PDF Solutions Inc (PDFS) 2015 Q4 法說會逐字稿

Good day, ladies and gentlemen, and welcome to the PDF Solutions Inc. conference call to discuss its financial results for the fourth-quarter and full-year ended Wednesday, December 31, 2015. (Operator Instructions) As a reminder, this conference is being recorded.

If you have not yet received a copy of the corresponding press release, it has been posted to the PDF website at www.pdf.com. Some of the statements will be made in the course of this conference are forward-looking, including statements regarding PDF's future financial results and performance, growth rates, and demand for its solutions. PDF's actual results should (sic) differ materially. You should refer to the section entitled risk factors on pages 12 through 18 of PDF's annual report on Form 10-K for the fiscal year ended December 31, 2015, and similar disclosure in subsequent SEC filings. The forward-looking statements and risks stated in this conference call are based on information available to PDF today. PDF assumes no obligation to update them.

Now I'd like to introduce John Kibarian, PDF's President and Chief Executive Officer; and Greg Walker, PDF's Chief Financial Officer. Mr. Kibarian, please go ahead.

Thank you, and welcome, everyone. Today I'll start a discussion with a brief review of our previously stated plans for 2015 and then give you an overview of our business and performance to plan in the fourth quarter and full year of 2015. I'll then talk about our plans for 2016 and review how some of the major activities we are seeing in the overall semiconductor logic market may impact us.

Greg will then walk you through the financial results in detail and give some comments on our 2016 outlook. We will take your questions after that.

We began 2015 with a plan to achieve four strategic objectives: first, execute on key 14-nanometer ramps to maximize our Gainshare potential. Second, establish PDF Solutions as an integral part of the Chinese semiconductor ecosystem. Third, make major strides on delivering our Big Data analytics product, Exensio, across the semiconductor supply chain. And fourth, accelerate the development and market adoption of our new design-for-inspection solution or, as we call it, our DFI system.

The business activity in the fourth quarter and year demonstrate the success we have achieved on this plan. As to 14-nanometer, during the year we successfully enabled two key customers to reach production capability and booked a new 14-nanometer engagement. Also, as we anticipated, we received the first 14-nanometer Gainshare in Q4.

Highlights of 2015 for China business includes: a 28 nanometer yield ramp engagement with the leading foundry in China, which was signed in Q4; a 28 nanometer R&D engagement with another Chinese foundry, which had been signed earlier in the year; a 28 nanometer DFM engagement with a leading fabless company, signed in Q4; and an Exensio Big Data engagement with a leading foundry in China, signed in Q1. Additionally, we are engaged in many pilot projects and other sales activities in China. Other activity in the fourth quarter additionally included the following contracts for our base business: a 7 nanometer R&D engagement with a foundry client and a 14-nanometer DFM engagement for a fabless company.

Regarding our Exensio Big Data platform objective, we closed over 50 contracts in the fourth quarter alone and over 200 for the year. We successfully acquired Syntricity, which enabled us to extend Exensio to a SaaS model; and acquired the assets of Salland software, which expanded Exensio modules into test.

We now have over 90 Exensio customers. We estimate that over 10,000 engineers are currently using Exensio and that over 10,000 manufacturing tools are connected to the platform. Our Exensio customer base includes foundries, fabless, IDMs, test and assembly houses, and equipment vendors. For 2016 we anticipate further growth in Exensio across the semiconductor industry supply chain.

We also made great progress in our fourth objective for 2015 -- DFI. During 2015, we completed design, development, and buildout of our initial DFI tool and related IP. Additionally, we demonstrated our ability to inspect and analyze results from actual wafers for both fabless and foundry customers. The first major fabless company that taped out multiple chips based on our template DFM layouts and embedded over 7 billion on-chip DFI instruments successfully achieved foundry baseline yields on their first wafers out.

This outstanding achievement demonstrates that chips based on the combination of DFM template layouts and DFI came up the yield curve faster than other chips in 14-nanometer production. PDF's value proposition in providing a virtual IDM experience results in faster bring-up, better yields, and better parametric performance. Further, in 2015 we placed DFI instruments on 19 tape-outs at four major foundries and have received back and inspected demo wafers from three of those foundries. We have demonstrated the ability to perform electrical inspection and the technical viability of our DFI solution. We have, in effect, brought the benefits of electrical characterization to in-line inspection.

In summary, we executed well on all of our 2015 strategic objectives and have strongly positioned the Company for 2016 and beyond. For 2016, we will focus on the following strategic objectives.

First, continue to deploy our Yield Ramp solution for leading-edge fabless and foundries -- which, in 2016, we anticipate will expand into 10- and 7-nanometer, with early technology development work on 5-nanometer. We will continue to drive our Yield Ramps on 28- and 14-nanometer.

We believe 10- and 7-nanometer are really somewhat of a combined node, much like 20-nanometer and 16-nanometer. 10- and 7-nanometer will become much more important part of our solutions business during the year, while 14-nanometer volumes grow and drive Gainshare revenues.

The second strategic objective for 2016 is to drive adoption of DFI by increasing the number of chips with our on-chip instruments by installing early tools at multiple fabs and demonstrating benefits while continuing development of our system. Our DFI investments during the year will increase significantly as we get closer to market release. Last year we said we anticipated that DFI would have small revenue impact late in 2016 and more material impact in 2017. As we begin 2016, we maintain this positive outlook.

Third, expand our overall business activity in China, supporting the entire semiconductor ecosystem. Our activity in China in 2015 sets us up well for a great 2016 and beyond. As way of background, I should point out that 2016 is our 10th anniversary in China. More than 30% of our employees are based there and were initially focused on our back-office analytics. Late in 2014 we expanded our focus in China to become the cornerstone of silicon characterization for the Chinese semiconductor customers.

It is been well covered in the trade press that China has committed to investing in leading-edge silicon, starting with 28-nanometer. We are uniquely positioned to support China's desires in semiconductors with our strong local engineering team and worldwide, proven IP systems and software.

Our fourth and final objective for 2016 is to further assist our fabless and foundry customers in applying Exensio, our Big Data analytics platform, to maximize their production efficiency. For some of these customers, we will couple Exensio with our Characterization Vehicle test chips and other IP to further optimize their production. For foundries, we call this combination of CVs and Exensio for production control YieldAware FDC. We expect further adoption of this solution in 2016.

From an industry perspective, we are excited to see the beginning of 14-nanometer volumes, and we anticipate those volumes building throughout the year. While 28-nanometer volumes stabilized in Q4, we remain cautious about 28-nanometer volumes in 2016. However, beyond 2016 we see many reasons to be positive on 28-nanometer volumes and our related Gainshare, as we have a number of foundry customers that are just beginning their journey on 28-nanometer volume ramps.

I want to thank our stockholders, customers, and employees for their support of the Company over the past years. I look forward to working with you all to deliver and increase PDF's value and make 2016 and beyond even better.

Now we'll turn the call over to Greg to discuss in detail our financial results for the fourth quarter and 2015. Greg?

Thanks, John. As a reminder, in addition to using GAAP results when evaluating PDF's business, we believe it is also useful to consider our results using other non-GAAP measures. For internal purposes, the Company focuses on non-GAAP net income and EBITDAR. Non-GAAP net income excludes some nonrecurring items, stock-based compensation expenses, and amortization of expenses related to acquired technology and other intangible assets and their related tax effects, as applicable. Additionally, the income tax provision has been adjusted in our non-GAAP income to reflect cash tax expenses only.

EBITDAR is equal to earnings before income tax adjusted to exclude nonrecurring items, depreciation, amortization, and stock-based compensation. You can access the earnings release that contains a reconciliation of EBITDAR and non-GAAP net income to GAAP results in the investors section of our website, located at pdf.com.

Now let's turn to a review of the financial results. Total GAAP revenues for the quarter were $24.1 million, resulting in GAAP net income of $2.8 million and GAAP EPS of $0.09 per fully diluted share. Revenues on a non-GAAP basis totaled $24.5 million, with non-GAAP net income of $5.7 million or $0.18 per fully diluted share.

Cost of sales and operating expenses together were $20.7 million on a GAAP basis and $17.7 million on a non-GAAP basis. For the year, total GAAP revenues were $98 million, resulting in GAAP net income of $12.4 million and GAAP EPS of $0.39 per fully diluted share.

Annual revenues on a non-GAAP basis totaled $98.9 million, with non-GAAP net income of $25.6 million or $0.80 per fully diluted share. Annual cost of sales and operating expenses together were $78.7 million on a GAAP basis and $69.2 million on a non-GAAP basis.

Moving on to revenue details: as stated earlier, total GAAP revenues for the quarter were $24.1 million. This was higher than the previous quarter by approximately $200,000. Total non-GAAP revenues of $24.5 million for the fourth quarter were approximately $100,000 higher than in Q3. Total non-GAAP revenues were comprised of design-to-silicon-yield solutions, or solutions revenue, of $14.7 million and Gainshare performance incentive or Gainshare revenues of $9.8 million.

Our top 10 customers represented 85% of total revenues in the current quarter. Two of these customers contributed 10% or greater revenue for the quarter, for a total of 66% as compared to one customer and 60% in the prior quarter.

Looking at solutions revenue in more detail, 16 project-based engagements contributed at least $100,000 to solutions revenue in the quarter, two more than in the prior quarter. Primarily as a result of Q3's strong software revenue performance, Q4 solutions revenue at $14.7 million was $3.1 million lower than in Q3.

Gainshare revenue for the quarter was $9.8 million, an increase of $3.2 million over the prior quarter. The total number of node sites, which we define as an individual fab and process node combination, contributing to Gainshare was 18.

On a geographic basis, North America accounted for 44% of total revenues, which is down 10% from the previous quarter. Europe accounted for 28% of total revenues, a decrease of 2% from the prior quarter; and Asia accounted for the remaining 28% of total revenues, an increase of 8% over the prior quarter.

Total GAAP revenues for the year were $98 million. This was lower than the previous year by approximately $2.2 million. This variance in year-over-year revenues was generated by a reduction in Gainshare revenues of $13.3 million related to depressed 28-nanometer volumes, offset by an $11.1 million increase in solutions revenues, primarily driven by growth of our Exensio Big Data solution.

Total non-GAAP revenues of $98.9 million for the year were approximately $1.3 million lower than in 2014. Total annual non-GAAP revenues were comprised of design-to-silicon-yield solutions or solutions revenue of $64.8 million and Gainshare revenues of $34.1 million.

Moving to expenses: cost of sales for the quarter was $10.1 million on a GAAP basis, which is approximately $200,000 lower than in the previous quarter. This decrease in GAAP cost of sales was primarily driven by one-time hardware costs recognized in Q3 that were not repeated in Q4. GAAP gross margin was 58% as compared to 57% in the previous quarter.

On a non-GAAP basis, cost of sales was $8.9 million, which was $200,000 lower than the previous quarter. This decrease in non-GAAP cost of sales was principally driven by the previously mentioned hardware costs in Q3. Non-GAAP gross margin was 64% as composed (sic - opposed) to 63% in the previous quarter.

Total GAAP operating expenses at $10.6 million were $300,000 lower than last quarter and 44% of total revenues, down 2% from last quarter. R&D expenses totaled $5.4 million, $200,000 higher than the prior quarter. R&D expense as a percent of revenue was 22% in the quarter, which was the same as Q3. SG&A expenses totaled $5.2 million or 22% of revenue compared to $5.7 million and 24% of revenues in the prior quarter. The overall GAAP operating expense decrease was primarily driven by the reduction in acquisition-related expenses as compared to Q3, partially offset by increased R&D expenses, primarily driven by new hires.

During Q3 2015 PDF completed its acquisition of Syntricity, Inc., the industry-leading hosting solution for characterization and yield management, and that drove the acquisition of costs. On a non-GAAP basis, looking at operating expense and cost of sales together, total spending was $17.7 million, which was flat compared to the prior quarter.

Cost of sales for the year was $39 million on a GAAP basis, which was $700,000 lower than in the previous year. GAAP gross margin for the year was 60%, which is the same as in 2014. On a non-GAAP basis, cost of sales for the year was $36.8 million, which was $2.4 million higher than in the previous year. Non-GAAP gross margin for the year was 63%, which is 3% lower than in 2014. Total GAAP operating expenses for the year, at $39.7 million, were $7.1 million higher than last year and 41% of total revenues, up 8% from last year.

R&D expense for the year totaled $19.1 million, $5 million higher than in the prior year. R&D expense as a percent of revenue was 20% for the year, which is up 5% from last year. SG&A expenses for the year totaled $20.6 million or 21% of total revenues compared to $18.5 million or 19% of total revenues in the prior year.

In 2015 the SG&A expenses included on a GAAP basis $1.2 million of acquisition-related expenses. The overall GAAP operating expense increase was primarily driven by increased investment in R&D related to our DFI initiative and acquisition-related expenses.

On a non-GAAP basis, looking at operating expenses and cost of sales together, total annual spending was $69.2 million, which was $7.4 million higher than in 2014. The GAAP income tax provision for the quarter was $620,000, which reflects an effective tax rate of 18% compared to 46% in the prior quarter. This rate decrease is primarily due to the reinstatement of the federal R&D tax credit program, which allowed the Company to benefit in the current quarter from credits related to the full year of 2015 for qualifying R&D activities.

Cash tax liabilities for the quarter were $1 million. This represents an effective cash tax rate for the quarter of 30% of pretax GAAP income. As we have stated before, our cash taxes are primarily comprised of foreign withholding taxes. For the year GAAP income tax provision rate was 36% compared to 34% in the prior-year. The cash tax rate for the full year was 22% compared to 17% in 2014.

GAAP net income of $2.8 million for the quarter resulted in GAAP EPS of $0.09 per fully diluted share compared to $1.5 million or $0.05 per share in the prior quarter. This increase in GAAP EPS and net income is primarily related to the increase in Gainshare revenues in Q4.

GAAP net income of $12.4 million for the year resulted in GAAP EPS of $0.39 per fully diluted share compared to $18.5 million or $0.58 per share in the prior year. This decrease in GAAP EPS and net income is related to the decrease in full-year Gainshare revenues and also one-time costs incurred for the Syntricity acquisition.

EBITDAR, which I defined earlier and is also defined in our press release, was $7.5 million in the quarter as compared to $7.4 million in the prior quarter. EBITDAR per fully diluted share was $0.23, the same as in Q3. EBITDAR for the year was $32.5 million as compared to $40.5 million from the prior year. EBITDAR per fully diluted share for the year was $1.01 compared to $1.27 in 2014.

On a non-GAAP basis, net income was $5.7 million and non-GAAP EPS was $0.18 for the quarter compared to $5.8 million and $0.18 in the prior quarter. For the year, non-GAAP net income was $25.6 million and non-GAAP EPS was $0.80 compared to $33.9 million and $1.06 in the prior year. This reduction in non-GAAP income is primarily related to the year-over-year decrease in Gainshare revenues and increased investments in R&D.

Total cash at the end of the quarter was $126.2 million, which is a slight decrease of $100,000 when compared to cash at the end of Q3. Cash generated from operations during the quarter was $6.5 million, representing an increase of $1.2 million over the prior quarter. This cash generation was offset by $4.9 million of stock repurchases, or 450,000 shares, and approximately $1.2 million of fixed-asset and intellectual property purchases mainly related to our tester and DFI initiatives.

As mentioned earlier, the Company repurchased 453,000 shares of its common stock for $4.9 million. After these Q4 repurchases, up to $10.5 million remains available for use by the Company under its current Board-approved stock repurchase program. Included as a use of cash from operations, we also repurchased $800,000 of additional shares to cover employee taxes on restricted stock grants released -- about 69,000 shares. For the full year the Company spent $14.5 million on repurchasing 1.1 million shares under the Board-approved program and additionally spent $1.8 million on 133,000 shares to cover employee taxes on restricted stock grants released.

Trade accounts receivable DSO was 83 days for the quarter compared to 76 days in the previous quarter. Trade accounts receivable balance at the end of the quarter was $22 million, an increase of approximately $2.1 million over the previous quarter. The unbilled accounts receivable balance was $11.5 million, a decrease of approximately $1.1 million from the prior quarter.

Of the $33.5 million of total receivables, approximately $76,000 -- or less than 1% -- was more than 30 days past due. Since the end of the quarter, $4.3 million has been collected. Total DSO for the quarter, including unbilled receivables, was 127 days compared to 124 days in the prior quarter. Headcount at the end of the year was 390 compared to 391 at the end of Q3.

Looking at 2016, we expect the growth rate of solutions revenue to be greater than 10% year-over-year, driven by increasing demand for our Exensio software solutions and the continued expansion of the Asia semiconductor market and, in particular, the Chinese market. For Gainshare we remain very cautious with regards to 28-nanometer volumes for the year. While we expect overall Gainshare to increase during the year, driven by the buildout of the 14-nanometer volumes, the magnitude of this increase may be negatively impacted by the lagging recovery in the 28-nanometer node.

From a spending standpoint, we expect that 2016 R&D expenses and related costs of sales will increase by more than 30% as we take our DFI initiative from development to commercialization. Additionally, capital equipment purchases should also step up substantially as part of the DFI initiative. As we've stated before, we expect first revenues from DFI before the end of 2016.

In regards to other operating expenses, they will reflect the full year of the impact of the Syntricity acquisition. Finally, we expect our full-year GAAP tax provision rate to be between 38% and 40%, and cash taxes for the year should be in the range of 24% to 26%.

Now I will turn the call back over to the operator for Q&A. Operator?

(Operator Instructions) Jon Tanwanteng, CJS Securities.

Are you starting to see any rebound or backfill at all at the 28-nanometer node? And was the strength in the Gainshare in the quarter all 14-nanometer?

Jon, this is John. The growth was primarily 14-nanometer. We had really gone into the quarter expecting 28-nanometer Gainshare volumes to degrade Q3 over Q4, and that didn't happen.

Okay, so not as expected on 28.

Right, exactly.

Okay, got it. And then can you talk about those new 10- and 7-nanometer engagements? Are any of these with the client that you had a dispute with in 2014? Or, perhaps, have you made any progress in penetrating clients that you haven't historically done at all?

Yes, so I think we are always very careful about talking about who customers are, because we have to abide by our confidentiality agreements with customers. The 10-nanometer/7-nanometer engagement was with an existing customer. There are other contracts, particularly in China, that were with growing customers.

Okay, great. And then just, Greg, on the R&D comments, I just wanted to clarify -- did you say you expect it to be up 30% year-over-year, the R&D expense?

At least.

Okay, got it. Just remind me, also: what is the remaining share repurchase authorization? And is there any (technical difficulty)

Yes. I believe it's $10.4 million, but let me validate that for you. $10.5 million remains.

Dollars?

And -- yes. And no other updates on cash plans at this point in time.

Okay, great. Thank you very much, guys.

Tom Diffely, D.A. Davidson.

I guess first, on the Gainshare side, for the 14-nanometer Gainshare in the quarter, does that represent some catch-ups from previous quarters as well? Or is it just kind of the ongoing level?

No, that's just the ongoing level.

Okay. I mean, some of these customers have been doing 14-nanometer for several quarters. How does that play out as far as the royalty [share goes]?

Yes, we talked about -- with respect to one of the contracts, the first Gainshare quarter was measured in Q3 and invoiced in Q4. And then the six or so years that that contract represents on Gainshare started at that quarter. So it's just for -- prior production was lost.

Okay, good. And then on the 28-nanometer side, is it the fact that the previous 28-nanometer capacity is been partially moved to 14-nanometers? Or is that 28-nanometer capacity still there and just underutilized? How would you characterize it?

Yes, it's different for different customers. So I think you can really take it into three buckets, Tom. So the folks that have a gate-last 28 that is alternative to kind of the industry standard -- they are seeing demand for their high-k technology. On the gate-first technologies, we see that is very spotty in terms of the number of designs that have been put to that.

With respect to one of the customers that has a gate-first technology, a lot of their 28-nanometer capacity was retargeted to 14-nanometer.

Okay.

But they still have some capacity in 28-nanometer. In the case of other customers, they have a combination of gate-first in polysilicon. And we watch those two closely to see what the demand is for each one of those.

Okay. Over time, would you expect some of the older designs to migrate down to the 28? Or would you expect some of those 28 to get migrated down, actually, to 2014?

Yes, I think for the customers that are primarily in the foundry business, usually they don't migrate capacity. They -- more often than not, that capacity is left. So the 28 factories are, generally speaking, the last planar node factories. And if you remember the 0.18 micron node, eventually the foundries repurposed that 0.15, 0.13. But it was pretty much the last 200-millimeter aluminum node.

We believe 28-nanometer capacity for the most part is the last planar node. So you'll see it repositioned for FinFET -- I mean, for FDSOI technology. In one customer's case, in tweaks on 28, you know, slight shrinks; parametric tuned versions. You know, the foundries have now HPM, HPC, HPC plus. Many versions that are slightly improved. We anticipate that happening, but we don't expect any more capacity being transition from planar to 3-D technology, other than that one case that we saw in this past year.

Okay.

Did that answer your questions?

Yes, it does; that's very helpful. And then when you look at the 28-nanometer down the road, when the capacity comes back, or when your royalties come back, is the biggest driver for you just the ramp-up of the newer, largely Chinese/Asian customers? Or do you think that this existing capacity gets fully utilized again at some point in the future?

This -- we anticipate the capacity that exists being utilized both on 28 and derivative nodes like FDSOI. We also see an awful lot of capacity going in in Taiwan and in China on the second-wave foundries. And that represents a pretty big slug of volume, should they complete those shells, primarily targeted -- I mean, the shells are completed -- complete the capacity buildouts, primarily targeted at 28 nanometers.

So, you know, 28-nanometer -- for us, we still see a number of years of meaningful Gainshare. And so the pause in 2016 is really those newer customers are still kind of coming up on the volume curves. We don't expect to see -- we'll see some Gainshare from them in 2016, but see more as we get out to 2017 and beyond. And we also expect the customers that have more capacity still available to pick up as you get through design wins, which will take a good part of 2016 and really build into 2017.

Okay. And then the last question on 28-nanometer. You mentioned that -- it sounded like the 28-nanometer did not degrade as fast as you thought it would in the fourth quarter. I guess that means you're expecting a bit of a drop-off here, then, in the first or sometime this year?

You know, it's a great question, Tom. We proved to ourselves that weren't great forecasters in Q4, because we expected it to go down, and it didn't really. And so when we had modeled that in Q4, we had modeled that to stay at a depressed level throughout the first part of 2016 as well. And now -- you know, how do you change your model, right? Based on, you know, one data point, I can put any line through it I want.

So we decided to just remain cautious with respect to the 28 and watch how it goes out throughout the year. We'll figure out more over the next couple of quarters. But, you know, we don't have control over that.

Okay. And then moving over to Exensio. When do you think this becomes a meaningful part of the revenue stream? And at that point, what does it do to the margin structure?

Yes, I think as we went through 2015, it was becoming more meaningful and significant. As we go into 2016, I think you will start to see it becoming a more dominant part of the overall margin picture on solutions. Offsetting that will be as we move more and more work to the China and Asia markets. You typically don't see as great a margin there as you would on the rest of the world. But yes, as we move into 2016, the Exensio business becomes a very material part of the business.

Yes, I think one way to think about it, Tom, if I can add a little bit of color to that: Gainshare was down, what, $14 million year-over-year, right? And gross margins were basically flat. Normally, if we were to lose $14 million of Gainshare revenue and replace it with solutions revenue, you wouldn't hold margins.

Yes, okay.

So that's one way. You could go back and do your own modeling on that and figure stuff out.

Yes, okay. And then finally, Greg, on the R&D increase this year, the 30%-plus: what kind of linearity are you looking at for that?

Yes, my model -- which is no more accurate than anybody else's, because, you know, these brand-new development projects; they can go in spurts and fits. But I have a fairly aggressive ramp-up in the first half of the year and then flattening a little bit in the second half.

Okay, great. Thank you.

Drew Fanberg, Pennington Capital.

Just one question for me. You have quite a bit of cash on the balance sheet. I know you have a little bit left over on the current buyback program, but can you comment -- are there any future plans to return capital to shareholders in terms of either dividend or more buyback activity?

Yes, I think, Drew, that the amount left open on the buyback may be somewhat misleading. We've never actually gone through a period where we haven't had an open buyback. So while this one will expire, and it's set to expire in 2016, we will announce another buyback on the heels of -- you know, slightly before it will expire, which will refresh the amount of buyback we'll put in place in 2016 and beyond.

Okay, thanks.

Gary Schnierow, RiverPark Funds.

Can you guys characterize your DFI solutions -- where it fits in the inspection ecosystem? And also talk about how you are going to sell it, and if there's Gainshare tied to that? Thanks.

Sure, Gary, thank you. You made it so it's hard not to answer that question. You said, can we characterize -- which is, I guess, a big part of our business. So we have to answer that question.

So where does that fit in? What we believe we are doing is bringing an electrical measure in line. So it doesn't directly replace what they do, but it gives them a new piece of information, we believe, related to electrical performance on the part -- and, to some extent, potentially, reliability metrics.

With new 3-D structures, the way layers interact becomes more and more important. Conventional inspection kind of lets you look at the in-layer behavior, but it doesn't give you a good way of understanding how that layer is interacting with the layers below it. And we're really targeting this application to looking at the interaction between the layer that has just been completed with the layers below it, to look for defects that are basically covered and/or interactions of layers.

In terms of the business model, we really are still exploring that. So it's hard for us to kind of go and say what the business model will be for it. We do anticipate that the business model will somehow be ratable and be proportional to the amount of information and the way the system is used.

However, we don't really -- because there's an element that is a machine, but there's a lot of it, which is the IP and software. But without a signed contract that says the market also accepts the way that we are thinking about going to the market, it's probably speculative for us to say that's the way we're going to actually price it and bring it to market.

Okay. And you mentioned that you have, I think, 19 tape-outs for DFI?

That's correct.

So what does that mean, as opposed to being in-market, and those being contract -- you know, revenue-generating -- gain -- possibly ratable generating contracts?

Yes, so one thing I didn't bring up, point out -- I think we did something like 80 tape-outs last year in 2015 on test on CVs. So we tape-out more than a vehicle a week. Every vehicle, every chip in the world has dummy fill; test vehicles, product chips, every chip in the world does.

What we started doing with our fabless customers and our foundries is going back and saying, look, this dummy fill -- it's going to be wasted space. Let's use it productively. And we started putting -- wherever we could, given engineering constraints and timing, et cetera -- we started putting on-chip instruments inside fill for both test vehicles, fabless customers' test vehicles and fabless customers' full products.

It's like putting an instrument, then, inside potential customers' foundries. Then our fabless customers or us are able to go back to the foundries and say, oh, these instruments are running in your factory; we'd like to see how they perform. And we're able to then take wafers, bring them to our lab here in California, excite those instruments with this -- our own tool, read the results, use the Exensio analytics platform to characterize how that wafer behaved, and report that information back to the fabless and foundry about how you can get a better characterization of the design process interaction with this technology.

So for us, they are at this stage like selling tools, because they are our way of marketing what this capability can bring to people. And given we do so many tape-outs at pretty much all the foundries in the world every year -- you know, as I said, many more than one a week -- we always have an opportunity to in effect market the solution, provided we have the capacity to address the opportunity, right? We've got to be able to put the instruments in.

Got it, thanks. And switching to your Gainshare, can you characterize -- like, a few years back, I think you had a few peak quarters of $12 million of Gainshare. And at the time I think that was on 12 engagements. And today you're at 18 engagements, but I don't know if you are defining engagements differently. So how -- if on the 18 engagements today, and everything was a success, volumes and customers, can you talk -- what the dollar amount of Gainshare would be kind of compared to the $12 million on 12 engagements a few years back?

It's a difficult thing to say. Given our conservative view of the 28-nanometer volumes, we are modeling numbers that are not as high as that in any given quarter for the year. I won't give you -- you know, we don't give guidance, so I won't go beyond that.

But I think Gary's question -- maybe, Gary, your question is if all the factories filled up --?

Then we would be above that number.

We would be well above the previous peak.

Right. I mean can you give -- I assumed you were going to say that, but can you give a little bit more clarity? If it's 18 engagements versus 12 engagements, and I'm guessing your economics are better, and the volumes are bigger, that an engagement -- a single engagement today is better than a single engagement three years ago? Is that fair?

Yes, I mean generally. I think that we -- we don't measure the number. What we look at is just the number of wafers out there in the factories. And if the factories produced at full capacity, what would the dollars that would come out of each of the factories be? That's our kind of way of modeling it.

When you look at it under that model, the number of potential wafer output has grown from -- I think the peak you're referring to is a 2014 peak. So the number of wafer -- the factory capacity has grown over that time period for the most part. And so, therefore, the total potential volume it would represent and the Gainshare that would result from it is up substantially. I don't know the exact multiple over where it is over the previous peak, but it's a pretty sizable growth.

Okay, great. Thanks, guys.

Gus Richard, Northland.

You mentioned you had 19 tape-outs with DFI. How many customers does that represent?

Gus, I think it's somewhere around five or six fabless entities.

Okay. And then in the solutions business, can you give a little bit of color -- right now you've got a number of contracts. How many are at 10-nanometer and below? How many are at 14, and how many are at 28?

We generally won't specify those numbers, but --.

Are you talking about Gainshare or solutions customers?

Solutions.

Solutions.

So, yes, I mean, we have multiple contracts at 10 and 7, I think is probably the most clarification we would give you. And our older contracts at 45 and 32 are starting to roll off.

That's Gainshare, though. I think he was expecting --.

Oh, the -- yes.

Yes, I'm just looking at solutions.

So Gus, there's still a substantial -- there is, as you can (technical difficulty) the quarter, we are still signing contracts at 28-nanometer, right? I think one of the points that we've been maybe not doing a great job of communicating to the community: the 28 and 14 ramps (technical difficulty) we look at China and parts of Asia now.

So those are all starting all over again. So you're going to see us have activity on 28 and 14 for a number of years beyond this point. That said, there's a number of activities going on in 10 and 7 as well. So the number of nodes that we'll be engaged in in 2016, 2017, and 2018 is a very broad number of nodes.

Yes.

Okay, okay. That's helpful. And then I didn't catch the Gainshare at 28. Was that up, down, or sideways for the quarter?

It was primarily sideways for the quarter.

Sideways for the -- okay. So it has stabilized, but you do expect it to kind of stabilize to be weak through 2016?

That's how we're modeling it. The answer is we really won't know until we get into the year.

Fair enough. And then just real quick, your solutions business dipped in the quarter quite a bit, and you're guiding for the year to be up 10% or greater. And I just want to sort of understand what you expect that ramp to look like in 2016 in terms of your solutions business. Is it going to be a steady increase, or --? And what you see out there to give you comfort that you're going to have a pretty good step-up in the run rate off of Q4?

Yes, so to answer the first question, will it be a steady increase? No. Historically, our solutions business is fairly choppy and moves up and down quarter to quarter. Two reasons for that.

One is that on the Yield Ramp solution side, those are percentage of completion contracts, generally. And they tend to have very lumpy revenue recognition as different parts and deliveries on individual contracts are made and how much effort you've put forth to complete those.

And then on the software side, we're coming from a software business -- the Exensio business -- that historically had a significant portion of revenue that was perpetual and the annual maintenance renewals related to that. While that is shifting more towards a ratable model, in some cases, even a hosted model, there are still quarter-to-quarter impacts on occasion where you will get the lumps of revenue coming in, either catch-up on maintenance or a mix of perpetuals. And in those cases, the combinations of those two can generate you material variances quarter to quarter.

If you look at the Q3 to Q4 variance, that's exactly what happened. In Q3 we had a very strong software quarter, as some existing contracts that were fairly substantial actually got to revenue recognition that had been under some services that was being performed. And then in Q4 those contracts are not there. You're on your regular steady base. So that dropped off, and that's what drove most of the variance from Q3 to Q4.

We'll see that kind of activity up and down all throughout the year. But the overall direction is one where we're pretty comfortable that we'll be up at least 10%.

Perfect, thank you so much.

Jon Tanwanteng, CJS Securities.

Can you clarify that up 10% in the solutions business? Does that include the lump sum you received Q1 in 2015? Or is it up 10% from that? Or are you actually backing that out when you calculate?

Yes, that's very good question. I would count that as up 10% from the absolute number, including the $6 million.

Okay, great. That's very helpful. Thank you.

Rob Ammann, RK Capital.

Thank you. Just a clarification on the expense growth. That 30% -- is that just R&D? Or was that 30% growth in R&D and cost of goods sold?

That's just R&D.

Just R&D. Okay, okay, great. And then past 2016, is that a level that you think it plateaus and sort of holds there? Is there a chance it steps down? How do you view that longer-term?

I don't think it will step down significantly, but it will flatten out as we move out of 2016. This is R&D expending into 2017.

Okay, okay. And then the increased CapEx this year was, what, somewhere around $5 million or so? Is that something that you see a couple million extra, or any way to size that a little bit?

Yes, I would expect it to be a higher growth than we saw in 2015. And it could be materially higher.

Okay, thank you.

Andrew Weiner, Samjo Capital.

With respect to 14 nanometers, did you have one customer recognizing -- recognize revenue against, or was it two on the Gainshare side?

It was one this past quarter.

Okay. And then when you look at the second customer, do you have an estimate at this time when you'd expect to see first revenue from them?

We expect it in 2016, Andrew. But timing is always hard to say when they first start shipping for revenue. We believe they are at strong enough yields to.

Okay. In the beginning comments, did you say -- I may have misheard, but did you say that you added another 14-nanometer customer -- some sort of development deal?

That is correct.

Okay. So that would make -- we now have three customers at 14-nanometers at various stages.

That is correct.

Okay. With respect to the 28-nanometer and 14-nanometer ramps in China and Taiwan, to what extent -- you know, Greg, you had talked about potential margin pressure on solutions because of sort of their willingness to pay for, I guess, labor and that type of stuff. Is there a learning curve, though, you benefit from from the previous 28-nanometer ramps that you did for other customers that might allow you to have greater efficiency or productivity in getting customers in China up and ramped?

Andrew, I think the vehicles and systems -- I think my choice of words was worldwide, proven IP in systems. And I think there is tremendous benefit for our customers there and everywhere that our systems are proven against -- you know, we've run vehicles at every 28-nanometer node in the world, and virtually every 14-, 16-nanometer node in the world.

So there's a lot of validity in the structures and the capability, which does drive benefit to them and some efficiency to us. But you know, they all make different tool choices and different materials choices, and things are -- that you have got to run the program to run and characterize.

So the hours devoted to run up an engagement is relatively consistent across the customer base. A lot of that will be served out of our China office; and obviously, over time we have some efficiencies there.

Correct.

And then with respect to Exensio, you obviously mentioned you had 200 odd contracts for the year against 90-plus customers, which obviously means more than one contract for customer. Perhaps you could talk about how deeply you're penetrated within the existing customers from an Exensio perspective? And what opportunity is there for you to expand that business within the existing customer base?

Andrew, yes, so (technical difficulty) we have a varying degree of penetration in the solution. Only a handful of customers have the broad solution with the full analytics capability across all of the data types. And those are by and large the largest (technical difficulty) the modules that they've purchased, and then giving ourselves the ability to demonstrate the other capabilities that are out there.

So we believe that the customer base itself is a big upsell opportunity. And that's a significant part of our go-to-market strategy.

Okay. Greg, and then as a follow-up to the CapEx question, do you expect, particularly in the back half of the year, the CapEx associated with DFI -- will it be reactive to customer demand? Or do you anticipate the need to build tools in advance of getting interest in orders?

Andrew, it's John. Some of it is early buildout of initial tools for our lab and demo capability for second-generation tooling. Some of it is for customers. So obviously, if we don't see success throughout the year, we would fill up some of that buildout.

But as we said in my prepared remarks, we feel comfortable about achieving our goals of seeing customer impact by the end of 2016 -- financial impact by end of 2016, so -- with more material on 2017. So we would need to prepare in advance of that a little bit, which is what we've built into our operating plan.

Okay, great. Thank you.

Tom Diffely, D.A. Davidson.

Just a quick clarification here. So if a lot of the increase in R&D is for beta tools, Greg, could you put that in inventory instead of R&D?

The R&D expense is not really for the tools. You'd see that more in our capital line. There's some parts that get expensed that don't have commercial capability and things like that. So it wouldn't be in inventory. It would either be in capital for the moment or in R&D expense. And both of those will be going up.

Okay. That helps, thank you.

At this time, there are no more questions. Ladies and gentlemen, this concludes the program. Thank you for joining us today.

Thank you.