Good day, ladies and gentlemen, and welcome to the FormFactor Inc. first quarter financial results conference call. At this time, all participants are in a listen-only mode. Later, we will conduct a question and answer session, and instructions will follow at that time. If anyone should require operator's assistance during today's conference, please press star then zero on your touch-tone telephone. As a reminder, today's conference is being recorded. I would like to introduce your host for today's conference, Mr. Jason Cohen, company's general counsel. Sir, please go ahead.
Thank you. Today, the company will be discussing GAAP P&L results and some important non-GAAP results intended to supplement your understanding of the company's financials. Reconciliations of GAAP to non-GAAP measures and other financial information are available in the press release issued today by the company and on the investor relations section of our website. Today's discussion contains forward-looking statements within the meaning of the federal securities laws. Examples of such forward-looking statements include those with respect to the projections of financial and business performance, future macroeconomic conditions, foreign exchange rates, business momentum, business seasonality, the anticipated demand for products, customer requirements, our future ability to produce and sell products, the development of future products and technologies, 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 2017, 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 2nd, 2018, we assume no obligation to update them. With that, we will now turn the call over to FormFactor CEO, Mike Slessor.
Thanks, Jason, thank you, everyone, for joining us today. The first quarter of 2018 demonstrated the benefits of FormFactor's opportunity set as a diversified leader in electrical test and measurement. As anticipated, revenues from our largest customer were at their lowest level since 2016. We continued to capitalize on our line of sight opportunities in advanced packaging, mobile data, and automotive applications, enabling us to deliver revenue and non-GAAP earnings per share at the top end of our outlook. Highlights were solid probe card demand in both foundry and logic and DRAM markets, driven by growth in advanced packaging applications a robust memory investment environment. Momentum also continued in our engineering systems business, driven by multiple applications and regions in what is typically the seasonally weakest quarter of the year.
These drivers are continuing in the second quarter, augmented by a forecasted sequential increase in demand from our largest customer. For 2018 overall, we continue to expect and plan for revenue growth as we progress through the year. Our revenue trajectory will depend on three key factors, namely, recovery and production probe card demand from our largest customer, further adoption of our technologies in advanced packaging applications at the world's largest foundry, and a continued strong investment environment in DRAM. Each of these is a major focus area for our worldwide team, and I'd like to provide more details on each of them here. As recently announced, yield challenges at the 10 nanometer node continue to be significant and have caused a further delay in the manufacturing ramp of this node to 2019 at our largest customer.
They also announced that new products will continue to be released on the 14 nanometer node. As a reminder, the probe cards we provide are a device-specific consumable, and these new 14 nanometer designs require new probe cards, which generates some incremental demand for FormFactor products. We've shipped initial units of probe cards for these new 14 nanometer devices and expect shipment volumes to grow as these designs transition to high volume manufacturing. Accordingly, we continue to expect second quarter revenues of this customer to increase from the first quarter lows. Looking a little further down the road, we maintain our view that this business will build throughout 2018, albeit at lower levels than from a full-fledged 10 nanometer ramp, driven by a combination of sustained 14 nanometer activity and gradually increasing 10 nanometer activity.
We continue to actively monitor and manage this very dynamic situation, leveraging the learnings from our 4 generations as a key supplier to this customer. Advanced packaging continues to be a key driver for FormFactor's probe card business, and in the first quarter, we shipped significant volumes for customer projects in both foundry and logic and DRAM. In advanced packaging applications, FormFactor's MEMS probe technology is one of few viable options to meet high density and stringent electrical performance requirements. In the foundry and logic space, we delivered record quarterly shipments to the world's largest foundry, supporting a major seven nanometer application processor with integrated fan-out packaging, and these shipment levels have sustained through the start of the second quarter.
Although this business is still concentrated on a single design at present, we are also beginning preparations for second half shipments to this customer, supporting 10 and seven nanometer designs from multiple fabless chip suppliers. This is an important step as we expand our footprint and business with this strategically important customer. Staying with our advanced packaging opportunity, our first quarter DRAM revenues contained a significant number of probe cards for testing of stacked die high bandwidth memory devices, also known as HBM. This example of advanced packaging adoption in DRAM enables manufacturers to generate new and innovative products with higher densities and performance, and FormFactor's technology allows them to effectively test these products.
In DRAM broadly, as you have heard from our customers and other suppliers, there continues to be significant but sustainable investment in both the migration to new nodes and the release of new designs. We expect these investments to continue through the balance of 2018. Our engineering systems business produced strong financial results in the first quarter, with gross margins increasing following our fourth quarter investments in strategic tool placements for silicon photonics applications. In addition to the revenue growth and financial contribution this business continues to deliver, it also provides strategic value from early customer engagement in development of next generation semiconductor and related technologies. Silicon photonics is a good example of this, along with the RF components that will power upcoming 5G devices, as well as the microLED display technology we have discussed in the past.
In these emerging applications, along with others in our development funnel, we are working with the world's leading customers to develop the required electrical test and measurement technologies. When these applications eventually reach volume production, we will be ready with the right technologies and products. It is a key part of our Lab to Fab strategy you can read about on our website. I now want us take a step back and take a broader view of our largest business. Earlier this week, VLSI Research published its annual survey of the probe card market, showing FormFactor again grew share from our leadership position to capture over a third of the $1.3 billion advanced probe card market.
Demonstrating the increasing benefits of scale in our space, FormFactor, together with the number 2 and number 3 suppliers, increased our combined market share to over 60% of the market, up from less than 50% five years ago. According to VLSI, the largest and fastest-growing part of the market is the nearly $1 billion spent on probe cards fabricating using MEMS processes. As I mentioned earlier, the scalability and performance of FormFactor's MEMS technology is helping enable industry adoption of advanced packaging. Although there will always be short-term volatility in our individual businesses, the large investments required to innovate and develop advanced probe technologies such as MEMS generate competitive advantage with significant entry barriers and will drive secular share growth for FormFactor in the years ahead. Finally, I'd like to introduce and welcome our new Chief Financial Officer, Shai Shahar.
Shai's been part of our team for two months. In that time has begun to leverage his extensive financial and operating experience in the semiconductor industry to help further improve FormFactor's performance and operational efficiency. I'm enjoying the process of building our partnership. I'm looking forward to working together to realize our target financial model that delivers $650 million of revenue and $1.50 of non-GAAP earnings per share. With that, I'll turn the call over to Shai for further details on our first quarter results and to provide further insight into FormFactor's financial outlook.
Thank you, Mike. Good afternoon. As Mike mentioned, I joined FormFactor in early March. I'm very excited to be a part of the FormFactor family. The company has great potential for growth and value creation given our leadership position in the electrical test and measurement space and the smart and passionate people that I've already met in the company. Having played a key leadership role in similar business situations in the semi and semi equipment industries, I am confident that we can continue to execute our strategy to drive revenue growth and margin expansion. As you saw from our press release and heard from Mike's comments, our largest customer's volume production delay in its 10-nanometer process resulted in an anticipated revenue decline in Q1 as compared to Q4.
Despite that decline, we achieved first quarter results slightly above our expectations, with revenues above the midpoint and gross margin slightly above our outlook ranges. These results, coupled with continued good operating cost control, allowed us to deliver non-GAAP EPS at the high end of our outlook range. FormFactor's revenues for the first quarter of 2018 were $118.3 million, down 10.3% from the fourth quarter of 2017. Probe card segment revenues of $94.9 million decreased 11.5% compared to Q4. Systems segment revenues of $23.4 million decreased by $1.3 million or 5.3% compared to Q4. The systems segment's revenue decrease is lower than the usual seasonal decline. Within the probe card segment, foundry and logic revenues of $58.4 million decreased 15% compared to fourth quarter. The decline was attributable to the anticipated lower revenue from our largest customer.
On a more positive note, during the first quarter, we saw a significant sequential increase in shipments of a new design to the world's largest foundry, which utilizes our technology in advanced packaging applications for leading-edge nodes. Some recent announcements indicated soft end-market demand for high-end handsets. However, as our customers continue their innovation in BAW and SAW filters and other RF components, probe card revenue for RF applications were essentially flat in the first quarter, consistent with the seasonality we see in this market. Foundry and logic revenues were 49% of total company revenues in the first quarter, down from 52% in the fourth quarter. DRAM revenues were $30.3 million in Q1, down $1.6 million or 5% sequentially. While the results were down slightly, overall, we continue to experience robust DRAM demand as technology node transitions and a strong data center demand positively increased probe card spending.
DRAM revenues were 26% of total company revenues in the first quarter, a slight increase from the 24% in the fourth quarter. Flash revenues of $6.2 million were $0.4 million lower than the fourth quarter. Approximately half of the flash revenues in Q1 were from the NAND flash application. The slight sequential decrease in system segment revenues was driven by a decrease in station sales, mostly offset by record sales of thermal subsystems. System segment margins in Q1 improved substantially from the Q4 low, as I'll discuss in a moment. GAAP gross margin for the first quarter of 2018 was $45.1 million or 38.2% of revenues, up 130 basis points compared to 36.9% in the fourth quarter of 2017.
Q1 included $6.1 million of GAAP to non-GAAP reconciling items, which you can find outlined in a reconciliation table available on the investor relations section of our website, and in our press release issued today. On a non-GAAP basis, gross margin for the first quarter was $51.2 million or 43.3% of revenues, up 150 basis points from 41.8% in the fourth quarter, and slightly above our outlook range. The increase was primarily due to a favorable product mix and lower warranty costs, partially offset by a negative impact of a stronger Euro. Our probe card segment gross margin was 42.2% in the first quarter, an increase of 50 basis points compared to Q4 on lower revenue, due primarily to the favorable product mix mentioned above, redeployment of direct labor to high-demand product manufacturing, and lower warranty cost in Q1.
Our system segment gross margin was 47.7% in the first quarter, a significant improvement of 5.2% compared to Q4 on slightly lower revenue. As you recall, our fourth quarter systems gross margin was negatively impacted by strategic tool placements. Q1 did not include this, and margins returned to a normal level despite a stronger Euro, which resulted in approximately 1% decrease in the Q1 systems gross margin. We partially hedge our foreign currency exposures. However, the benefit from our FX hedging activities is reflected in our other income line in our P&L. We expect the systems business gross margin to continue to improve towards a target level of high 40s, low 50s, including FX potential impact. Our GAAP operating expenses were $41.5 million for the first quarter, $1.7 million lower than the fourth quarter. The first quarter operating expenses included $4.9 million of GAAP to non-GAAP reconciling items.
non-GAAP operating expenses for the first quarter, including ERP implementation costs, were $36.6 million or 31% of revenues, up $0.6 million compared to Q4. The increase is mainly due to higher R&D investments and ERP integration and implementation costs, partially offset by lower performance-based compensation. Starting 2018, nearly two years from the completion of the acquisition of Cascade Microtech, as we achieved our initial cost synergies commitment, we view ERP integration and implementation expenses as routine. Accordingly, these expenses are not excluded from our non-GAAP operating expenses. ERP implementation and integration expenses were approximately $1 million in Q1, and we expect a similar quarterly expense level during the remainder of 2018. Company non-cash expenses for the first quarter included $7.2 million for the amortization of intangible assets, $3.8 million for stock-based compensation, and depreciation of $3.5 million.
The non-GAAP effective tax rate for the first quarter was 5.4% compared to 1.1% for the fourth quarter. The higher effective tax rate was due to higher percentage of profit in foreign tax jurisdictions versus the U.S. We ended Q1 2018 with approximately $240 million of remaining U.S.-based NOLs. As such, we continue to expect to have a non-GAAP effective tax rate of 4%-6% while we utilize our NOLs. GAAP net income was $2.1 million or $0.03 per fully diluted share for the first quarter, compared to net income of $0.07 per fully diluted share in Q4. First quarter non-GAAP net income was $12.7 million, or $0.17 per fully diluted share, compared to $18 million or $0.24 per fully diluted share for Q4. Moving on to the balance sheet and cash flows.
We generated $6.3 million of free cash flow in the first quarter, taking our total cash to $143.7 million at the end of the quarter. We spent $9.6 million of principal and interest payments on our term loan during the quarter, including $5 million of prepayment. Our total cash balance exceeded the balance of our debt by $46.7 million at quarter end. In line with our annual capital spending plan of $16 million-$20 million, we spent $3.8 million on capital expenditures during Q1. The decrease in free cash flow in Q1, as compared to the $23.5 million generated in Q4, is mainly the result of low Q1 revenue and inventory investment in anticipation of revenue increase in Q2. Turning to the second quarter non-GAAP outlook.
We continue to experience strong probe card demand in DRAM and foundry and logic markets, including stronger second quarter revenue from our largest customer. We expect revenues in Q2 to be in the range of $130 million-$138 million. Consistent with the higher revenue outlook, second quarter volumes and factory utilization are expected to be higher than in the first quarter, but will be partially offset by a memory-rich product mix. We currently expect second quarter non-GAAP gross margin to be in the range of 42%-45%. We expect to realize non-GAAP earnings per fully diluted share in the range of $0.20-$0.26. A reconciliation of our GAAP to non-GAAP Q2 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 to questions. Operator?
Thank you. Ladies and gentlemen, if you have a question at this time, please press star then the number one on your touchtone telephone. If your question has been answered or you wish to remove yourself from the queue, please press the pound key. To prevent any background noise, we ask that you please place your line on mute once your question has been stated. Our first question comes from the line of Edwin Mok with Needham & Company. Your line is open. Please go ahead.
Hey, guys. Congratulations for the great quarter and great guidance. First question I have, if you can kind of let us know how many 10% customer you had in the first quarter and how many you expect in the second quarter, and how big was your largest customer in the first quarter?
Yeah. Thanks, Edwin. It's Mike Slessor. We had two 10% customers in the first quarter. The queue will come out next week, and you'll see the details of that. Obviously, our perennial 10% customer, our largest customer, was significantly stepped down in the first quarter. Because of the heavy activity and demand in DRAM, we added one of our DRAM customers as a 10% customer. As we move here through the second quarter, I think there's a reasonable chance that some of our other memory customers will end up in the disclosable 10% customer line, but we'll have to wait and see sort of how that plays out as we finish up the quarter.
Okay, great. Then I guess focus on our large customer. As you mentioned, there's a push out or they have pushed out their 10 nanometer ramp, but it sounds like the kind of growing demand for 14 nanometer as they roll out new products is still supporting demand for probe cards, and that's why you guys are high for the second quarter. How do you kind of see that progressing into the second half? I think historically, you guys talk about potentially record second half. Obviously, that predicates on 10 nanometer is not happening. Maybe you can kind of give some color on how you kind of think about the demand dynamics into the second quarter and then second half relative to what you've done in the first quarter.
Yeah. I think directionally, we continue to see our largest customer's demand building throughout the year. As you noted, the addition of some new 14 nanometer designs that drive probe card demand is a helpful trend that's gonna directionally move in that way. As you also noted, the expectation that we'll have record demand and record revenue from this customer is probably not a realistic one given the push out of the 10 nanometer node. I think people over the last few weeks have seen various data points that would confirm that.
We still do see 10 nanometer activity and the existence and appearance of 14 nanometer designs and probe card activity is certainly a helpful trend that will allow us to continue to execute on growth throughout the year with this customer, but obviously not at the levels that would take us to record levels in Q3 and Q4.
Okay, great. Two more questions. One is on the largest foundry. Sounds like you guys have done really well in the first quarter and you expect the level to stay in that kind of similar level, at least in the second quarter. As you go into the second half, I know historically you guys are very concentrated on one particular design with that customer. How has that changed this year? I think that kind of let us kind of think about how we think about your opportunity with that customer if we look beyond this year, and how do you think that can potentially progress in 2019?
Well, I think it's a very important point in our opportunity set. Our opportunity, both last year and for the first part of this year at the largest foundry, really has been focused on a single design that's driving the most advanced node and the integrated fan-out packaging. As we move through the back half of this year, we're seeing some opportunities beyond that single design. They're still very much associated with the advanced node and the integrated fan-out packaging, where FormFactor's MEMS technology really enables these customers to do an effective job of wafer test. We're seeing the initial activity on broadening this business beyond the single design. We think that'll result in some probe card revenues in the second half of the year.
More importantly, as we move forward into 2019 and beyond, as more of that foundry's wafer mix shifts to the advanced nodes and employs advanced packaging, things like integrated fan-out, we think that's a real opportunity for FormFactor to grow our business much more significantly. The opportunity will shift towards us as their mix shifts more towards the advanced nodes and integrated fan-out packaging, where really FormFactor's MEMS technology is a key enabler.
Okay, great. Thank you. Last question I have on the system side. I think you mentioned on the call that 1Q was better than seasonal, right? I think it came off a really strong fourth quarter, and last year, I think it grew 15% last year. Historically, we don't expect much growth on the system business. How do you kind of think about that business this year? Seasonally, typically, you grow through the year, right? Do you still expect a similar trend this year?
We are planning on growth for the systems business again this year. The double-digit growth we delivered in 2017, to be honest, surprised us a little. I think it's a result of a couple of different things, but one of them is the broader customer engagement we have as part of the worldwide FormFactor team with the leading customers in the world in serving a whole vast plethora of applications, things like microLED, things like silicon photonics, that continue to broaden the opportunity set for the systems business beyond its sort of classical niche of CMOS characterization. We are planning for a growth year. I don't think we'll be able to match the double digits we put up last year, but the opportunities continue to present themselves to diversify our business and work closely with customers on these emerging applications in development.
We're pretty confident that's going to result in some interesting opportunities when these applications get to production.
Okay, great. That's all I have. Thank you. Appreciate it.
Thanks, Edwin.
Thank you. Our next question comes from the line of Craig Ellis with B. Riley FBR. Your line is open. Please go ahead.
Thanks for taking the question, Mike. Good job executing in a dynamic demand environment. Shai, welcome aboard. Mike, I wanted to start with questions on Intel, just to understand some of the points that you're making there a little bit better. When you talk about 14 nanometer volume ramping through the year, is the point there that there is more happening from a new product standpoint, or that with volume growth, they're going to need more probe cards to deal with that volume, or something different?
No, Craig, I think it's more of the latter in that, I'm sorry, the former, in that to fill the hole associated with the continued delays of 10 nanometer on the product roadmap, this customer's chosen to release some new designs, rather than rely on the node transition from 14 to 10 to get them increased performance, they're continuing to innovate on the architectural side. Well, doing that means new designs, new products, to remind everyone again, new designs require a new set of probe cards. It really is this design innovation on 14 nanometer that's going to provide some level of increased demand for us, compared to a base case of just pushing 10 nanometer out.
Now, I will temper enthusiasm a little bit, because 14 nanometer is a relatively mature node, and so yields are high, and test times are relatively shorter than they would be for, say, 10 nanometer, that this isn't a straight trade of 14 nanometer volume for 10 nanometer volume as far as FormFactor is concerned. It is a helpful trend and is one of the aspects of this business that lends a consumables-like flavor to it.
That's helpful. It leads to my next question, which is the probe card intensity of 10 nanometer versus 14 nanometer, and understanding that Intel's in a completely different point of their ramp with those two technologies, very mature, 14 nanometer plus, 10 nanometer just coming up the curve. As 10 nanometer gets to volume sometime next year or whenever it does, would probe card intensity for 10 nanometer be similar to 14, or would it be, for whatever reason, higher than 14 nanometer?
I think we've seen in pretty much every node transition. This is true certainly at all of our customers in Foundry and Logic. As I think about it's probably true in DRAM as well. When a node first comes out, obviously the yields are lower. Customers are working aggressively on yield enhancements and yield improvement activities. That drives up test times and test intensity. Therefore there's a higher probe card intensity associated with the first part of a node. Over time, we've typically seen that then moderate or normalize to kind of historical levels, right? Nodes tend to get to entitlement yield levels. Customers dial back test times as they solve their problems. We tend to see a pretty normalized view of the probe card intensity per wafer start, if you will.
I think the thing that may be different this time. This again is true across foundry and logic and DRAM, across the industry. I think the Moore's law shrinks, the node shrinks are getting so difficult and so drawn out that we're seeing that normalization process take longer and longer. Look at DRAM, for example. Even with the capacity additions that each of the customers are going through, I think the view of DRAM bit growth rates for the year is still something only like 20%. That's indicative of how hard it is to shrink and how hard it is to add capacity. It's true, not just in our foundry and logic business, but I think it's true across the industry.
That's very helpful. Just picking up on the DRAM point, the business has been tracking very well right at $30 million per quarter. I think a couple of DRAM manufacturers recently talked about moving more towards volume 1X nanometer output as they go through 2018. As you look at the year, can you just frame for us some of the puts and takes from what has been a nicely consistent DRAM performance over the last four to five quarters?
As we said in the prepared remarks, DRAM remains one of the exciting opportunities here for us in 2018 to continue to grow the business. It fits into the larger FormFactor diversified profile of businesses, and we still believe that DRAM is cyclical, but right now we're in a cyclical investment period. As you note, different customers are talking about and planning for different node shrinks and different design releases as we move through the year. They continue to be on, as we've described before, slightly different cadences on these different activities, which is nice for us trying to balance our DRAM capacity. The sum total of it is that we view DRAM continuing to be a pretty healthy business throughout the balance of the year. It'll depend a lot on the end-demand scenarios and our customers' continued discipline on not over-adding capacity.
For right now and for the balance of 2018, we see DRAM probe card spend continuing at these levels, and we're executing pretty well on helping our customers migrate through those node transitions and design releases.
Last question from me, Mike, and I appreciate all the color. It has been two years since the Cascade deal, and our view has been for a long time that that was just a very strategically robust deal, and I think the post-integration performance of the business shows that. Can you just update us on how you're thinking about inorganic growth and where the company is in its process of moving from integration with Cascade on to potentially other opportunities? Thank you.
Sure. To reiterate, we've talked about our strategy having two components. The first, continued share and efficiency gains in the markets we serve, and I think the VLSI report that I highlighted in my prepared remarks shows that we continue to execute well on that front. The other piece that's important for us is to continue to increase our addressable markets through M&A. We said we want to be a buyer of leaders like we were with the Cascade acquisition, and leaders in attractive markets like we were with the Cascade acquisition and RF in the engineering systems business. We've refocused ourselves in 2018 in taking a fairly broad view inside the electrical test and measurement space.
We continue to go through that process, I'm very pleased with the progress our team has made on identifying opportunities that will increase our addressable market, that are probably resonant with some of the other themes in our business, like advanced packaging, and will continue to allow us to grow the overall FormFactor business along these two components. M&A is part of our future. We're continuing to focus on where we want to execute that. As you all know, deals take a while to come together, but I'm pleased with the progress we've made on the analysis and defining our funnel and direction as we move forward.
Thanks, Mike.
Thank you. Again, ladies and gentlemen, if you have a question at this time, please press star then one on your touch-tone telephone. Our next question comes from the line of Edwin Roesch with Sidoti & Company. Your line is open. Please go ahead.
Yes, hi. Just wanted to ask another question about your largest customer. If there hadn't been the planned node migration right out of the gates kind of early January, would you expect their business to have fallen off as much in Q1? Would they have spread out more of this 14 nanometer probe cards for their architectural changes, and have some of those orders fall into Q1 more so than what occurred?
I'm not sure I really understand the question, Ed, but let me talk about some of the puts and takes that can help, hopefully provide some insight there. Obviously, the fundamental principle here is that this is a very dynamic situation. The mix between continued 14 nanometer innovation and ramping 10 nanometer is obviously one of the bigger dynamic pieces. I think with all of our customers, they're continuing to innovate, whether it's moving to advanced packages on the 7 nanometer node at the world's largest foundry or using what's a mature node, like 14 nanometer, to continue to release new designs. Those are choices that our customers are making in a pretty active and real-time way. Baselines, to be honest, are pretty tough to establish here.
I think what we've learned recently is that the 14 nanometer innovation and design releases are going to be helpful in us continuing to grow off of the Q1 lows. They will be a lower intensity, a lower demand of probe cards than if it were pure 10 nanometer, as we discussed a little bit earlier. I think it points to the drivers of FormFactor's business in this device-specific consumables arena, where, if a customer decides to innovate with a Moore's law shrink, that's very good for us. If they decide to innovate by releasing a new design on a more mature node, that's good for us as well.
Okay. On the DRAM side, it seems like trailing 12-month revenues settled in at about $120 million, which is a very healthy pace. I was a little surprised that you had, it appears to have been a 10% customer show up in that segment, and still kind of be at that $120. Do you think this is sort of a peak 12-month trailing revenue level for DRAM for this cycle, or would you expect to maybe build on that pace?
Well, I think a couple of comments on DRAM probe card spending overall. We do see it continuing at that kind of run rate level. There is probably some relative strength right now, as we indicated in our guidance comments. We're going to have a pretty memory-rich product mix, and that's indicative of, for sure, a component of strong DRAM demand. I think from an overall intensity standpoint, I would not expect DRAM probe card spending and revenues to go up too much from the run rate we've talked about it. We'll probably see quarters that are a little bit stronger and maybe quarters that are a little bit weaker.
That was part of the comment around DRAM probe card revenues, I think, from Q4 to Q1 went down by a little bit. That's not indicative of anything other than a short-term fluctuation and some customer delivery request timing. We see the DRAM probe card investment and spending to continue to be right at about that level.
Got it. Okay. Thank you for the clarity there. Congrats on managing in a very dynamic situation.
Thanks, Ed.
Thank you. Our next question comes from the line of Tom Diffely with D.A. Davidson. Your line is open. Please go ahead.
Yes. Good afternoon. Maybe another question on the foundry business. I know a year ago in the second quarter, you had some really nice margins based on the type of product you were producing for them. Is that still the case? Is it still a high-margin business, and it's just the margin profile is getting overshadowed by more of a higher memory concentration?
Well, Tom, Mike Slessor again. If I back up to Q2 of 2017, that was a unusually unfavorable product mix on a whole bunch of fronts. There were some very high-end DRAM cards in there, as we talked about at the time. I think the Q1 gross margins probably more closer to normal from a product mix standpoint. Although, as Shai noted, we executed pretty well, gross margins for those revenue levels were higher than you might have otherwise expected. With our business, with the 7-nanometer fan-out package part, as we commented on the last call, we're a little bit earlier in the cycle this time. Obviously, our 10-nanometer experience last year, we were in there late, and we were kind of under probation as a key supplier, if you will.
This year, we were involved much more early in the process, obviously, it was a contributor to some of our Q1 revenues where it was not in 2017. As I said in the prepared remarks, continues into Q2. I still would expect this to be pretty lumpy, I think the timing shifted a little bit earlier than it was in 2017.
Okay. From a margin point of view, it seemed like the foundry business, the 7-nanometer business, or 10 last year, was a pretty nice adder to margins in calendar. I was surprised that you didn't get more of a margin kick in the first and the second quarter of this year from that business.
Well, again, the Q2 2017 had a whole bunch of favorable factors. One was the heavy foundry and logic mix, but there was a lot of other DRAM mix in that Q2 2017 that really helped us out.
Okay
I think viewing at it in isolation, probably maybe a little too simplistic of a view. The business this year, although it's still foundry and logic, which bears our highest margins, is spread out a little bit more between Q1 and Q2.
I noticed that when you talked about the flash business, you said 50% or so was NAND. I guess that means the other 50% is a NOR flash. I'm kind of curious what that model looks like. Is it more like flash or is it more like a DRAM from a probe card point of view?
In general, as we talked about, our flash strategy is to be relatively opportunistic. Most of wafer testing in flash is a pretty simple endeavor for our customers, they don't need FormFactor's technology. It's almost the inverse situation of some of the high-end foundry and logic or DRAM HBM opportunities I talked about, that really need a MEMS technology to effectively test. Flash is a much simpler proposition, therefore, the pricing and the margins are not nearly what we need to support our overall business model. Having said that, at the high end of flash, whether it be NOR or NAND, there are opportunities with individual designs that do look a lot like DRAM. They're high density, small die sizes, relatively demanding electrical requirements that play more towards our strengths.
As we've commented probably for the last couple of years, our strategy is to be opportunistic about competing for those, but not really participate in the mainstream commodity flash market. For our business, the flash, whether it be NOR or NAND, looks pretty similar, and it looks relatively similar to a low-end DRAM configuration where we are being compensated for the technology and value that we provide.
Okay, great. Finally, I noticed that you had a nice little uptick in business in Japan. I'm curious what segment that comes from and how it is that you compete against the homegrown competitor there.
Well, I think with all of our customers, well, not all of them, but most of them have a pretty global footprint now. It's hard to think of any region, whether it be Japan or the U.S., as associated with individual customers. I can give you examples. Obviously, DRAM tends to be still pretty localized in Korea, but there are elements in China. The U.S. DRAM manufacturer's footprint spreads all over the place because of the acquisitions it has. I think we're doing relatively well from a market share standpoint against our major competitors. Again, if you dig into the VLSI report that I mentioned, I think it pretty clearly shows that for the first time in a long time, we had higher market share in DRAM than our Japanese competitor.
We continue to try and execute to really win opportunities and execute on them profitably so that we can continue to build our share position and the overall gross margin and operating profit characteristics associated with the businesses we currently own.
Okay. Thank you for your time.
Thanks, Tom.
Thank you. Again, ladies and gentlemen, if you have a question at this time, please press star then one on your touchtone telephone. Our next question comes from the line of Gus Richard with Northland. Your line is open. Please go ahead.
Thanks for taking my question. Mike, on the fan-out opportunity, is the probe intensity for that kind of packaging more sufficient, or is the probe card itself more expensive?
I think there's two elements to it, Gus. The simple answer to your question is both. When 10 nanometer and integrated fan-out first ramped a year ago, the ATE provider commented that there was a much higher test intensity. You can go back and look at their commentary around the test intensity being much higher, partially because the yields were lower, but partially because there's a whole bunch of new failure modes to take care of. The other piece to this is because they are a more complex, higher density, higher performance probe card, those tend to bear higher ASPs as well. We've got both things working for us. As customers migrate to these advanced packages, higher density, higher node, more content, tougher electrical test content, that does raise the value of what we're doing, primarily through the MEMS technology.
We get compensated for that value. It's a tougher test problem, people are willing to pay for it to enable things like integrated fan-out.
Got it. A very similar question for high memory bandwidth. Is that, again, a more difficult test, longer test times, more probe cards, are the ASPs higher because the I/O speed's higher?
HBM's an interesting one, it's why we chose to comment on it in the prepared remarks. You can imagine the final HBM configuration is this stack of die, it turns out that each of those die need to be tested pretty rigorously to make sure that they're all going to end up being good in the stack and not kill their neighbors. Several people have referred to this over the years as known good die. I think our customers continue to come short of that standard, but for sure, the test intensity's going up. In certain cases, you're right, the speeds are higher, the electrical performance associated with the card tends to be quite a bit higher, those are more complicated configurations, but they're higher ASPs.
Got it. Okay, that was it for me. Thank you.
Thank you.
Thank you. I'm showing no further questions at this time, I would like to turn the conference back over to Mike Slessor for closing remarks.
All right. Well, thank you all for joining us today. We'll talk to you in another quarter.
Ladies and gentlemen, thank you for participating in today's conference. This does conclude the program, and you may all disconnect. Everyone, have a great day.