time going to zero, can you bracket that revenue expectation?

Sure, Christian. And we've done this at different times in the past. But a good way to think about the past 2 quarters have been $50 million of revenue shipped into the region, shipped into China. But the majority of that, between 2/3 and 3/4, is to the multinationals that operate in the region.

We have big DRAM customers there. We have Flash memory customers there. And many people don't know this, but there's a very large microprocessor assembly and test facility in China run by one of the multinationals as well. So the bulk of that $50 million quarterly China revenue really is the multinationals.

Our domestic China revenue, obviously, is then the remainder. And that's been running somewhere between $10 million to $15 million a quarter. It's a little bit lumpy. We've seen nice activity out of a DRAM customer in China. The Systems business continues to perform well, so it does go up and down.

But if I were to size it for you, I think somewhere between $10 million and $15 million a quarter is an appropriate way to think about the domestic China business.

Great, thank you for that clarity. On the Systems business, you guys talked about silicon carbide. I wonder if -- which is a strong growth area, obviously, can you kind of talk about your revenue opportunities there and what it means to the Systems business as the number of customers expand and the number of wafer starts expand, how should we be thinking about that portion of the business?

Yes. So the Systems business overall is a pretty diverse mix of applications. As we've talked about in the past, running from mainstream CMOS right now working on gate-all-around 2-nanometer kind of things, all the way through quantum computing, silicon photonics and power applications like silicon carbide, which is obviously on a great growth trajectory right now.

But one of the things that's important to remember about the Systems business is it really is R&D focused. So as silicon carbide, wafer starts continue to accelerate and ramp as they are expected to, we don't have a ton of leverage and exposure there, because we're involved in the early development and the yield improvement associated with us.

There are areas of some of these applications. Probably silicon photonics is the most promising for us right now where we do have some production leverage. But I'd caution people on sort of using us as a proxy for silicon carbide production activity. We're really enabling initial yield improvement, customers moving from 4-inch to 6-inch to 8-inch wafers, things like that in their early R&D associated with silicon carbide.

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

Mike, I listened to your prepared remarks, and I kind of get the sense that it looks like your mobile side of the business, which includes both SoC and RF seems to be that it's probably moderately strengthening. DRAM, I think based on the commentary last quarter and the upcoming quarters, it may be as bad as it gets. But things are probably in a stabilization process in terms of how -- the rate of decline.

But I do want to ask you more on the microprocessor side. I mean, stripping out any contribution, let's say, from the (inaudible) or Mac OS-based, the microprocessor demand, I believe March quarter, you will see some of the strength from that particular customer. But x86 side, do you see stabilization from the December quarter level? And I think that particular customer, they probably have prematurely called bottom 2 quarters ago. And -- but going forward, do you see where things will bottom out from your #1 customer in terms of your demand?

Yes. So a couple of different things in that question. Let's start with the last one first. We are seeing incrementally stronger x86 microprocessor probe card demand here in the first quarter. Again, making a comment about anything beyond that is extremely difficult, given the short lead times we have in the business well within a quarter.

And we've seen overall the market, not just in the microprocessor probe card space, but overall in the probe card space, be pretty volatile over the last couple of quarters. So I think right now, again, we are encouraged by what's clearly a stabilization in our overall revenue and some strength in the Foundry and Logic space, driven both by the x86 microprocessor also, as you noted, in the mobile business.

Now shifting gears to the mobile business for a second. It really is isolated to mobile application processors. We are not seeing any significant strengthening in RF. Some of that's due to the different product launch cycle times, but I think you can draw a pretty straight line between the mobile application processor probe cards we're shipping now and major mobile handset releases that are scheduled for later in the year or expected for later in the year.

Got it, got it. That was very clear. I appreciate that. Maybe I want to ask you a little bit more on the DRAM side. Maybe I should ask from a historical perspective. I think back in 2019, that was a prior industry downturn.

Your DRAM business was bad in the first quarter, but recovered in -- starting from the second quarter. I know your drivers for your DRAM probe cards are not exactly the same as the WFE folks. But do you think the same condition may exist this year as we go into the downturn '23, that DRAM probe card business may start to recover a little bit earlier than people thought? Or do you think at this time may be a little bit different compared with 2019?

I'll caveat my answer with the usual understanding of lead times being very short and this being a very dynamic environment across all of our businesses, including DRAM probe cards.

What I will say is the depth of this DRAM downturn. If you look at the DRAM spot market pricing, what our customers are doing, it's very difficult to draw a parallel to 2019. This is a much more severe pullback than 2019. And as a consequence, our customers, although they're behaving differently, are certainly behaving differently than in 2019. They're behaving differently from each other and behaving differently from 2019.

So I think that's a parallel. If inventories get consumed in a relatively short term, in a handful of quarters, I think you could see customers take some of these new devices, things like high-density DDR5 for mobile, and ramp those more aggressively. But as long as this overhang of inventory and what's a pretty unhealthy (inaudible) DRAM chip market persists, we don't anticipate any significant recovery as we go through 2023 here.

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

I wanted to start with just a few follow-ups. So Mike, very helpful to get the color on what's happening in Foundry and Logic and a better sense of what's going on with your largest customer. On that front, they've been public, and I think you in the past have talked about some early shipments to a tile-based product, which for them launches in the second half of the year.

Can you just give us an update on how trends are looking there and the degree to which that's contributing to revenues in the current quarter or could in the coming few quarters?

Yes. I talked about a little bit in the prepared remarks, because I think it's a very important trend for FormFactor and for the industry overall. And we are continuing to ship in volume for what's really the first client-based CPU on a tile or chiplet architecture.

And we certainly, along with other people I know, view this as a significant event in the industry because of the significant volumes for the client-based CPU drives throughout the entire supply chain. So that will be if you like a really good high-volume (inaudible) for tile architectures, for chiplet structures and for advanced packaging in general.

Obviously, the magnitude of that opportunity though is, as we go through the year, really going to depend on the decisions our customer makes on how aggressively they ramp that product. And presumably, a lot of that will have to do with the client PC market. The inventory stabilization there, the inventory consumption there, and then their willingness to drive the new product through the channel.

I do think it's an interesting example we're seeing in all of our customers in Foundry and Logic and DRAM. Although volumes are obviously down from the peak levels we had in the first half of 2022, each of these customers are continuing to invest in new designs so that their road maps are differentiated. They're not doing it at the volumes, requiring the volumes of probe card they would in a cyclical upturn, but we are seeing very healthy new design activity that leaves us pretty optimistic about growth when growth returns to the overall industry and this inventory correction gets done with.

Got it. That's helpful. And then I wanted to follow up on DRAM and just try and dig in a little bit. So I think oftentimes when we talk, we talk about the generations of DRAM that might be in user coming into production DDR4, DDR5, DDR6, et cetera.

But I wanted to talk a little bit about DRAM from an application basis, if the team has optics there. Because it seems like what we're hearing from all our checks is that -- the mobile market is starting to find an inventory bottom. And obviously, PCs or progress and the server correction was latest to the party, which is why we're just starting to see server DRAM collapse in the first quarter.

But the question to you is, as you look at your DRAM business, are you seeing signs of relative stability in some of the application areas? Is it all acting the same? How would you frame things up for us on those parameters as you look at the market now and what's possible in the first half of the year?

Yes. I'll go back to some of the questions we've answered earlier. Given how dynamic things are in the industry and our short lead times, it's pretty difficult for us to make any coherent remarks on those subsegment levels and how the inventory corrections are going.

What we are seeing, although there's a sequential reduction in our projected DRAM revenue going from Q4 to Q1, we are seeing relative stability in the design activity. That's part of the reason why we're encouraged that overall Q4 revenues for the company are nominally the same as we're seeing here in demand that's going to drive Q1 revenue.

Different puts and takes, obviously. As you note, the different end markets applications for DRAM are all on a slightly different cadence in terms of supply-demand imbalance and inventory corrections. But for somebody -- for a company that's operating with lead times well within a quarter, it's pretty difficult for us to say anything coherent about the health of those end markets.

Got it, got it. Okay. And then if I could just sneak in a last one before I hop back in. Shai, you talked about the $55 million in CapEx for calendar '23. Can you just give us the top 2 or 3 buckets that we should expect that would deploy into? And any color around those would be appreciated.

Sure. The majority of this $50 million to $60 million investment in Q3 will be in tools and equipment, to continue investing in the long lead time items that are required to increase our capacity.

Like we opened a manufacturing -- a new manufacturing center in Livermore more than a year ago. We started with (inaudible). We are since then populating it with tools. What we're doing is because these tools have long lead time, we don't want to slow down the purchasing process.

But we do slow down the process of putting them in service and starting depreciation because we want to align as much as possible this capacity online with the demand coming in from our customers.

And our next question comes from the line of Hans Chung from D.A. Davidson.

First, I want to follow up on the mobile commentary. So can you give me an idea like what's the (inaudible) revenue cycle for that -- the mobile customer for new product ramps? And since that, we start to see from the application process and then (inaudible) just give me a general idea like maybe like 1Q, 2Q, third quarter, which quarter will be strongest in terms of (inaudible) with that customer.

Yes. So Hans, the mobile application processor piece, although I think in -- earlier in the Q&A session, I gave some pretty decent visibility about where that project is headed, that historically that has been primarily a Q1, Q2 activity.

Given where the industry sits now, I think we see perhaps it being a little more spread out. But one of the other interests -- so maybe some contribution in Q3. But think of it mostly as a first half of the year kind of thing. I think the other interesting thing is the mobile application processor activity, although it's dominated by the one big project in the industry, we do see other customers also releasing new apps processors, new mobile application processors that are going to go into other handsets in the Android ecosystem.

So some interesting multifaceted activity there that all goes back to this theme of despite the demand downturn overall in the industry, our customers continue to innovate and release new designs so that they're ready for the upturn and ready to differentiate their product road maps when things resume to grow.

Got it. And then is RF coming later than the (inaudible) like by a quarter or so?

Yes. So RF is typically later than the apps processor in the overall cycle because it has shorter lead times. It has shorter fab cycle times. It has shorter lead times for our probe cards. It has shorter assembly cycle times. So it's just more compressed and closer to the actual handset launch than the more complicated silicon like the apps processor.

Now the one thing we're keeping a very close eye on is, obviously in RF, there's still a pretty good inventory build of things at the component level, things like (inaudible) SAW filters, and you've heard from our customers that there still is an inventory correction ongoing there. So keeping our eye on that. But again, new design activity would be expected as we go through, call it, the middle part of the year to support the late-year handset launches.

Got it. That's very helpful. So the next question is about the gross margin. And -- so it seems like the improvement in Q1, mainly driven by the inventory reserve back to normal level and then also the restructuring effort. And the price mix helped for Q1, given we think that we have some strength in Foundry Logic and a weakness in Memory. And then how do you think about the rest of the year? Like just in terms of the -- maybe the mix or the efficiency gain and et cetera. Just how should we think about the gross margin trend throughout '23?

Yes. So I think you listed exactly the 3 main factors that impact the expected increase in gross margin from Q4 to Q1, right? You mentioned the no special or excess E&O reserve, the restructuring benefits to be a full quarter benefit in Q1 and the more favorable mix in Foundry and Logic and DRAM. These are the main factors that are taking us back to 37% plus/minus 50 basis points.

As for the rest of the year, it really depends on revenue and depends on mix. And as Mike mentioned, I think a few times during this Q&A, we still don't have a lot of visibility. But we are encouraged by the stabilization of the revenue between Q4 and Q1. We are adding capacity to make sure that when our customers ramp, we are ready to be there and supply to them.

And as historically, we demonstrated our ability to perform at high 40s if you go back to Q1 and Q2 of 2022, when revenues were around $200 million, we were able to achieve our gross margins around the close or even exceeded our target model. So as the industry recover and our revenue increase, we still -- we are still confident in our ability to achieve our target model and our gross margin as well. And it's going to fluctuate as we go up there along that trend line.

And our next question comes from the line of Krish Sankar from Cowen.

This is Steven calling on behalf of Krish. I guess if I could start, I had also one more question on the Foundry and Logic business. I wanted to ask you in entirely different way as opposed to from an end market perspective. Do you have any perspective on from a leading-edge versus trailing-edge perspective, how the Q1 sequential improvement in demand or flattish to improved demand, how that's breaking out across, call it, 16-nano smaller geometry products versus a lot larger geometry products, first of all?

Yes. Steven, it's an interesting question for us because most of our exposure is on leading-edge nodes. If you think about the probe card business and even the Systems business, really what we're doing is enabling customers to improve their yields on these advanced nodes. And in the probe card business, screening out bad chips before they go downstream to what are becoming more expensive assembly processes.

So if you think about where that's going to be deployed most, for sure it's on leading-edge nodes and on brand-new nodes where customers are very focused on yield improvement and trying to get yields up to entitlement levels. Our exposure on trailing-edge nodes is quite a bit more limited. It's restricted to things like microcontrollers in the automotive segment, where there's requirements like high temperature, high parallelism.

So we're not a great read through on the mix of leading-edge versus trailing-edge. So I think you can conversely look at the stabilization of at least our outlook as we go Q4 to Q1, as a commentary on some of the stabilization that leading-edge nodes. We really are much more exposed to leading-edge nodes than we are trailing-edge.

Great. And then one quick one on Systems. Just given the strength of that business and the, I guess, the multitude of sort of R&D applications that are driving that currently, do you expect or foresee any, call it, your pockets and demand just from -- given the strength from all the different applications that have been driving revenue in that business so far?

Well, one of the nice things about that business is it is rather diverse, right? Across semiconductor, optoelectronics, broader sets of applications. And as we talked about the past -- in the past, a wide variety of customers and applications.

I think as long as customers are continuing to drive their R&D budgets, their innovation road maps forward, we feel pretty comfortable about the strength of that business, primarily because of its diversification. So I don't see anything that's flashing any warning signs in the space across all of those customers and applications.

The one place that is a headwind, and has been a headwind, is obviously our ability to supply into the domestic China market. But I think that's -- as we said, in Q4 and here in our Q1 guidance, largely reflected in that overall view.

And maybe one last one for Shai, also on gross margins. For the non-GAAP gross margin guidance at 37%, what is the embedded underutilization charge in that? And what would be the sort of revenue level you need to get back to, in order to get those charges to be de minimis?

Yes. So in terms of realization and utilization of our factories, if you think about the 3 main components that are contributing to utilization, you have labor, tools and facilities.

With labor, with the restructuring we had in Q4, we are basically 100% utilized, right? We took down the workforce to labor, to the levels required to support this level of revenue. And we're going to ramp it as needed in the future.

We do have enough tools and enough facilities and footprint to support larger revenue as we saw in the first half of 2022 and enough to support our target model of $850 million of revenue. And -- can you repeat the second part of the question?

Yes. I guess what would be the revenue level that's needed to not have any, I guess, additional charges exercised from an equipment and facility standpoint?

Right. So what we said in the previous call and since then is that with the new cost structure at revenue levels similar to Q3, which was around $181 million to $184 million, we expect gross margin to go back to the low 40s.

And then in order for gross margin to reach the target model, mid-40s and high-40s, we need revenue to grow back to the $200 million plus on a quarterly basis. And we need the growth to come from Foundry and Logic because that's our higher-margin market.

And our next question comes from the line of Gus Richard from Northland Capital Markets.

There's going to be a microprocessor using chiplets, which I would believe is about an order of magnitude larger than any other chiplet product in the past. And I'm just wondering if you could talk about, sort of, how that changes the incremental opportunity per 1 million or 10 million units for you all?

Right. Gus, we touched on this a little bit prior Q&A, and I told you we were active in this project in the prepared remarks. I think this is a great example of why we're excited about the chiplet opportunity.

If you think about the test intensity and test complexity required for chiplet designs, test intensity because you want to make sure that each of the chiplets is good before you assemble them in the stack, before you tile them together. But also the complexity. Things like speeds are going up, the temperature ranges with which the chiplet tests are broadening.

So the technical requirements for what we need to deliver to our customers for chiplet designs are substantially more capable probe cards than you have to do for a single die. You roll all that together, and we've looked at it in some of the ATE manufacturers have come to a similar conclusion. It looks like about a 20% to 25% uplift on a like-for-like basis.

So the end good die out, you're going to get 20% to 25% more opportunity associated with the probe card spend when moving to a chiplet architecture. Now over time, that's going to decrease. Customers are going to get better at yield improvement, better at their test methodologies. But nonetheless, chiplet processes, tile processes, are a much more test-intensive process that has good ROI for our customers, right?

Packaging bad chiplets together with good chiplets, really not very economically viable. And so we view ourselves as a key enabler in the probe card business and in other businesses to helping the industry innovate and drive advanced packaging strategies like chiplets forward and we view it as a great financial opportunity, too.

That was super helpful. And then my follow-on is on your System business. On photonics in your prepared remarks, you mentioned you were sort of on the cusp of photonics opportunity that was moving into production. I was wondering if either: A, you could quantify it; or give a little color on what exactly in photonics the application is?

Yes. So most of the activity we're seeing in silicon photonics and the Systems business falls into 2 categories. Some of it's detectors, things like components for LiDAR. But probably the more exciting one is co-packaged optics, where data center applications are taking an optical chip and packaging them together with an electrical chip.

These are the predominant applications that we're working on. Now again, the Systems business is -- we've been engaged in silicon photonics for several years in the systems business in R&D labs. Now moving to pilot production, we're still engaged.

Whether we have a play in full high-volume production is something that we're working with customers to evaluate right now. Probably requires a few changes in our road map, but a really interesting area where the combination of our electrical test and optical test products and technologies appear to have some pretty significant value for customers as they're ramping these co-packaged optics applications.

And just a clarification for me, and I'll stop. Is that (inaudible) folks are working on? Am I getting that correct?

Yes. Yes.

Got it, got it. And do you expect -- or you're seeing visibility into that going into production within the next few years?

Yes, early pilot right now and a lot of bugs to work out, but it is an exciting area, where we're working with customers on this fusion of optical and electrical technologies to help push the industry forward.

(Operator Instructions) And our next question comes from the line of (inaudible) from Jefferies.

So I actually wanted to touch upon some of the questions you had earlier regarding the chiplet architectures, right? So I understand you talked about that 20% to 25% increase in test intensity.

But for me, on a simplified way that I was looking at this was if you have four chiplets, you'll need four different probe cards. So why is it this a 20% to 25% increase versus like a 4x increase? And so I guess if you could help me understand that what are some of the strategies that are being used that your probe card requirement isn't as high?

Right. So the reason it's not 4x, yes, you need 4 different probe cards. But if you think about how customers are partitioning the end product, which used to be a single die into individual chiplets, you can imagine that the test coverage, the test complexity or the number of tests you need to do for each of those chiplets is substantially less than you need to do for the composite die, right?

If I'm going to test an entire microprocessor, the SRAM, the inputs, outputs, all of the buzzes, you could imagine -- that's a fairly long test time because you have to go through a broad suite of electrical tests to get all those things. You break it up into chiplets and now you're also breaking up the test coverage into the individual chiplets.

So yes, I need 4x the number of probe cards. But simply put, I'm also testing 1/4 of the transistors on each of these chiplets, so the test times are shorter. That's why it's not a 4x uplift, it's a 20% to 25% upward.

Got it. This is very helpful. And for my follow-up, so in 2022, you had market share losses to one of your #1 competitor in Foundry and Logic. So how should we think about this going into 2023? Are there any low-hanging fruits that help you kind of reverse this -- the market share?

Yes. I think -- so you've hit on an area that's a key area of focus for FormFactor right now. You're right, we did lose market share to our primary foundry and logic competitor for different reasons at different customers.

But if I take a holistic look at it, it involves a couple of different things that we're now very focused on. One is making sure that we're aggressively delivering new technologies for those customers and engaging them very early on. I think in some cases, we lost market share because our competitor was faster to deliver their technology.

I think one of the other things is -- if you look at some of the customers where on an industry-wide basis we have lower than benchmark or lower than entitlement market share, we're working pretty hard on gaining share at those as well. So there is a very good competitor, a very viable competitor and we're going to -- we're in an industry where you need 2 suppliers.

So I think us executing better, us being more aggressive on getting our leading technology into customers' hands are a key focus area for us now. We've made some organizational changes. We've made some incentive changes, and are very focused on making sure that we're a sharper and stronger competitor against these guys in Foundry and Logic.

Is pricing sort of an area where -- did that drive the market share gains? Is that (inaudible)?

No, not really. One of the things I think all of us understand in this business is there really are 2 supplier markets. And you're not going to be able to price your way to market share gains.

Customers depend on us for critical product ramps. We've talked about lead times being short. That means there's a real premium on execution, right? Our customers can't ramp if they don't have the probe cards in.

And the value of the technology that all of us provide us, our competitors, is pretty compelling. And so you're not going to win on price. And we, as a set of suppliers, sure, you've got to be cost competitive, but it's not about the lowest price winning.

Okay. That's helpful. And if I may squeeze in one more. So I know you talked about China being -- domestic China being 10 million to 15 million. So I think a bulk of that is DRAM. Have you -- could you sort of give some color on how much is systems portion of it? Or -- and is -- are you able to ship into China domestic customers for system revenues?

Yes. It's actually -- I wouldn't say the domestic China revenue is dominated by DRAM. DRAM is a good chunk of it. And given that the major China DRAM manufacturer, domestic China DRAM is really just ramping up, that can pretty be a pretty lumpy revenues quarter-to-quarter.

We do have pretty significant systems exposure and systems opportunity in China. Again, we've, by and large, found a way to -- at least what we're booking, understand whether it's going to be -- that we're going to be able to ship. So we understand the trade compliance implications.

So at least in the short term, again, these headwinds are baked into our Systems business into our outlook for the Systems business. Over the longer term, again, we do expect that business to continue to decline and eventually go to zero. The question is a question of when and when a local China supply chain is able to support its industry. That's probably a multiyear event.

This does conclude the question-and-answer session of today's program. I'd like to hand the program back to Mike Slessor for any further remarks.

I think we're out of time. So thank you, everybody, for your participation. Thanks for the questions. And we'll see you either at some upcoming conferences or on our late April, early May earnings call. Take care and stay safe.

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