FormFactor Inc (FORM) 2017 Q1 法說會逐字稿

Thank you, and welcome, everyone, to FormFactor's First Quarter 2017 Earnings Conference Call. On today's call are Chief Executive Officer, Mike Slessor; and Chief Financial Officer, Mike Ludwig. 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 security laws. Examples of such forward-looking statements include those with respect to the anticipated effects and benefits of the completed merger between FormFactor and Cascade Microtech, projections of financial and business performance, future macroeconomic conditions, business momentum, business seasonality, the anticipated demand for products, our future ability to produce and sell products, the development of future products and technologies; 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 2016 and 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, May 3, 2017, and we assume no obligation to update them.

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

Thank you, Jason. And thank you, everyone, for joining us today. FormFactor delivered a strong first quarter to open 2017, setting an all-time revenue record and posting earnings above the high end of our guidance range. Here in the second quarter, as indicated by our guidance, we're efficiently executing against even stronger demand for our products.

In addition, our visibility into the second half of the year has improved, making us more positive on 2017 overall than we were at our early February conference call. Despite the limited nature of that visibility, along with the likelihood that we will experience some degree of typical probe card seasonality, we now expect revenues for the second half to approach the revenue levels we're delivering in the first half.

As we have discussed previously, our acquisition of Cascade Microtech provides FormFactor with a broader customer and end-market footprint and the ability to capitalize on a set of diverse demand drivers. This diversification is paying off as we're experiencing positive momentum across the board in data center, mobile and automotive applications. And we expect these themes to continue into the future.

In recent weeks, you've heard both our customers and other suppliers discuss these end markets. And I'd like to describe how developments in each end market are favorably impacting both current results and future opportunities for FormFactor's broader product portfolio.

Here in the first half of 2017, FormFactor is experiencing strong demand for both probe cards and engineering systems to test the microprocessors and memory chips that are the heart of today's data centers. Despite some recent industry debate on end-market growth rates, the underlying factor, that's our large microprocessor customer double their business with FormFactor in 2016, remains solidly in place. This is a major contributor to our record first quarter Foundry and Logic probe card revenue. And we're adding capacity even as we operate at historically high shipment levels.

During the first quarter, we also experienced significant sequential growth in probe cards for server DRAM, primarily for DDR4 designs at the 20-nanometer node. We also delivered strong shipments of 300-millimeter engineering systems for both 10-nanometer yield improvement as well as 7-nanometer and 5-nanometer development.

Another source of current demand strength is the mobile sector, primarily with our Foundry and Logic and DRAM probe cards. Our volumes here have been driven by 3 factors: one, recent application processor design wins at all leading 10-nanometer foundries; two, accelerating midyear seasonal demand from the RF/BAW and SAW filter manufacturers; and three, ramping 18-nanometer and 20-nanometer mobile DRAM production. The themes of increased complexity and more wafer test content we have described before are especially evident today in Foundry and Logic probe cards for a major 10-nanometer mobile application processor project. Consistent with recent comments by one of the leading ATE suppliers, we are shipping a large numbers of probe cards to support the test flows associated with wafer-level fan-out packaging used in these devices. This volume is a result of longer test times needed to ensure high-quality individual devices prior to wafer-level fan-out packaging to avoid killing good devices by packaging the most effective devices.

We also continue to see growing demand for all our products and supporting the automotive end-market. With stringent test quality requirements and challenging test environments such as high temperature, automotive applications represent significant current business and a growth opportunity across our probe card and systems products. For example, our engineering systems business is benefiting from a customer need to collect large volumes of ultraprecise electrical performance data, while subjecting the devices to the extreme environment characteristic of drivetrains, a complex combination of factors where we have a significant competitive advantage.

In addition, we're excited about the opportunities for our Foundry and Logic RF probe card business, given new millimeter wave sensing and 5G communication requirements being contemplated to support autonomous driving.

Turning to a broader view, VLSIresearch recently released its 2016 survey of the advanced probe card market, which reported continuing strong growth paired with FormFactor's sustained leadership. They expect this $1 billion dollar market to grow at nearly 7% CAGR over the next 5 years. This growth is driven by increased wafer starts in new designs as well as by adoption of advanced packages that require increased wafer test content and longer test times, as we're experiencing in our Foundry and Logic business. These in turn require more probe cards. At the same time, the complexity and performance requirements for these probe cards continues to advance. This moves technical requirements towards FormFactor's differentiated strength, such as the use of MEMS technology for probe fabrication.

Lifted by these trends, in 2016, we increased our share to over 30% of the advanced probe card market. FormFactor now holds more than double the market share of our nearest competitor. We're very proud of our share gains, which came from both organic growth in existing products as well as the addition of 0.5 years of Cascade Microtech RF probe card revenues. The combination of our scale and technology leadership together with our strong multi-generation customer relationships will enable us to continue to outpace the market growth forecasted by VLSI, just as we've done in recent years.

As evident from our strong first quarter results, operational execution continues to be a very high priority for our team. At present, we are increasing capacity to meet our increased customer demand. We are doing so in a flexible and cost-effective way, as we anticipate some degree of historical calendar seasonality, which typically appears as fourth quarter demand softness for production probe cards. The combined FormFactor and Cascade Microtech teams are now focused on both efficiently delivering the record demand for our products as well as executing on the joint R&D projects we described on the last call. These projects are planned to fuel future growth with corresponding gross margin and EPS expansion as we continue to leverage our investments in fixed cost over a larger revenue base.

I'll now turn the call over to our CFO, Mike Ludwig, for further details on our first quarter results and to provide insights into FormFactor's outlook.

Thank you, Mike, and good afternoon. As you saw from our press release and heard from Mike's comments, our year is off to a strong start as we delivered record quarterly revenues in the first quarter and the highest non-GAAP earnings per share in almost a decade. Our financial performance was well above the midpoint for all measures with our revenues, earnings per share and free cash flow exceeding the high end of the first quarter guidance we provided.

For the first 3 quarters together, the combined business continued to stay focused on execution and integration, delivering revenues of $377 million, non-GAAP earnings per share of $0.66 and free cash flow totaling $47 million.

Turning back to the first quarter results. FormFactor's revenues were $128.8 million, up $4.3 million sequentially. Probe card segment revenues of $106.5 million increased $7.9 million or 8% compared to the fourth quarter, while Systems segment revenues decreased $3.6 million or $22.3 million in the quarter.

Within the probe card segment, Foundry and Logic revenues of $74.3 million increased $5.9 million or 9% compared to our fourth quarter resulting from continued strength in the data center, mobile and automotive applications. Foundry and Logic revenues comprised 58% of total company revenues in the first quarter. Year-end revenues were $29 million in the first quarter, an increase of 20% sequentially, as we saw the demand environment continue to strengthen throughout the first quarter, a continuation of the trend we experienced in the second half of 2016.

Technology node transitions, increased year-end content per handset and a stronger data center demand environment are all contributing to an increase in DRAM probe card revenues. DRAM revenues comprised 22% of total company revenues in the quarter. Flash revenues of $3.2 million for the first quarter decreased $2.8 million from the fourth quarter. As we have discussed previously, we expect Flash revenues to continue to be lumpy as our exposure to the market remains relatively concentrated.

The decrease in the Systems segment revenues was driven by the typical seasonality in this business where we saw system shipments decline relative to Q4. The mix of the Systems revenues, however, was favorable with a higher percentage of revenues coming from the 300-millimeter platform.

GAAP gross margin for the first quarter was $47.6 million or 36.9% of revenues compared to 32.5% of revenues for the fourth quarter. In the quarter, there were $7.3 million of reconciling items outlined in our GAAP to non-GAAP reconciliation available on the Investor Relations segment of our websites. On a non-GAAP basis, gross margin for the first quarter was $54.9 million or 42.6% of revenues, a slight increase from the 42.4% of revenues for the prior quarter.

In the probe card segment, Q1 non-GAAP gross margin was $42.9 million or 40.3% of revenues, slightly higher than the 39.8% of revenues in Q4. Solid execution in our factory and higher fixed cost absorption was able to overcome a challenging product mix to produce a slight increase in the margin for the first quarter.

In the Systems segment, the non-GAAP gross margin was $12.1 million or 54% of revenues in the quarter compared to 52.1% in Q4. The improved gross margin percentage in the first quarter was driven by the favorable product mix of 300-millimeter Systems revenues.

Our GAAP operating expenses were $40.5 million for the first quarter compared to $54.8 million for the fourth quarter. The first quarter includes $5.2 million of GAAP to non-GAAP reconciling items. Non-GAAP operating expenses for the first quarter were $35.4 million or 27.4% of revenues compared to Q4 expenses of $36.2 million or 29.1% of revenues. These results reflect our stated focus on leveraging our operating expense structure more effectively across the business. Non-GAAP operating expenses benefited from $1.8 million of realized synergies in sales, general and administrative expenses or $7.2 million annualized. We're on track to capture previously committed $10 million to $12 million of synergies in 18 to 24 months, principally from sales, general and administrative expenses.

The non-GAAP effective tax rate for the first quarter was 4.4% compared to 7.3% for the fourth quarter. Consistent with our expectations of a continued low effective tax rate while we utilize our U.S.-based net operating losses and continue to carry evaluation allowance against our deferred tax assets. The lower effective tax rate in the first quarter reflects the anticipated higher mix of U.S. -generated pretax income in fiscal 2017. We expect our non-GAAP effective tax rate for 2017 to be between 4% and 6%.

GAAP net income was $5.2 million or $0.07 per fully diluted share for the first quarter compared to a loss of $15.4 million or $0.22 per share for the fourth quarter. Given the strong revenue outlook for 2017, provided by Mike in his comments, we expect fiscal 2017 to be GAAP profitable as well.

First quarter non-GAAP net income was $17.3 million or $0.24 per fully diluted share compared to $14.4 million or $0.20 per fully diluted share for Q4, reflecting higher revenues, good factory execution and disciplined OpEx spending in the first quarter.

Company noncash expenses for the first quarter included $8.5 million for the amortization of intangible assets, $3.3 million for stock-based compensation and depreciation of $3.2 million.

Moving on to the balance sheet. The company generated $15.5 million of free cash flow in the first quarter, another strong quarter of cash generation. In the first quarter, the company spent $7.9 million on principal and interest payments, including a $5 million accelerated principal payment; repurchased shares for $2.7 million; and spent $3.5 million for capital expenditures.

Consistent with our February earnings call, we expect 2017 annual CapEx expenditures to be in the range of $16 million to $20 million. In total, cash comprised of cash, short-term investments and restricted cash ended the first quarter at $123.1 million, $13 million higher than Q4.

The balance of our bank term loan was $131.5 million at April 1, a decrease of $6.7 million from December 31. With another solid quarter of free cash flow expected in the second quarter, we anticipate being net cash positive at the end of Q2.

Turning to the second quarter non-GAAP guidance. We continue to experience strong broad-based demand momentum in our probe segment, both in our Foundry and Logic applications and in our DRAM applications. The demand in our Systems segment remains strong at both the 200-millimeter and 300-millimeter systems. As such, we expect our revenues for the second quarter to grow sequentially and be in the range of $130 million to $138 million.

We anticipate FormFactor's gross margin to expand from our Q1 margin due to a slightly better product mix and improved factory efficiencies from higher production levels. Therefore, we expect the gross margin to be in the range of 42% to 45%. We expect to realize fully diluted earnings in the range of $0.24 to $0.30 per share. Our Q2 non-GAAP guidance assumes consistent foreign currency rates.

Having completed 3 quarters as a combined company, we have a better view of the opportunity to grow the business and leverage technology investments, operational investments and the SG&A operating structure across the combined company. As such, we have made significant progress in developing a long-term financial model. We plan to hold a conference call to review our go-forward strategic focus areas and review our financial model on June 28. We hope you'll be able to participate in our conference call.

With that, let's open the call to questions. Operator?

(Operator Instructions) The first question comes from Patrick Ho with Stifel.

Mike, first off, in some of the strength that you're seeing on the DRAM probe card side, would you say it was based on several customers? Or was it concentrated primarily with maybe one customer?

Thanks, Patrick. The question on DRAM probe card customer concentration, it was certainly broader than a single customer. We did see in the first quarter a significant strength associated with a node transition at a single customer. But even in the first quarter and through here in the second quarter, we're seeing broader customer involvement in, call it, a DRAM recovery or DRAM production ramp than just a single customer. So I characterize it maybe not quite as DRAM industry-wide, but more than just a single customer driving our DRAM probe card demand.

Great. That's helpful. And maybe looking at your systems business, I think one of the key things you talked about in terms of the Cascade Microtech deal itself was some of the leverage and some of the synergies you'd be getting the cross-selling opportunities. I know it's only a couple of quarters since you closed the deal. But have you seen any of that type of leverage yet to date in terms of whether it's helped your probe card sales or whether it's helped your Systems sales or the leverage between the 2 businesses?

I think the revenue synergies are the cross-selling opportunities between our probe card business and our Systems business. We've seen maybe some opportunities associated with that. They're certainly not material to our current financial results. And maybe providing a little more color on the opportunities, they're outside of our mainstream semiconductor businesses. They're in things like silicon photonics, maybe micro LED testing, scenarios that are really new areas for us as a company and markets and applications that are maybe less well defined than mainstream semiconductor test. We have an opportunity to bring our products together, offer a solution. And we're optimistic that -- not going to be a 2017 event, but we're optimistic in future years that we're going to be able to show some revenue leverage associated with that.

Great. And final question from me for Mike Ludwig in terms of the cash flow generation. Obviously, you had a very strong quarter this year -- I mean, this quarter. And it looks like, based on the revenue levels, you'll have a pretty good cash flow generation for the year. Is it fair for me to characterize that a lot of it will be paying down the debt and maybe do some of these accelerated principal payments that you did in the first quarter?

Yes, that's our plan, Patrick that -- what we'd like to do is, I would say, use about half of our -- half of the excess free cash flow, so after we pay down our debt, our regularly scheduled debt payment. We then take half of our free cash flow above that and pay down additional principal payments. But really sort of pay it down to the level where the unhedged level of the principal balance, right -- we actually have about $95 million, $96 million of our balance -- of our principal balance is hedged, has interest rate hedge on it. So it's at a fixed rate -- at a pretty good fixed rate. But we still have some at floating. And the floating part of that, we definitely want to pay down on an accelerated basis.

Our next question comes from Craig Ellis with B. Riley.

I'll start with one for Mike Slessor. Mike, just clarifying a couple of things. One, more backward looking, the revenues in the quarter were above the high end of guidance. Where, on a segment basis, good things turned out better than you had expected? And even before the very strong (inaudible) guide, a statement that the second half could be as strong as the first half would have been quite significant. But now it's off a much higher base than we were looking for previously. Where across your end markets are you more confident in the second half? And if there is risk to the second half, where would it be?

Thanks, Craig. And I just so want to clarify, we didn't say that the second half was going to be stronger than the first half. We did say that based on what we delivered in the first quarter and the revenue guidance range we've provided for the second quarter, we see the second half approaching that. So I just want to be clear that we don't currently see the second half higher. But that goes to some of the fundamental reasons, why our view of 2017 has improved. The first of which is we're 3 months closer to the second half, and we have a much better view of what some of our key customers and the demand dynamics that are going to drive our third quarter and fourth quarter revenue. I'd say, from an overall upside theme, there's probably 3 elements to this. The first is our Foundry and Logic probe card business. It is continuing to show strong growth around both existing customers doing more wafer tests and therefore requiring more probe cards for it. Obviously, this was a theme that drove our 2016 results. But we're also seeing strong accelerating demand from the RF market here as we go through the middle part of the year. We're also more optimistic about the sustainability of the DRAM investment cycle this time. Obviously, pretty solid first quarter DRAM probe card results, but we see that having some lag, certainly into, if not through the second half. And then finally, our Systems business has a seasonality profile, which tends to increase throughout the year, which helps us -- obviously still at 20% piece of our business, but that seasonality dynamic helps us out with pulling up the back end of the year on a pure seasonal basis. So I hope that answers the question.

Yes, that's helpful. And maybe as a follow-up to revenues before I move on to gross margin. As we look at the second quarter, can talk about some of the gives and takes as you look at the business. Headline is moving up nicely by about $5 million quarter-on-quarter. Where do you see gains, and are there any of your market segments where we should expect to see business decline quarter-on-quarter?

I think of our major businesses, we see the same themes I talked about in answering the questions about the whole year sort of building through the second quarter. Foundry and Logic continues to, I'll use the word, surprise, but pleasantly surprise us to the upside around the themes we talked about, both customer wins and more wafer starts and more probe cards. We expect DRAM to be comparable in the second quarter, again, a data point associated with the sustainability I talked about. We might see -- depending on different acceptance and revenue recognition, we might see Systems down a little bit sequentially in the quarter. But the overall demand there continues to be quite strong associated with leading-edge node development and the yield improvement activities we talked about.

And then moving on to gross margins and perhaps more for Mike Ludwig. It's nice to see gross margin guidance range that's got 45% on the high end. The question is, with the range that's 300 basis points from 42% to 45%, can you walk us through some of the risks, both on the positive and the negative side? For example, comments that the company is adding capacity, is there risk that capacity could be added ahead of demand and dilute gross margins? And alternatively, as we look at the gross margin range, if the company were to come in, say, at the high end of the range, 44% to 45%, what would be the factors that would lead to that?

Yes, I think when we look at the gross margin, Craig, I think there are a couple of things that we want to think about. So one is, with increasing revenues, we generally have better factory utilization. Better factory utilization generally leads to higher gross margins, right? But all of that -- the caveat around all of that really is also the product mix. So I think if we were going to have a -- if there is a risk on the downside, I don't see it so much with respect to putting capacity in place prior to really getting orders from the customers. I think as Mike said, we're -- I think we've learned a lot over the last 12 to 18 months with respect to increasing capacity and customer demand and whatnot. So I don't see that so much of the risk. But the risk again we do see could be along the lines of product mix and standard margins associated with that. So again, I think in general, there's upside to the margin with respect to increasing factory utilization gives us, again, higher lift on margins. And if you can combine that with a more favorable product mix, then sure -- then I see that we could get toward the higher end of that range. The risk, I think, is more on the product mix on the downside, but again, not so much on how we plan our capacity.

And last one before I get back in the queue, bouncing it back to Mike Slessor. Mike, it was a quarter ago when you identified that you would be integrating R&D for the Form business and the newly -- recently acquired Cascade business, can you give us an update on how that's going, and what the milestones are that we should be focused on as we move through 2017 and into 2018?

Sure. Thanks, Craig. So we continue to operate in a business unit structure where the old Cascade probe's business is in its own business unit with its own R&D, same for the old FormFactor business. What we have done is decided to fund 2 initiatives here in 2017 associated with technologies from each of those business units producing a new product or new subcomponent for the FormFactor overall road map. One milestone or update I'll give you on that is a product to address an area of the RF test space that we do not currently serve is progressing nicely, being run by a joint old FormFactor and old Cascade team. And we expect to shift the first articles of that combined product to a customer in the third quarter.

Our next question comes from Edwin Mok with Needham.

First question, just a clarification. You mentioned that you're adding capacity. It's not specifically for the microprocessor customer, or is that generally adding capacity on your end?

I'm sorry, can you repeat that, Edwin?

So you talked about adding capacity, right? Is that specifically for the microprocessing customer, or are you talking about adding capacity for your MEMS (inaudible) or making a probe card?

We're adding capacity in a variety of places. I would say, generally, it's around the theme of stronger Foundry and Logic probe card business than we originally started the year with. I wouldn't classify it as being specific to any one customer in that Foundry and Logic space. I'd say that overall business is stronger than we planned to start the year with. The other piece of it, though, as we brought together product roadmaps and started to leverage our manufacturing assets, some of the additions we're making that maybe are associated with Foundry and Logic demand, we're also able to flex over into memory and DRAM in particular. So the capacity additions are not quite as specific as they might have been several years ago.

Okay. That's helpful. We have heard some people in the supply chain talk about kind of how the Chinese handset market is going for inventory correction. But looks like the RF business is doing pretty well. Do you think that as potentially a risk for the RF business as you go through second quarter and beyond?

So our RF probe card business has typically seen seasonal growth with stronger in the second quarter and third quarter, mostly associated with new BAW and SAW filters that are going into the late year-end tech releases. We see that theme continuing to play out this year. And whether it's demand signals from customers, our current backlog or our view of RF business and its continued growth here in 2017, we don't -- we hear the noise around a lot of the China handset demand. But to be quite honest, we're not seeing it in any of the ways our customers are behaving and the capacity and orders our customers are putting in place so that they can produce to this [filter] demand for the next-generation handsets.

Great, that's helpful. Lastly, just on the auto commentary. Actually, thanks for providing color around that. We obviously got questions around that as well. Just curious, is there a way to kind of break down and think about business? How much of the exposure do you have on auto right now? And [it sounds like some of the new]product you're developing is going to be -- is going to have high exposure for that market? So do you see that mix increasing as we go through -- over, let's say, the next few years?

Yes, I think we see automotive as an end driver increasing for us for a variety of reasons. First of all, anybody's semiconductor forecast for end units associated with automotive shows some pretty healthy growth rates associated with that. And given that probe cards and consumables essentially are driven by wafer starts and design starts, we see some upward momentum there. As I mentioned in the prepared remarks, there is also the complexity associated with automotive test. Obviously, anything goes into a car has a much higher quality bar than something that goes into your cell phone or your laptop. And that drives things towards FormFactor's historical strengths: our technology leadership, our ability to do a very accurate electrical test at extreme conditions, things like higher temperatures. So we're seeing more demand for automotive, and we're seeing that demand be better matched to our differentiated strength. That's one of the reasons why we're optimistic. It's one of the areas that when we do our analyst call on June 28, we're going to try and tie together some of these end-market drivers for you and show how that translates into some of the opportunities for FormFactor's probe card and engineering systems businesses going forward.

Great. We're definitely looking forward to that. But just to kind of touch on the engineering system. Do you see a scenario where you're -- historically, engineering systems is mostly for development and R&D lab and not so much for production, right? But like given the different (inaudible) on auto, do you see that potentially create an opportunity for you to start selling some of these systems into the more production environment for this higher requirement in auto?

Yes, maybe not the entire system, but certainly some of the subcomponents associated with our engineering systems are already beginning to serve these requirements. One example, again, is temperature. Some of the subcomponents that we use in our engineering systems are also applicable to production test. Maybe not the whole system, because as you say, we've kind of optimized it for the flexibility and precision required for R&D, but some of the subcomponents do transfer over. So we're participating in some interesting ways in the subcomponent supply chain for production automotive test.

Our next question comes from Jagadish Iyer with Summit Redstone.

A couple of questions. This is for Mike Slessor. So in the apps processor, you talked about the fan-out packaging. I was just wondering what kind of market share do you have with this particular customer. And can it help you with other customers who want to embrace this packaging technology?

Yes, Jagadish, good question. So we're exposed to this 10-nanometer application processor project at this foundry really for the first time in FormFactor history. And as a result, it's really some share of a single design at 10-nanometer at this customer. So if I were to characterize our share overall, at this customer, it's going to be single digits. As you probably know, we've made it a strategic effort to work with all of the leading customers in the world and finally breaking down the door and supplying significant production volumes, even if only for one design on one node, represents a big step forward for us. I'd anticipate that we'd be able to, again, show our value, show the value of FormFactor's technology leadership, our performance and our customer support and be able to grow our share at this customer into the kind of share positions we have across the broader market. That will be a multiyear effort, however.

Fair enough. Then the second question is what is your level of engagement with customers in China, who are planning to start up memory business next year? Are there any competitive threats there?

Yes, I think, China for us represents a really interesting opportunity. As you may know, FormFactor has had a pretty strong the China footprint for the better part of a decade. We started there some local manufacturing. We also, last year, if you recall, opened up quite a large facility in the Hsinchu Science Park and have been investing in the region pretty broadly. As we see some of the memory fabs start to produce first silicon and articulate their plan for ramping in 2018, we see a really robust opportunity for FormFactor. As the world leader in memory wafer test, I think we're recognized, given our China footprint and historical China footprint and having the technology that's supported by a very strong local team there, we see the ability to compete pretty effectively as significant wafer starts start to flow through this facility. Now it's still -- right now, if we're assessing things, it's still early days. A lot of these things are just shells. Initial equipment deliveries have obviously started. But there is a lot of cards to be played before we see significant probe card volume associated with China. I think we're in a good position, and we view it as a large opportunity. It'll happen sooner or later.

Okay. And this is my last question is for Mike Ludwig. How should we think about the OpEx as you focus on investment priorities while balancing synergies from your Cascade acquisition?

Yes. So obtaining leverage on our operating structure continues to be a focus of our integration and execution. And of course, we're pleased to have achieved an OpEx spending of $27.4 million in Q1. But when we balance our growth requirements with our remaining synergies to capture, we anticipate that we'll be managing our OpEx expenses to the previously committed range of $35 million to $37 million per quarter, certainly in 2017. So I think that's where we expect to end up. That translates to about 27% of revenues. And I think that's -- from our perspective, I think that's really sort of our goal as we go further into 2017 is to really kind of keep an eye on that. But again, I think the number $35 million to $37 million is really what we expect to -- how we expect to manage that. And by the way, we think that does allow us to achieve our growth requirements.

Our next question comes from Tom Diffely with D.A. Davidson.

Just following up on that last question then. Is some of the COGS at $35 million to $37 million, does that include any additions for this capacity add, or is most of that in the COGS line?

Yes, most of that, Tom, is going to be in the COGS line in terms of -- when we were referencing capacity, we were really talking about our manufacturing structure.

Okay. Great. And then when you look at the new 10-nanometer apps processor customer, is that for full probe cards at this point, or is that just probe components?

No. Tom, great question. As we've said historically, the business model for some customer -- and customers and that customer in particular is a little bit different. We are selling subcomponents to them that they integrate together into a probe card for a variety of both historical and operational reasons. So it is subcomponents.

Okay. And I assume that's probably a pretty good margin business for you?

Yes, I think we're being competitive for the technology and value that we're providing.

Okay. Moving over to the automotive section, just one more question there. Are you using your standard probes inside of these burn-in systems, or is it a specially designed probe just for that?

Well, it's not entirely burn-in. So obviously, higher temperatures are often associated with burn-in, but we have customers who are trying to test and characterize their devices for short periods of time at high temperatures as well. So we see some applications where our standard probes work just fine. There are other applications where we have had to do some development. And we continue to do development, because we see this as a trend that will continue as automotive standards move to slightly higher temperatures but become more and more stringent with the quality requirement as cards become maybe not fully autonomous, but more autonomous, and the centers have to be much more rigorously tested.

Okay. That makes sense. And then moving over to some of your earlier comments. You did mention that maybe it's just a little too soon to know exactly what the fourth quarter looks like, and there is some often a seasonal downtick. Is the biggest risk to the fourth quarter on the DRAM side?

I think, given the historical volatility we've seen in DRAM, I think that's a fair assessment. It's certainly beyond the -- the fourth quarter in total is beyond the visibility we have right now and I would argue beyond the visibility our customers have right now. So I think that is probably the key risk associated with the fourth quarter. And when you add on top of that, because seasonal swings we've seen associated with fourth quarter softness have been coupled to DRAM, I think it's fair to assess that as the biggest late-in-the-year risk to the top line.

Okay. And then finally, when you look at the Flash business, is the -- obviously, it's very lumpy. But is the $6 million last quarter and the $3 million this quarter more of a representative run rate for that business?

Well, I think given the lumpiness because of our concentration at a few customers with a few designs, it's hard to say what the run rate is at this point. I think given our current loading and situation and where we're engaged in supporting both Foundry and Logic and DRAM, we want to make sure that we are judicious about the Flash designs we're competing for. Obviously, long term, given 3D NAND wafer starts, we view this as a secular growth trend we want to grow into. But we're also managing in the short term to make sure our mix and our factory loading is where want it to be in the very short term. So if you made me answer, I'd probably tell you somewhere in between the $3 million and $6 million is the right run rate.

All right. And then, I guess, following up on that. Is there a difference in probes going from Gen 2, 3 and 4 for the 3D NAND products?

So that's one of the interesting places where we have carved out a bit of a niche is as big entities go up because of the stacking, at least for the smaller capacity chips, things like 128 gig, you have a smaller die size, which then results in things being pad-size limited, which moves things back to FormFactor's strength for a MEMS probe. Those are the places really where we're participating. And how that evolves as layer counts continue to increase is something that we're looking at pretty closely and something that could represent a big opportunity for things to move back towards FormFactor's strength.

Okay. And actually just one more here. When you -- some of the tester guys are saying that they're seeing a slowing down of the increased parallelism of their testing? (inaudible) the growth of parallel testing is slowing over time. Does that impact your business either way?

Well, so that one is an interesting one, because it's very segment specific. I think in memory in particular, we're seeing no abatement associated with parallelism, right? Essentially, customers want to test every die on the chip at once. And as their densities goes up and die counts increase, that trend continues. So parallelism is going up. There's some interesting dynamics in the Foundry and Logic business. Certainly, for high-end processors, you can imagine on leading-edge nodes where yields are relatively low. The testing with high parallelism maybe isn't the best economic choice. And so without getting into all the details, there are pockets of the market, at least at this point in time in yield and device ramps, where we are seeing parallelism back off a little bit, primarily at the high end of Foundry and Logic.

Okay. But does that have an impact on your business, whether or not it continues down...

I don't think so. I mean, the first order, we sell more probe cards to test the same number of devices. And so we're not really seeing an impact associated with those puts and takes.

Our next question comes from Ed Roesch with Sidoti.

Lot of good information already, but I did want to ask a few questions. First, did capacity constrain you at all on your Q1 shipments?

I think if you look at how our demand and then shipment commitment cycle works, I don't think we were fundamentally constrained by capacity in the first quarter. The -- our comments associated with adding incremental capacity in a somewhat guarded way here, are associated with the increased demand. And I think that's consistent with how you see our revenue stepping up from what was delivered in Q1 to the range we've guided to in Q2. We are seeing some nominal increasing strengths in a few businesses. And so we're adding capacity there.

And then when I look at your revenues by geography, the Asia-Pac region was really stellar for you in Q1 moving from Q4. I was wondering if you could talk about some of the dynamics there. I know you do have a new customer relationship, but if you could speak to some of the drivers, that would be great.

Yes. And obviously, Asia Pacific is an awfully big region with an awful lot of different puts and takes in it. I do think maybe a couple of themes. Obviously, the foundry business has its epicenter in Taiwan, and that's an area that we've obviously had a nice initial customer win. But we also see that business being strong compared to historical levels. The other piece is that some of our memory customers, one in particular, has significant operations in Taiwan as well. And so that's helping results as well, at least on a sequential quarter basis.

And last one, maybe bit of a stretch. But listened to one of your customers discuss kind of a new way of stitching together different die, multiple die, and interconnects via bridge. And I was just wondering if that's a challenge in design probe cards that landed on FormFactor's plate or is in your radar at this point?

It's most definitely in our radar and is one of the more specific drivers that lead us to talk about advanced packaging driving our business. For us, that kind of thing where you're taking multiple constituent die and then putting them together on some sort of bridge offers us a couple of different kicks at the can. Obviously, the quality requirements associated with the individual constituent die go up, because you want to make sure, as I said in the prepared remarks, that you're not killing good die with bad die. So you want to make sure that each of the components is good. It also offers us an opportunity, somewhat interestingly, to test the bridge itself, because you don't want the bridge taking down the known good constituent die. So definitely on our radar screen, an area of development and R&D spend for us. Not really material to our Q1 revenues, at least in testing the bridge part. But some of those dynamics I think associated with both the industry moving closer towards known good die and offering us another test insertion are opportunities for us and opportunities that we're focused on and investing R&D dollars in.

Our next question comes from David Duley with Steelhead.

Just a couple of follow-ups on things that people have been asking about. As far as your sequential improvement in the DRAM market in the March quarter, was that mainly driven by 18-nanometer or 20-nanometer conversions?

There was certainly a strong component of 18-nanometer designs in the first quarter result. However, there was also a good chunk of 20-nanometer, especially server DRAM, as I talked about in the prepared remarks. So DRAM overall quite a bit stronger for us, a healthy chunk of 18-nanometer for the first time, but a reasonable mix of 20-nanometer as well. Obviously, the whole industry is not moving to 20 -- sorry, the whole industry is not moving to 18-nanometer all in the first quarter. There is different customers placing different technology node transitions at different timings and priorities.

And as far as the June quarter goes, did you mention that DRAM business is going to be flat sequentially? And I was just wondering what the mixture of business would look like at 18 and 20 nanometer in that quarter? Would you have another 18-nanometer customer or it'd just be a single customer at that node?

Yes, I think in the answer to a previous question, I may have said that DRAM would be comparable Q1 to Q2, will have some pretty strong levels in Q1. I think we see 18 nanometer as still being largely a single-customer event in the second quarter and perhaps even longer as well. Having said that, the other 2 major DRAM manufacturers are still making significant manufacturing investments and releasing new designs even on the 20-nanometer node. And as you well know, new designs mean new probe cards. And so honestly it's six of one, half-dozen for us if they're are sitting on the 20-nanometer node producing new designs or shrinking to the 18-nanometer node.

Okay. And then, switch gears to 10-nanometer foundry business. It sounds like it could be quite an exciting piece of business. Did you actually have revenue with the foundry customer at 10 nanometers in the March quarter? Or is that going to be more June and second half oriented?

We did have revenues associated with that customer in the first quarter, in the March quarter. We have planning significantly more revenues in the second quarter and into the third quarter. Given, again there is a relatively concentrated opportunity at this point for us, one node, one design, we have some work to do for it to really continue to propagate into the late second half.

Well, good thing about one node, one design, it's an awful big design.

Yes, you are right.

Perhaps the biggest out there. Anyway, do you have an idea about how you're splitting that business with your other advanced probe card competitor?

Yes. I think it's probably a fair estimate that we're in second place. They had been in there before us. They had had historical relationship that went back maybe one node on a couple of other designs. And as I think you know, in these businesses, your customer relationship and incumbency is an important factor, given the lead time pressures, given the various service and support elements of the business, it's important to be the incumbent. And so although it's nice business for us, we're fortunate that it's a big design and there is plenty to go around for all of us.

Final thing from me. Again, I guess back to this 10-nanometer foundry win. You have mentioned that, that's, I think, produced with a fan-out packaging scheme.

Yes.

How -- and that [fab] requires longer test times. I was wondering if you might be able to characterize how much that would increase probe card demand? If the test times are longer and you consume more probe cards, is it like 10% or 15% kind of more probe intensive? Or is there some sort of metric you can help us understand how fan-out impacts the probe cards?

Yes, so it's a great question, but a bit of a moving target unfortunately, because as you can imagine, one of the key inputs is yield. So if -- the reason fan-out requires longer test times is because you want to make sure that or our customers want to make sure that they're not taking bad die and packaging them together with good die. That drives more test content at the wafer level. And with 10-nanometer yields relatively low right now, that's driving a really big increase in test times, and therefore, the number of probe cards. I have to assume that over time, yield improves and that delta, that incremental test time goes down a little bit. But for right now, it's significant. It's much more than 10% or 15%.

And I'm not showing any further questions at this time. I would like to turn the call back over to management for closing remarks.

Great. Thank you for joining us today, and thank you to the worldwide FormFactor team for delivering strong results to start 2017 with positive momentum building for a strong year. We're looking forward to meeting with you at several upcoming investor conferences and hope that you'll join us on June 28 for our analyst call.

Ladies and gentlemen, this does conclude today's presentation. You may now disconnect, and have a wonderful day.