FormFactor Inc (FORM) 2016 Q1 法說會逐字稿

Thank you, and welcome, everyone, to FormFactor's first-quarter 2016 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's discussion contains forward-looking statements within the meaning of the Federal Securities laws. Examples of such forward-looking statements include those with respect to the anticipated timing, completion, effects, and benefits of the proposed merger between FormFactor and Cascade Microtech. Projections of financial and business performance, 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 2015 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, April 27th, 2016, and we assume no obligation to update them.

Today's discussion will also include some important non-GAAP financial measures which are intended to supplement your understanding of the Company's financial results. 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.

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

Thank you, Jason. Earlier today, we released first-quarter 2016 results in line with the updated guidance we provided on March 29th. As we have stated, our Q1 guidance and results were adversely impacted by two factors. The first factor was our delayed capacity ramp to meet customer-specific SoC probe card demand. And the second factor was the push out of DRAM probe card deliveries by certain customers.

As we will describe in more detail, both of these issues have been resolved. While we are not pleased with our performance in Q1, we dealt with these issues swiftly, and they are now behind us. As you will hear later in the call, we anticipate significantly improved results in Q2 and beyond.

We want to give you more specifics on what went wrong on our SoC probe card capacity ramp and how we resolved the issue. First, this was a highly complex transition that required doubling of capacity over a three-month timeframe to serve expanding customer demand on multiple nodes and products.

The root cause of our execution miss was that we didn't hire, train, and qualify key direct labor and manufacturing engineering talent fast enough to meet the required timeframe. And we were therefor unable to achieve the required output in the originally planned time period.

Through a combination of headcount additions and redeployments from our memory product lines, we have now successfully increased productive headcount by more than 25% in the SoC product line, and are running these products at the targeted capacity for both 14 nanometer and 10 nanometer technology nodes.

We still have some ramp-up costs to absorb in the second quarter, but have now matched our capacity with the sustained increased customer demand. We have strong technology leadership at the very challenging 10 nanometer node, and the customer relationships remained very solid.

We ended Q1 with the highest every backlog in this business. We are not aware of any lost orders to competitors. And we continue to receive purchase orders that fully utilize the increased capacity. So while we did not initially execute this complex transition effectively, we are now consistently delivering at a run rate that supports a doubling of this business in 2016.

In addition to the growth in this microprocessor-centered business, we are also beginning to deliver probe cards to support the typical mid-year mobile application processor and modem release cycles, and are gaining further momentum with our market-leading non-memory probe card products and technologies.

Moving to the DRAM customer shipment push-outs, all products that were pushed out in March have already shipped and we have experienced no customer cancellations. We see DRAM probe card market conditions gradually improving consistent with our view of a stronger second half.

We are encouraged by our ongoing interactions with all three major DRAM manufacturers who are at various stages in implementing plans to make measured investments in both leading-edge capacity and new designs. Our worldwide team continues to work closely with all three customers on new designs and new capacity additions that are planned for production in the second half of 2016.

We'd now like to look more broadly at FormFactor's market position and strategy. As we announced earlier today, the VLSIresearch's annual survey on the probe card market reported that FormFactor again led this growing $1.4 billion market in 2015.

We also expanded out leadership position by gaining market share and growing faster than our served market. Compared to FormFactor's historical dependance on the DRAM industry, these results demonstrate our ongoing repositioning of the Company to address multiple components of growth and diversification in both the memory and non-memory probe card markets.

This results in an expanded opportunity and revenue base from which to operate the business. And we expect to continue to outpace the market in 2016 and grow FormFactor's revenues for the fourth consecutive year.

We'd now like to turn to the Cascade Microtech acquisition where we are continuing to execute on this theme of growth and diversification. We are combining market leaders across complementary semiconductor test and measurement applications, creating significant scale and customer diversification, as well as expanding our served market by approximately $400 million.

In addition to the significant earnings per share accretion we have discussed in the past, one of the many things that excites us about the Cascade Microtech opportunity is that they too have been outpacing market growth in the markets they serve. For example, in 2015 the Cascade probe card business grew by more than 20%, fueled by rapid adoption of their leading RF products and capabilities. As announced yesterday, Cascade delivered strong first-quarter 2016 results and provided second-quarter guidance for double-digit quarterly revenue and earnings growth.

The deal continues to be on track to close mid-year, and we have now received HSR clearance and filed a registration statement with the SEC. We are in the midst of detailed integration planning with working groups staffed from both companies. We remain confident we can achieve our initial target of $10 million to $12 million of operational synergies within 18 to 24 months.

Our integration approach is to move quickly to integrate areas where there is little to no impact on our ability to deliver revenue and profitability while moving more cautiously on areas that directly influence operational business performance. Accordingly, we are not planning any near-term manufacturing facility integration to ensure the core systems and probes businesses continue to deliver the strong profitability and growth exhibited in Cascade's Q1 results and Q2 guidance.

In summary, while we did not deliver as initially planned in the first quarter, FormFactor has started the second quarter with these execution issues squarely in the rear-view mirror and a transformative acquisition in front of us.

I'll now turn the call over to our CFO, Mike Ludwig, for further details on our numbers.

Thank you, Mike, and good afternoon. As Mike commented, the financial performance of the Company in Q1 was disappointing but in line with our advised guidance. Revenues for Q1 2016 of $53.6 million decreased $18.2 million, or 25% compared to our fourth quarter of 2015. The decreased revenues were driven by a measurable decrease in our DRAM revenues as well as a decrease in our SoC revenues.

SoC revenues in Q1 of $36.1 million decreased $5.3 million, or 13% compared to our fourth quarter. SoC revenues comprised 67% of our revenues in the first quarter compared to 58% in Q4.

Mobile processor revenues decreased in the quarter as our customers transitioned to new designs. We did, however, achieve slightly higher microprocessor revenue from a significant customer. With record backlog for this customer and continued order flow, we expect to completely utilize the doubled capacity now in place through the remainder or 2016.

First quarter revenues for DRAM products were $16.3 million, a decrease of 42% compared to the fourth quarter. DRAM revenues comprised 30% of our revenues in the quarter compared to 39% in Q4. As highlighted in recent customer's earning calls, the DRAM demand environment was weak in the quarter.

Flash revenues were $1.2 million for the first quarter, a decrease of $1 million from the fourth quarter.

First quarter GAAP gross margin was $9.8 million, or 18.3% of revenues compared to $21.2 million or 29.5% of revenues for the fourth quarter. On a non-GAAP basis, gross margin for the first quarter was $12.5 million, or 23.3% of revenues compared to $24.6 million or 34.3% of revenues for the fourth quarter.

The decrease in the first-quarter non-GAAP gross margin was due primarily to significantly lower DRAM revenues resulting in lower absorbed fixed overhead and the cost in investments incurred to address the capacity ramp for the significant SoC opportunity.

The cost incurred to double our capacity including additional direct labor resources, training cost, manufacturing and efficiencies and material scrap resulted in a reduction of gross margin in the first quarter of approximately 4%.

Our GAAP operating expenses were $23.4 million for the first quarter, an increase of $1.4 million compared to Q4. The increase is due to $2 million of acquisition-related expenses recorded in the quarter. Non-GAAP operating expenses for the first quarter were $18.6 million, a decrease of $0.3 million compared to the fourth quarter.

Variable compensation costs were not earned or accrued due to the loss in the quarter.

GAAP net loss was $13.8 million, or $0.24 per share for the first quarter, compared to a GAAP net loss of $0.6 million or $0.01 per share for Q4. Non-GAAP net loss was $6.3 million, or $0.11 per share for Q1 compared to net income of $5.8 million, or $0.10 per fully diluted share for Q4.

Turning to the balance sheet, cash comprised of cash, short-term investments, and restricted cash ended the first quarter at $187.6 million, $0.4 million lower than Q4.

Here are some other financial details for the first quarter. Our depreciation and amortization expense for the quarter was $5.2 million, including $2.4 million for depreciation and $2.8 million for amortization of intangible assets.

Our stock-based compensation expense for the quarter was $2.7 million. Our capital expenditures were $1.1 million. Excluding the impact of the Cascade acquisition, the Company expects to spend $8 million to $10 million on capital expenditures in 2016.

Turning to the outlook for the second quarter, we see strong SoC revenues resulting from our improved execution of our capacity ramp for a significant customer, and the beginning of the typical mid-year mobile application processor and modem release cycles.

The DRAM environment remains soft, but we see market conditions gradually improving consistent with the stronger second half. As such, we expect second-quarter revenues to be in the range of $76 million to $80 million, a tightening of our revenue range around the top end of our previously provided second-quarter guidance.

While we expect to achieve greater efficiency from the investments made to double our capacity, we still expect to experience an approximate 1% impact on gross margin from manufacturing ramp cost in the second quarter. Therefor, we expect the non-GAAP gross margin to be in the range of 34% to 37%, with non-GAAP fully diluted earnings in the range of $0.10 to $0.14 per share. We expect cash flow will be positive $4 million to $6 million. Our guidance excludes any impact from the Cascade Microtech acquisition.

For details on Cascade Microtech's first-quarter financial results and their second-quarter guidance, please see their press release dated April 26th, 2016, which can be found on Cascade Microtech's website at www.cascademicrotech.com.

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

Thank you. (Operator instructions)

Arthur Su, Needham & Company.

Hi guys, this is Arthur in for Edwin. Just congrats on the great guidance and the good rebound in the outlook. First question is on the logic segment, on the SoC segment, really glad to hear that the capacity ramp is going to translate to great business from this one customer for the rest of 2016. But beyond that, is there any commentary that you can provide on how copper pillar adoption has been with other customers and if you're seeing demand from other logic customers in the segment?

Yes, this is Mike Slessor. So when we look at what's driven FormFactor's leadership position in the SoC or non-memory probe card segment, a lot of it, as you indicate, has been around our customers adoption of copper pillar packaging technologies. Obviously, the leading customers in the logic and SoC space have been doing this for a few years.

We've seen increasing movement of devices, not just microprocessors, but parts like application processors and modems move to the copper pillar packaging technology. And as a consequence over the past few years, we've seen increased adoption and growth associated with our probe card technologies to serve those applications.

I think that's a trend that we're in the mid to late innings of. We still see devices at 28 nanometers that are a mix of copper pillar packaging and flip chip solder bump packaging. But we do see a continued migration of high-end devices so that customers can get the benefits of increased packaging density, reduced packaging cost, better power dissipation -- all the things that have cause the leading SoC companies in the world to adopt it. So the trend's continuing. It's not over yet. But we are getting into the middle age of the trend, I would say.

Great. Thanks for that. Then on the DRAM side of things, it's great to hear that the traction is getting back. Wondering if you can provide any commentary on the possibility of any follow-on orders and what might give you the confidence for DRAM to rebound in the second half?

There's a few elements to it. The first is, generally, as you can probably deduce from our guidance, we are seeing stronger shipment and revenue activity and demand for our probe cards in the second quarter from all three DRAM customers. As we continue through the second quarter, we're seeing that momentum build as different customers execute different parts of their technology transitions, and for example, the ramps of 20 nanometers.

As Mike Ludwig indicated in the prepared remarks, there's been a couple of recent customer earnings calls where they've talked about their ramp plans, especially around the 20 nanometer node. And obviously as they ramp the 20 nanometer node, they require probe cards to test those new designs. And both in the second quarter and we see building through the back half of the year, the demand's beginning to build to support those ramps.

Great, thanks. That's all I had.

Christopher Longiaru, Sidoti & Company.

Hey, guys. I'll add my congrats on the guide. My first question just has to do with how these DRAM orders kind of are trending, right? I mean, it sounds like you just had a couple weeks in terms of a push-out and you shipped a big chunk of revenue over the course of, we'll say, really April, right? Or April's not even done yet. So I mean, is this going to be kind of something that we can be cautious about going forward in the sense of, are these orders going to be this lumpy going forward? Is there any sign of that continuing?

Chris, it's Mike Slessor again. A good question. Obviously DRAM right now and our customer's behavior is in a very dynamic state of affairs. As we looked and talked and talked about in Q4 and into Q1, there was very little spending or investment in probe cards or any other piece of consumables or capital equipment for that matter in the DRAM industry. And I think you've seen that be consistent through the DRAM supply chain.

With their various yield-improvement activities and their desire to move to the 20 nanometer node, we've had several customers get pretty active as they've managed to get their yields up and managed to increase the number of wafer starts. So as long as that trend continues -- and the commentary both in public and we have in our planning sessions with customers is that that trend's going to continue. But when we looked at Q1, it was obviously a situation where, as I said, it was very dynamic in terms of their wafer start plans and their ability to ramp.

Okay, that makes sense. And so as this positive trend continues, it gets, I would say, a little easier for you to pinpoint maybe the timing -- not necessarily the level of revenue, but the timing -- whether it's going to be shipped in March or April or June or July. I mean, is that kind of just fair to say, or are you still being cautious in terms of your expectations? Because you could ship the last week of June, but you could ship the last week of July. I mean, would you say your guidance range has something to do with that?

I don't know that it has a lot to do with specific timing of shipments like that across the end of the quarter. Clearly, that's one of the factors that impacted our Q1 results, as we talked about.

Having said that, I don't want to paint a picture that everything is rosy in DRAM. We are seeing the indications and some of the orders associated with the ramp activity that our customers have talked about publicly. It's got a long way to go before it gets to the run rate levels where we would consider the normalized DRAM business to be and where, you guys, I think, understand where the DRAM probe card market should be on an equilibrium, normalized basis.

Okay, that makes sense. My next question is, so you hired some people to deal with the SoC capacity ramp. Can you give us a little idea of how that affects your expectations for operating expenses going forward?

Chris, this is Mike Ludwig. So the people we hired were primarily sitting in -- they're direct labor and engineers that help us with that ramp and execute that ramp. Those expenses are in cost of goods sold. So in essence, they're already baked into the guidance that we gave for gross margin. So if you're talking about below the line, below the gross margin line, they'll have no impact on what we would call OpEx expenses.

Great. And just in terms of the 1% kind of lag that you said would drag a little bit at the capacity expansion, is that really just getting guys more up and running and more productive as they move through? Is that kind of the idea about what the 1% (multiple speakers) --?

Yes, certainly. So you bring on a lot of people. there's training costs. There's efficiencies in terms of getting them up to speed. So that's really what we're talking about. Again, first quarter we had about a 4% drag on margin. Second quarter looks like it's a 1% because, again, a lot of these people have already been hired and they're already getting up to speed. So there's just a lag on some of the tail end of it.

Chris, it's Mike Ludwig again -- oh, go ahead. Sorry.

One other point on this. With memory being as weak as it was in the quarter, we also took the opportunity to redeploy some of our manufacturing resources into the SoC product line. And so the net addition to SoC capacity got us to the doubled output level. But we did it also by using some of the fungibility of resources and tools that we have at our disposal as we serve all three major probe card markets.

Can you guys quantify maybe how much -- as you doubled capacity, it's basically you kind of almost probably doubled personnel, right? So how much of the personnel came from internally redeploying and how much was hiring?

So first of all, we didn't really double the resources. I think as Mike really kind of mentioned, we probably increased resources by, in the ballpark of 25% in order to do that. And I would say still, probably if you had to put a number on it, I'd say 80% to 85% were outside, and there were probably another 15% were we redeployed resources from other factories.

Got it. Okay, and just, I know you don't want to comment too much on Cascade, but you talked about just kind of low-hanging fruit in terms of the $10 million to $12 million that you're still confident about over 18 to 24 months. Should we still be thinking about that $10 million to $12 million as kind of a straight line over the next 18 to 24 months, or is there kind of a bigger portion in the beginning? And then does it flat line? Can you give us a little bit of direction on how to model that?

Yes, again, so we're still relatively early in the process. As Mike says, we have started the integration meeting, so we probably have a little better sense on that. And again, we do feel confident that we will achieve the $10 million to $12 million. I think if you looked at it this way, let's assume that we closed this some time in the second quarter. I think in the third quarter, there are some real obvious ones.

So we look at this and there are some real obvious ones, probably more so in G&A than any other areas when you have multiple Board expenses, public company expenses, those sorts of things will be pretty easy to get at. So I would expect in the first quarter, we're probably going to see about a million dollars hit there.

As we end the year, we're probably going to be looking at maybe another incremental million dollars. And then as we get into the second half of 2017, we probably have another incremental million dollars. So that gets us into the run rate of $3 million a quarter, which is $12 million on an annual basis. So that's kind of the timeframe. So I think the first million probably isn't too difficult after that. Again, it takes some time to get some of these other ones.

That makes a lot of sense. That's very helpful. Thank you, guys. I'll jump out.

(Operator instructions)

Patrick Ho, Stifel Nicolaus.

Thank you very much. Looking forward on the DRAM market and in particular some of the transitions that are going on in the industry right now, Mike, can you give us a little bit of color in terms of the potential, I guess, capacity additions or capacity ramps given we're talking about different designs, different nodes. You've got some going to 20, you've got some going to the 1X and 1Y. How do you feel about your current capacity ramp for these different designs?

Patrick, this is Mike Slessor. So capacity for us for DRAM is largely limited by the existing capital equipment that's in place in one of the major FormFactor facilities. What we have done obviously in the movement of some of the direct labor resources from that facility from DRAM into the SoC area, we would need to be back-filling those resources to get back to the DRAM levels that we had maybe in the Q2, Q3 2015 timeframe.

We've not made the decision to do that yet because, as I said, although we see gradual improvement in DRAM, we're not really (inaudible) the demand levels that we've seen in the middle of 2015. And so we don't want to incur those costs, make those investments, hire those people to get the capacity up before we really clearly see the demand in DRAM.

Having said that, given that it's a much more capital-intensive business for us and less labor intensive, I don't see a lot of issues responding quickly to the DRAM demand when it comes back as we plan it to in the second half.

Great. And maybe looking forward also on the Flash business, I know that market's quite weak right now. But you've had now probably about a year of the new Vector product. Can you just give us an update on, I guess, the trends there, the adoption, and when we may be seeing, I guess, a more meaningful revenue ramp on that front?

Yes, Patrick. It's a good question. So Vector for us, our product to address the Flash market, obviously you can see from the revenue levels, it has been disappointing for us. As we have been qualified at two of the four major NAND Flash manufacturers, we view the middle part of 2016 as 3D NAND really starts to ramp on some key designs, and I think you've seen indications of that from various capital equipment companies over the past couple of days. We're competing for those designs, and it's a real key test for the Vector product.

So I would set the expectation, as we have in the past, that we expect to see the $2 million to $3 million of incremental revenue from the Vector product exiting 2016. But it's a very key point in our ability to go address the NAND Flash market and build on the qualifications and initial penetration we've had. Obviously, the $1 million Flash result we had in Q1 is well below what we need to be a leader in that segment. Having said that, we remain committed to the Flash segment. And if we need to go address it in a different way, we will do so.

And maybe just to follow up on that, is it fair to say that part of the slow adoption is the industry itself where, I would say that most of the industry suddenly went to 3D NAND. I would say, obviously over the last year or so, they made the decision that we've got to transition now to 3D NAND versus extending their planar nodes. Do you feel that was part of, I guess, the slower than expected traction?

Certainly that was some of it. But as we saw that trend at the customers we were engaged with, we started to position the Vector qualifications and the designs we were competing for 3D. So I certainly wouldn't say that the slower than expected adoption of Vector has been purely associated with the industry transition from planar to 3D. But it has had an effect because there's not sort of the continuous running of the designs that we were initially qualified on.

Great. Thank you very much.

Tom Diffley, D.A. Davidson.

Yes, good afternoon. So I guess following up on that last question, when you look at the vector or the NAND business, where does that fall on the spectrum of people-intensive work for the SoC versus the capital-equipment-intensive side of the DRAM? I know it used to be in the DRAM side, but has that moved now?

So Vector for us, I'd characterize, Tom, as a mix between the more people-intensive SoC business or labor-intensive SoC business and the more capital-intensive DRAM business. It's part of the general theme of us post MicroProbe merger, aligning product roadmaps and manufacturing facilities so that, again, we have this ability to respond to demand in the different segments as one goes up and the other goes down.

Now, if they're all going up at the same time, then as I said in response to Patrick's comment about DRAM capacity, we're going to have to add net headcount. But as Vector represents a place where we can use a lot of the existing capital and represents one of the areas where we pulled direct labor resources from to go into the SoC business. So you can think about some, again, flexibility and fungibility of our resource and capital equipment pool there.

Okay. And then I guess, when you look at the added headcount that's in COGS at this point, did that change the incremental gross margin landscape? I mean, before it was somewhere in the 55%, 60% incremental range. Is that still the same range?

Yes, we think it's still the same range, Tom, yes.

Okay. And then on the DRAM side, you talked about how the thee customers that are ramping from a bit of a depressed level. Curious though, was the push-out just a single customer, one large project, essentially?

No, it was certainly more than one customer. It was a couple of different designs, as I think we discussed on the last call. While some of those same customers were, if you like, taking shipments for different designs in the first quarter. So to go back to a previous question, it really, I think, was a good indicator of the dynamic state of these customer's ramps, yields, their ability to yield different designs on these advanced nodes.

Okay. So was it just coincidence that several of these projects happened to coincide very late in the quarter then?

Yes, not to be too flippant about it, but it's often for budgetary reasons that you also get push-outs late in the quarter. So it certainly was the confluence of these different designs in the seat part of a 20 nanometer yield curve and ramp, and customers trying to, let's say, optimize their spending for the period end.

Okay. And then I do have a couple questions on Cascade I hope you can answer, but maybe you can't. First one is, obviously they just reported some very strong results, and really high gross margins. In your view, is that margin profile for that product repeatable?

And then more on the negative side, obviously we just got some worse than expected forecasts from Apple. And I'm curious if you think that trickles down into their business at some point over the next couple quarters?

Yes, John, it's Mike Slessor again. Maybe two comments on the Cascade gross margin profile. As you might have noted from their commentary, a lot of that was mix-driven. They had a very favorable product mix across all of their segments.

And so although a very impressive performance from that team, and we're obviously even more excited about bringing that team aboard the more we learn about them and the more they grow and produce these numbers that are going to be part of FormFactor, I do think there are some short-term mix advantages that are occurring in this period that we might not look for to be long-run sustainable. It does point to some of the potential for the business, however.

On the RF handset side, one of the key drivers, although there are multiple drivers that I'll get into in a minute, one of the key drivers for the Cascade production probe business has been mobile handset growth. One of the key applications where Cascade has a strong position is in testing the surface acoustic wave and bulk acoustic wave filters that go into the front ends of these phones.

So certainly, a reduction in handset growth or a reduction in handset demand is not good news. However, I think there's a couple of other underlying secular trends there. As the filter customers have talked about, they're seeing increased demand even as handset growth slows because of band proliferation. Essentially, they sell more filters per handset as people progress from, for example, 4G to 5G. And so we think there's an underlying buoyant trend there that helps out even in the face of slowing handset growth.

The other thing I'll say is, the trend to RF and RF test is much broader than just the front end of iPhone 6s. The applications associated with things like millimeter wave radar, with various internet of things applications that have to communicate with each other, we see those as fundamental, systematic drivers of RF component growth, and therefor, RF test and RF probe card growth.

Okay. And even if there isn't much unit growth, I assume you probably see new design activity on an annual basis just when the new models come out?

That's correct. I mean, the probe card business, as you infer, is really driven by new design releases and new parts which require new probe cards. And so even if unit growth is level, as long as there's new designs inside that unit growth market, there's pretty significant probe card demand.

Great. Thank you.

(Operator instructions)

Craig Ellis, B. Riley.

Thanks for taking the question. And gentlemen, thanks for all the transparency on the first-quarter issues. My first question is just regarding the modest change to the second-quarter outlook versus 30 days ago. I think I heard that was really an improvement in the DRAM market, no change in SoC, but I wanted to verify that.

Yes, so on the $76 million to $80 million, I think we just have, again, greater visibility, Craig, on both our ability to execute the ramp and what we're seeing there as well as some, I think, gradual improvement that we're seeing in the DRAM market. I think we're getting a little more comfortable around that. So that's why we tightened the range and did it toward the upper end.

Okay, great. With that tightening, it looks like the Company's recaptured about 45% of what it lost in the first quarter. And I think the view was 30 days ago that you had recaptured all of that $14 million through the year. Can you give us some color on when that remaining $8 million might come into the model? I'm not asking for guidance for 3Q or 4Q. I know you can't do that. But maybe qualitatively, you can talk about when we would expect that incremental revenue to come in?

There is a couple of components to it, Craig. The SoC ramp, now that we have a significant number of consecutive weeks under our belt operating at or above the doubled capacity level, we're -- and as Mike Ludwig said, that's a piece of the guidance being in the $76 million to $80 million range, whereas before it was in the $72 million to $80 million range.

So we're recapturing some of that demand in the second quarter. But given the demand levels and the continued order patterns and backlog associated with that product line and that customer, that will continue through the third quarter and possibly even into the fourth quarter where different elements of that capacity are being utilized in a burst mode to capture above this doubled run rate.

If we look at DRAM, again, that's a pretty dynamic situation. And so the notion of recapturing lost demand, I think, is a little more elusive in the DRAM case. We do have obviously, as we said, shipped the cards that have been delayed from Q1, and have now shipped them in Q2. So I would consider that demand to be recaptured.

And we are seeing, as we said, a gradual increase in the activity, our shipments and our revenue associated with DRAM in the second quarter. So we just view everything as kind of pushed out to the right by half a quarter or so. I think we're able to recapture that as long as DRAM continues on this ramp in healthy trends. Probably as we move through the second half, but probably some of that is a 2017 event as well.

Okay. That's helpful. Thanks, Mike. And then the follow-up, I think 30 days ago you also voiced strong confidence in year-on-year growth this year. Is that still the case? And would you like to put any further color around some of the things that you're seeing that give you that confidence?

So Craig, we are still confident in that. We made a point of reiterating that comment in the prepared remarks and in our press release, which went out earlier today. The color around that is the components we've talked about. It's one of our major customers where we have obviously a strong relationship and strong share position operating at a doubled capacity level, plus now executing and delivering at that doubled capacity level for the rest of the year. We see the typical mid-year application processor and modem cycles beginning their shipment cadence. Those designs releasing, us delivering cards into that demand in the middle part of the year.

And then the third part of this is obviously our assumptions and commentary around DRAM. We need DRAM to be in the second half at its healthy, normalized levels. We do see indications of that. We do see fundamental shipments, bookings, purchase orders supporting that trajectory. But those are the three things that go into our statement that we're confident we will both outgrow the probe card market and grow FormFactor's revenue for the fourth consecutive year.

Thanks for that. And then my last question is just on the transaction. And not about the transaction itself, but about one of the business model differences, and hoping to get some of your views on one attribute. In the just-reported quarter, Cascade systems revenues and probe revenues were about equal, and that was, I believe the first time that's happened.

Can you talk about the interest you have in a model that is more balanced, system versus probe, than what you have now, and what the milestones or steps would be to move in that direction if it was something that you chose to pursue?

Well, I think for us, we don't have necessarily a model that basically gives you a systems versus probes. I think from our perspective, they've got two very good business there. I think for us, one of the, again, nice things is the systems business adds a growth leg to FormFactor's portfolio, so we're excited about that.

Certainly, we're excited to get a probes business that is as profitable as that one is, and look to combine that probe business in terms of those revenues and profitability profile of ours on the probe side to, I think, have a much more competitive and a good company. But I don't think we have targets at this point in time in terms of sort of percentages of our probe business versus systems business. Each of the businesses is independent, important, and we want both to have good growth. But there is no target percentages for those in terms of going forward.

All right, thanks, guys.

Amanda Scarnati, Citigroup.

Hi, thanks for taking the question. And I apologize if this has already been answered. I've been kind of back and forth a little bit. What gives you confidence in the strength of the DRAM business in second half given memory makers like SK Hynix and potentially Samsung taking down CapEx numbers for the year? Do you see growth outside of that? Or where are you kind of seeing that growth coming from?

Right, and Amanda, we did have a bit of a discussion on this earlier. But no problem, we can go through it. If we look at certainly the yearly CapEx numbers for somebody like SK Hynix, as you said, have been taken down. One of the things that has us optimistic about the second half though is if you dig into those numbers a little bit, and you look at how they're spending and how they're talking about their new node, their 20 nanometer, for example, ramp, and their wafer starts as they go through the back half of the year, that's going to require some significant probe card demand to support, and probe cards that they don't have now.

We are, as we said, beginning to see the initial stages of our shipments, revenues, and activity to support those ramps. And our business is going to be largely symmetric to the trajectories of those ramps and those product release cycles for customers on the 20 nanometer node. So although their yearly CapEx is certainly down 2016 over 2015, the dynamic of them ramping these nodes and needing the tooling to sort the wafers as they produce as they go into the back half of the year, is something that leaves us pretty optimistic.

SK Hynix had an earnings call the other night, and I think their commentary about the trajectory through the year of their ramp, their wafer starts, and their design releases is pretty consistent with the trajectory and timing that we're talking about.

Great. And then just finally on the margins and how you're looking at them through two halves. Typically margins tend to peak earlier in one half and then fall off in the second half. Are you seeing a reversal of that off of the weaker March quarter? Or where are you thinking of (inaudible) margins going forward?

I think margins going forward really depend a lot on both the business mix as well as the customer mix within each one of the businesses. So last year if you look at it, the first half was very strong from a DRAM perspective. We had the factory filled up, and we had higher margins, let's say, non-GAAP margins in the 37%. Towards the second half of the year, they tend to be, as you noted, lower because, again, DRAM wasn't nearly as strong.

If you look at the first half of the year, so the margin profile here is probably, again, looks more like the second half of last year because, again, DRAM probably isn't as strong. We're fortunate that we have a good SoC business that will certainly help us in the second quarter. But we would look to have margins be stronger, I think, in the second half of the year than the first half of the year as DRAM again picks up and we get more loading of our factory here. So I think that's how the margin profile looks. So it may look opposite this year versus what it was in 2015.

Great, thank you.

Thank you, and if we have no further questions, I'd like to turn the conference back over to Mike Slessor for any closing remarks.

Thank you very much for joining us today. We look forward to delivering a strong second quarter and closing the Cascade Microtech acquisition. Thanks again.

Ladies and gentlemen, thank you for participating in today's conference. This concludes the program. You may now disconnect. Everyone have a great day.