PDF Solutions Inc (PDFS) 2002 Q4 法說會逐字稿

Welcome to PDF Solutions, Inc.’s conference call to discuss its financial results for the fourth quarter ending December 31st, 2002. By now you should have received a copy of the corresponding press release. If you do not have a copy of the release and would like one, please refer to PDF’s web site at www.pdf.com, where the press release has been posted.

Please be advised that some of the statements that 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 solutions. PDF’s actual results could differ materially. You should refer to the section entitled Factors Which May Affect Future Results on page 20 through 28 of PDF’s annual report on form 10(k) for the fiscal year ending December 31st, 2001 and similar disclosures 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.

At this time, all participants are in a listen-only mode. Later we will conduct a question and answer session, for which instructions will be given at that time. If you need assistance during the conference, please press star then 0 on your touch tone telephone. As a reminder, this conference is being recorded.

Now I will turn the call over to John Kibarian, PDF’s President and Chief Executive Officer, and Steve Melman, PDF’s Vice President, Finance and Administration, and Chief Financial Officer. Mr. Kibarian, please go ahead, sir.

Thank you, and welcome, everyone. For the fourth quarter of 2002, we are reporting total revenue of $9 million and a pro forma loss of one cent per share. During 2002, PDF continued to create traction for its process design integration solutions. On our last conference call, we told you that by the end of the year we expected to enter into two more 90 nanometer contracts. We also said that we expected to maintain flat total revenues, even though gain share would be down sequentially. In fact, in addition to other new business, we entered into two 90 nanometer contracts in the fourth quarter.

As to maintaining total revenues, we closed business to achieve this during the quarter, but due to revenue recognition we missed our projections. The shortfall in reported revenue came about after our extensive review of one contract in particular.

I want to talk more about this contract and its importance to PDF. During the fourth quarter, we structured a comprehensive contract with a key player in the semiconductor business. This contract includes our patented characterization vehicle test chip, proprietary analysis software, our PD Fast Test wafer electrical test system, and a term license for a proprietary yield ramp simulator software as well as related support.

The license is consistent with our other engagements and engagement contracts, where the use of our technology is limited on a per factory, per technology node basis. Also license fees are, like most of our gain shares, dependent on wafer out. We believe this structure to be the most effective means to satisfy the needs of the strategic account while establishing the basis for a long term relationship.

We made a comprehensive and lengthy analysis of this unique contract under revenue recognition guidelines, which concluded earlier this week. Based on that review, which Steve will explain in more detail in a moment, we applied the most appropriate revenue recognition in accordance with our commitment to fiscally conservative policies. In spite of the negative impact on short term quarterly revenue, we are confident that the contract represents the best solution for PDF’s long term business and shareholder value.

As we step back from the specific business activity in the fourth quarter and look at the state of the industry, it would appear that the semiconductor industry has weathered the worst. Even though current volumes remain flat, forecasts are getting better. As anticipated in our last conference call, gain share in the fourth quarter of 2002 declined sequentially. We still expect downward pressure and gain share revenues in the first quarter, but expect gain share revenue in the second quarter to begin building on this modest base.

As it has been a relatively prolonged slump, we find that semiconductor companies continue to be cautious about their spending. As a result, we will continue to be cautious in our outlook. Steve will discuss our guidance for the first quarter of 2003 in more detail shortly. It is important to keep in mind that 2002 was the year in which PDF achieved key milestones. In 2002, PDF gained earlier market acceptance of our 90 nanometer solution by the semiconductor companies leading the development and production of this technology node. In 2002, we increased our penetration into foundries by entering into a multimillion contract with our second foundry customer, one of the world’s leading foundries.

In 2002, we successfully introduced our design based yield improvement solution and entered into multiple contracts. In 2002, we expanded our process design integration offerings to include PD Soft Test, a proprietary wafer electrical test system. And in 2002, we achieved profitability, positive cash flow and record revenue. We believe that the milestones reached in 2002 create a solid foundation for PDF’s future growth. We are excited about the opportunities out there and the market demand for PDF that we experience on a daily basis.

Now I will turn the call over to Steve, who will discuss our financial results for the fourth quarter and provide our projections for the first quarter of 2003. Steve?

Thank you, John, and good afternoon to everyone. For the fourth quarter ending December 31, 2002, revenues and pro forma earnings per share were below the guidance we provided in October, obviously an unexpected report. However, as mentioned in our press release, this shortfall was primarily the result of business judgment and a decision to choose what was in PDF’s best long term interest versus short term financial accounting results. The pro forma loss per share was one cent versus prior guidance of pro forma earnings per share of four cents. This shortfall, with expenses on target, was entirely the result of the revenue shortfall. These quarterly results reflect application of accounting standards set forth in the SOP 97-2, which pertained to software revenue recognition. These standards specify that revenue related to contracts closed in the quarter that included software license provisions different from PDF’s regular business arrangements be recognized in future periods.

We conducted a comprehensive review of all contracts closed in the quarter and related accounting literature and a final decision was made earlier this week not to recognize revenue until later periods. The total impact of this decision on fourth quarter revenue was to delay recognition of $2.1 million in revenue to future fiscal periods. This delay accounts for the entire shortfall from prior revenue guidance of $11 million. This revenue will be recognized over multiple years, after the software components are delivered. Revenue for the fourth quarter ending December 31st, 2002 was $9 million, a decrease of 16 percent compared to last year’s fourth quarter and a 19 percent decline sequentially from last quarter. The pro forma loss per share for the quarter was one cent versus earnings per share of four cents for both the comparable period last year and that reported last quarter, primarily due to the decrease in revenue.

Design to silicon yield solutions revenue for the fourth quarter totaled $7.6 million, a 9 percent decrease from the comparable period last year, and a decrease of 8 percent from last quarter. The decreases were the result of the previously mentioned delay in revenue recognition. Had we focused solely on short term revenue, design to silicon yield solutions revenue could have been up over both those periods. However, our priorities are PDF’s long term business.

Gain share revenue for the fourth quarter generated from three customers and four engagements totaled $1.4 million, a 41 percent decrease over the comparable period last year, and a 50 percent decrease from last quarter. As anticipated and communicated during our last conference call, volumes earnings remained anemic at the newer technology nodes, while our gain share contracts on older technologies for the most part have run their course.

We expect gain share revenues to remain near this level until the industry shows clear signs of a sustained recovery, something that we have yet to see on the horizon. As one follower of PDF said to me a number of weeks ago, we are standing on two stepstools. Under my left foot is a stool representing older contracts that are drifting into the sunset, and under my right foot is a stool representing technology nodes where volumes continue to be pushed into future periods. Needless to say, it is a challenge not to fall between these two stools.

While the industry puts us in this position with regard to gain share, we will continue to promote our infrastructure technologies and our business model leveraged to wafer volumes, and as a result, we believe we will be well positioned for the upturn.

Gross margin for the fourth quarter was 63 percent total revenues, a decrease of 3 percent from the fourth quarter of last year, and 4 percent from the 67 percent reported for the third quarter of 2002. We had held our pricing and margins on infrastructure implementations, but the decline in higher margin gain share revenue has taken its toll on overall margins.

Total operating expenses before stock based compensation were $6.6 million for the quarter, up approximately $250,000 or 4 percent from the fourth quarter of 2001, while remaining flat from last quarter. The increase from last year was primarily the result of higher research and development expenses partially offset by a decrease in SG&A expenses.

Research and development expenses for the fourth quarter increased approximately $675,000, or 20 percent from $3.4 million in the fourth quarter of 2001, while sequentially increasing just $100,000 or 3 percent from the third quarter of 2002, primarily due to unanticipated NRE. The increase from last year was primarily also the result of NRE for the development of joint products under our OEM agreement with a large semiconductor equipment manufacturer, and increase in personnel related expenses.

[Some] general and administrative expenses were $2.5 million in the fourth quarter of 2002, down 14 percent from the fourth quarter of 2001, while decreasing $150,000 sequentially from the third quarter. Pro forma net loss for the quarter, excluding amortization of stock based compensation and intangibles, totaled $264,000 or one cent per share. This compares with pro forma net income of $1 million or four cents per share for the fourth quarter of 2001. Unfortunately, this quarter’s results breaks our streak of six consecutive quarters of pro forma net income and four consecutive quarters of net income on a gap basis. However, we have met many fundamental business objectives, are making the right business decisions, and continue our proven track record of being a fiscally responsible company. We will continue to watch our costs very closely during these difficult times, and believe our return to profitability will return immediately upon a reasonable uptick in revenue.

For the year, covering a period characterized by the worst recession in semiconductor history, we are pleased to report overall revenues of $43.7 million, an increase of 19 percent over last year, and a record for PDF. We also finished the year with gap net income of $524,000, or two cents per share, and $71.5 million in cash, $655,000 more than we had at the end of 2001.

Looking at the balance sheet as of December 31, 2002, we continue to maintain our excellent financial position. Cash remained above $70 million, accounts receivable at $7.9 million. We have no bank debt and our current ratio remains over eight to one. Additionally, 22 percent of outstanding accounts receivables at December 31st have already been collected. And lastly, included in our accounts receivable balance is unbilled accounts receivable of $1 million, down from $1.6 million last quarter and again reflecting nothing more than the inconsistency between contractual payment schedules and percentage of completion accounting.

Before I turn to guidance, I will state again that some of the statements made in the course of this conference call, including the ones that we are about to make with respect to Q1 2003, are forward looking. These statements include expectations about future financial results and performance, growth rates, the success of any business objectives, product and service features and introductions, client products, and demand for PDF design to silicon yield solutions. PDF’s actual results could differ materially. You should refer to our current SEC filings and understand that forward looking statements and risks made during this conference call are based upon information available to PDF today. We assume no obligation to update them.

Now for the first quarter of 2003, we are providing guidance for revenue in a range of $8.5 million to $9 million, a result of continuing softness and gain share revenue, very cautious spending on new technologies in our customer base, and the impact of the contract that will delay revenue recognition well beyond the first quarter. We are also providing guidance for a pro forma loss in the range of four cents to six cents per share, reflecting projected revenue levels and product mix, and increases in some costs ranging from sales and marketing efforts to explored opportunities in new markets such as Korea and China, and personnel and other expenses typically associated with a new calendar year.

With that, I would like to turn the call back over to the operator to open the floor for questions. Operator?

Thank you. At this time I would like to inform everyone as to how to ask a question. You can press star then the number one on your telephone keypad. If you are using a speakerphone, please pick up the handset before asking your question. Please wait one moment for our first question.

Your first question comes from Garo Toomajanian.

Hi, guys. I guess I want to start off with a number of questions around this new foundry contract. I know that you’re probably limited as to what kind of color you can provide, but anything additional would be useful. For example, was there somebody in Taiwan, and I think that what I’m particularly interested in is what is the revenue recognition going to look like for this contract, and also just what the payment terms are going to be like as well.

Garo, this is John. I think Steve will answer the questions on the revenue recognition and the payment terms. You’re right about the color around the contract. Obviously the majority of the larger foundries are in Taiwan. So it would be a good, if you’re going to place bets on a probabilistic basis, it would probably be a good way to bet. But I don’t know that we can describe who it is at this time. And as far as the revenue piece, I’ll turn that over to Steve.

Yes, I need to expand a little bit on a statement I made a couple of seconds ago where I said the impact of the contract that will delay revenue recognition will be on the first quarter. What I really meant to say was that revenue recognition will extend for a period of time well beyond the first quarter. The fact is, we will not begin recognizing revenue under this contract until we deliver the undelivered software elements of the contract in 2003. Under the guidance of SOP 97.2 you will then take revenue on a pro rata basis over the term of that license. Now, from the, so it will be a couple of years of recognizing revenue on a pro rata basis.

From a payment term perspective, this contract is exactly like our other engagements. We have a fixed fee period, we’re installing our infrastructure implementation, we will be paid multi-million dollars of this fixed fee, and we expect to collect all money associated with the fixed fee by the early to middle part of 2003. Relative to the back-end license and the graduated scale payment depending on output, that will be invoiced and paid on a quarterly basis over the license period.

Does that answer your question, Garo?

Yes, it does. Can you give us a sense of the size of the deal? You did say multi-million dollar a couple of times. Is this maybe the biggest contract you guys have done?

I can answer that. I think we’ve done contracts with some of the IDMs recently that have been in the tens of millions of dollars. So this one is not as large as that. However, it’s scoped in one particular fab, and it’s in the multi-million range, under ten and much greater than one.

And the opportunity, of course, is to expand utilization of the technology beyond the specific fab.

Exactly.

And did you say that this contract did involve a gain share component as well?

In fact, it does not contain a gain share component. What replaces the gain share component is this graduated scale license that will extend over a period of time. It will be hooked to output and it could contribute a significant amount of money, which has the same feel as the gain share component relative to margin.

But the difference is, it’s not dependent on the actual yield performance of the customer site. In the nature of these kinds of businesses, the customer may have multiple products. In that case, from our standpoint, how you would measure the yield to cross all those and establish a baseline would be more difficult for what we’ve established here is something where it’s proportional to the factory’s output, but not proportional to the yield performance. And hence, it doesn’t get, like a royalty does not actually fall into the gain share bucket on our P&L.

So it’s not dependent on yield, but is it dependent on volume?

Yes.

Yes.

Okay.

Now, if the contract had a traditional gain share component, as opposed to a software license, we would not have run into the revenue recognition restrictions that we have.

Got you. You mentioned that there were three customers who generated gain shares. Were those the same three as last quarter?

I need to look at some statistics here. Yes, they were.

Okay. And we’re hearing at 90 nanometer that mask costs are going up above a million dollars. I’m wondering what you guys are seeing or what you would anticipate in terms of new design activity at that node with basically those mask costs so high. Do you think that will be restrictive, or do you think that it will mean that only designs targeted to super high volume applications would be applicable at that node?

That’s a very good question, Garo. This is John. You know, I’ve also seen data that I think takes, turns out a .13 design into a $30 million to $40 million activity. So I think people focus on the masks, because masks used to always just be this cheap, insignificant thing. And yes, the numbers at 90 nanometer is around a million and a half dollars, and the .13s are still in the mid-sevens, the 700K range. So the mask is becoming a more substantial dollar amount, but in terms of the total dollar amount, to get a design into volume, the numbers that folks are telling us are in the $30 million to $40 million range. A lot of that obviously outside the mask shop.

I think that is suggesting something that we felt for quite a while, which is, you’ve got to look at the markets that can support that much NRE. And of course, the consumer electronics markets have been at least through most of this downturn some of the stronger markets, because they can support that much NRE. And I think as we look at the leading edge technologies, it’s going to be more and more the consumer products. I think that drives the acceptance and demand of those technologies, because these huge NRE costs can be amortized over hundreds of millions of units.

Got it. Okay. And just one last question and I’ll let someone else jump in here. John, you mentioned last quarter that you think the volumes at 90 nanometer could actually come up at the same time as 130 nanometer. We have heard about yields picking up finally at .13. I’m wondering of you’re still thinking that these two nodes could actually ramp up at the same time.

Yes, I think that depends a lot on geographically. And in Japan, I think we did see that, where the IDM 130 are coming up fairly close. We look at our customers’ wafer projections on those technology nodes. There is a lot of similarity. They should be shifted off by a year and a half, typically. And they’re not anywhere near that far apart. And I think that, the crop that came out of Japan stirred a bunch of equivalent press in a lot of other parts of the world, North America and, particularly over the last couple days, there’s been nonsense about 90 nanometer adoption or shipping the 90 nanometer sample. And I anticipate that that means that folks will push to bring that technology out, just because it will offer, especially folks with 200 millimeter fabs, a way of being competitive with the 300 millimeter .13 product. So you may see that. I still anticipate that.

Okay, great. Thanks very much.

Your next question comes from Bill Frerichs.

Hi, Bill.

Hi, Bill.

It’s funny, I was reading the SOP the other day, I’m not sure why, but now I know. The difference that triggered the different treatment was the fact that there was not a gain sure element to this deal?

Well, the fact that there was not a gain share element and there was a software license that was not non-incidental to the arrangement, it triggered consideration of SOP 97-2. That SOP provides guidance for multi-element software arrangements, which include a service element and a software element together. And in our situation, we were providing a comprehensive solution with software that was not incidental to the arrangement, so we needed to look at it and we needed to provide the guidance. So specifically, we have undelivered software elements and PCFs. And it’s our obligation to provide vendor specific objective evidence as to fair value of these undelivered elements.

We could not do that on the software license because we don’t have a history of standard pricing of software components. As a result, the rules say that you cannot begin revenue recognition until you’ve delivered the undelivered elements, and then you need to take the revenue in a pro rata portion over the life of the undeveloped, undelivered elements. So specifically answering your question, Bill, yes, the fact that we didn’t have a gain share component and we have a software license arrangement drove us to 97.2.

Okay. Basically, the theory behind this, so that I can re-explain it to other correctly, would be that since this is a foundry, the foundry has essentially different needs from your IDM customers, correct?

If I can jump in here, Bill. In our IDM customers, we use that same software during the design to silicon yield part because we usually have one product that is the bulk of the volume in the factory, there’s often a lot of other products. So the software, the yield ramp simulator is really the software that is more than incidental, as opposed to some of the analysis software that works with the CVs, characterization vehicles.

So for the IDM, we’re going to use that technology during the design silicon yield solution period, and then after that, when they’re running the factory, they may continue to use the characterization vehicles. But we typically take that yield ramp simulator away. And it’s okay in their case, the part counts are small. In the business arrangements where you would look at the foundries, they would have multiple products coming in. In fact, they ramped again almost without a product. And then there’s many products that come in as you go down through the production life.

So the importance of that software for that part of the industry is actually higher than it would be in a place where there’s a lower part count. And the need to incorporate that in our solution is part of what the customer gets to use, and the nature fundamentally would change in the way we look and account for the contracts. If that makes sense, that’s the techie reason why.

Well, that’s pretty important. And I think also that the trailer, if you will, must be a wafer based royalty, then, right?

Yes, the license is tied to the wafer output of the factory. The value of the technology, the value of our characterization vehicles, of our tester, our software, is really tied to how much volume they run through that factory, and hence the license amount is tied to the output of the factory. The license dollar amount that we receive is tied to the output of the factor.

Okay. And is the rest of the contract deliverables, are those on project accounting basis?

No, they’re not. That’s the whole issue with this SOP. If there was clear vendor specific objective evidence of the undelivered elements, you could bifurcate the contract. You could do percentage of completion accounting on the beginning and you could do variables on software revenue recognition at the end. As a result of not having vendor specific objective evidence on the undelivered elements, you don’t get to do percentage of completion accounting at the beginning, and you cannot recognize any revenue until you first deliver the undelivered elements.

So should we therefore then expect deferred revenue to build sequentially throughout the next year or so?

It will build as we invoice the customer, right. At that point you will see a blip in deferred revenue on our balance sheet.

Okay. And this quarter was roughly flat, and therefore you haven’t billed them yet?

We have billed an amount, yes. That was, I’m not sure of the exact numbers off the top of my head. But it’s flat because the normal degradation or the normal amortization of the deferred revenue back to the P&L was offset by the increase as a result of invoicing against this contract.

Okay, but I would expect at the end of Q1 to see a higher number deferred, correct?

Yes.

Okay, thanks a lot.

The next question comes from Gus Richard.

Good afternoon. I just need a little more color on the foundry contract. Is this for a 90 nanometer process 30 millimeter?

Thanks, Gus. This is John. No, we describe the technology nodes that we signed contracts up with and then something about the types of customers. But at least until we can make a press announcement with respect to the customer, we can’t describe the technology node that it’s associated with.

Okay, and can you talk just a little bit about the sales pipeline and what you see as far as new customers and contract in the coming quarters?

Sure. We did sign up two new customers, actually, in the last quarter. We have a number of sales things in the sales pipeline this quarter, including both existing customers and new customers. As I said in my color, we do see customers sharpening their pencils and being very concerned about how their spend is and what things, how they’re going to get a return on their investment. We feel our technology that we’ve been able to demonstrate to customers affords a wonderful return on investment for our customers and clients.

When we look at this quarter, we expect that we will see a lot of new, a lot of the existing customers building business, partly because some of those existing customers are in Japan and Q1 and Q2 are when you do expect, their fiscal or annual is April 1, so you do see activity in this quarter in Japan typically, and we do see that today. Plus, Japan is probably one of the markets from a capital equipment perspective where you do see spend.

We do also see customer activity in Europe and other parts of Asia, Taiwan and Korea, as well as in North America. And we have activities ongoing in all of those places. We have made an investment in sales in the last couple of quarters, in terms of staffing, and we are seeing some of the benefits of that.

Thank you.

Your next question comes from Dennis Wassung.

Thank you. A couple of quick questions. First, one more question on the foundry contract, I’m sure you’re tired of talking about it by now. But you talked about $2.1 million that would have fallen into the Q4 results. I’m assuming that that’s assuming the contract would have been treated as a percentage of completion up front, with gain share later. And that’s kind of the equivalent of what would have fallen. Is that an accurate way to think about it?

Yes, if we would have had percentage of completion accounting, the majority of that number would have resulted from this contract.

Okay. So if it’s fair to say that you’ve done a pretty substantial amount of work on this already in terms of implementing some of the solution with the customer?

Yes, we have.

Okay. And secondly, in terms of the two 90 nanometer deals that you’ve completed in Q4, is there any other color or information you can give regarding those two deals?

Actually at this time I don’t think that we can. We do anticipate in both customer cases being able to talk about them as we get through 2003. And in one customer’s case, we do know that our activity with them is something that they’ve talked about with their customers as being their way of leading the way with this technology node and anticipate being able to talk more about that. But at this point we can’t.

Okay. Can you say whether or not those were new customers or existing?

I believe it’s one of each.

Yes, that’s correct.

Okay. And lastly, I was just wondering if you could give an update on the PDFast Test product, how that’s been received. It sounds like it’s being at least coupled into the foundry contract. Any other installations there? And if you can talk a little bit about the progress of that.

I think there have been more than a handful of installations at this point at existing customers. Some customers actually looking to have multiple units. We’ve gone through a lot of the working of the application with the customers, customers seem to be very happy, among some analysts, from different customer things, I’ve seen very positive e-mails from the customer base about how much they love it and what it does for them.

So we are very happy about what performance we see with PDFast Test, and we think it’s a really important part of a total solution, just because of the difficulties in quantifying the failure rates and the characteristics, the yield characteristics of 90 nanometer technologies and the time required to do that. It really does address a big need for the customer.

Okay. And actually, one quick follow-up on the foundry contract, just want to make sure I get this straight. One of the things you talked about early on in the call was, I believe this is something to do with, it’s a contract that’s per technology node and per fab? Is that fair, or did I misunderstand that?

Thanks, Dennis, let me make sure I make that clear. So all of our contracts, the technology that we provide, the test vehicles, the software, the tester, are licensed on a per, typically I think on a per process, per factory basis. So you may apply this vehicle from this software in your factory in San Jose, I know it’s not here any more, but at your .13 node, for example. And if you were to, say, want to use this same technology at your fab in Milpitas, you would need a new license, or if you were to bring in the 90 nanometer process technology, you would need a new license to use our technology on 90 nanometer, even if it was in the same factory. So this contract has the exact same characteristic that we typically start for in our contracts that are more traditional.

Okay, so no difference there?

Right.

Okay. And lastly, if you could just talk about the overall environments, what changes you saw in Q4 from the customers in terms of their willingness to go forward with new projects or cancel projects, if you could talk a little bit about that, that would be helpful.

Sure. I think we have seen some of the leaders make a move on 90 nanometer, and we see it more in the process as well. That we anticipated on the Q3 conference call as happening in Q4. I think we now have three 90 nanometer contracts in place.

With respect to the overall environment, we still see customers very cautious about how they spend. The spending does seem to be technology related, you know, new technologies more than existing technologies. And they will make investments in process technology for next generation processes, we do see that as well. But we’re not seeing a tremendous amount of legacy, you know, older, people wanting design based improvements on .18 technologies. That’s mostly on the newer technology nodes as well.

So people are very cautions. We do see strategic spins, and I think that’s what we’re going to see at least in Q1.

Okay, thank you very much.

Your next question comes from Joan Tong.

Good afternoon. I actually still have one question regarding the foundry structure deal. I just want to see, is this the first time you guys structured a deal like that? Did that happen before, or do you actually see more deals in your future are going to be structured in this kind of way?

Thanks, Joan, this is John. This is our second contract with this foundry, however, the first one in this characteristic. So obviously a very strategic account, we spent a lot of time working on this for months, for a good part of 2002, to determine a good relationship for them and for us, a good start of a relationship for them and for us. There are few accounts that are as strategic as that account. I don’t know that we would apply this same type of contract unless we had an account that was that strategic again and had the right mix and characteristics.

So I think we are not necessarily going to apply a contract exactly this way again. However, we did learn a lot about elements of this contract that make sense with our traditional contracts, and we will probably add. We felt for quite a while that irrespective of the yield the customer ultimately achieves, the customer gets value out of our technology. And we wanted to be able to monetize the value as a function of what their wafer outs are. And we’re seeing, we think that is an element in this contract that we think is an appropriate one for all of our customers. And we may bring that into the contracts on an ongoing basis.

Actually, I just want to get it straight. I think Steve mentioned about the fixed portion or the fixed fee for this particular contract. Does that $2.1 million he mentioned about represent the ball park of this fixed portion?

Yes, the fixed fee is greater than that. It’s quite typical to the normal guidelines, you know, that we’ve always talked about relative to the fixed fee portion of the normal engagement. But quite candidly, I can’t get into the specific terms of the contract.

Okay. Regarding the gain share revenue, I believe you said that there are actually three customers contributing to gain share revenue. How many engagements are we talking about?

Four engagements.

Four engagements. Any engagements going to be, are entering the gain share revenue phase in the next quarter or even the June quarter?

I think as we have looked at it, we expect Q1 to be a lower point in gain share. As I said, I would expect modest improvements in Q2. Part of that is because I think there are some newer contracts that are coming into modes. We’ve also seen some customer estimates in the outer quarters about their volumes that are newer technology nodes, primarily 130 and 90, that look somewhat encouraging. That would happen in the latter part of the year. Of course, as we recognized, typically we would have some disability into that.

But I think it’s too early to really say if those customers are, got solid forecasts or if those forecasts are less solid and so we’re still waiting to see how those work out. However, it is correct to assume that in Q2 we expect to see a new contract contribute to gain share, as I remember looking at the forecast.

Okay. And then you mentioned about two new customers, you added two new customers this quarter. Are there any danger elements included in the engagements?

One yes, one no.

Okay. Thanks.

Your next question comes from Erach Desai.

Hi, Erach.

Hi, how are you?

Good, how are you doing?

Okay. I have two sets of questions. I guess one is that your gain share was, at least by my model, and I know what you had guides that suggested it would be down sequentially in the fourth quarter, but down 50 percent sequentially was a little more than I was thinking of. Even if, let’s just say that you’d been able to recognize this deal, it closed, I’m guessing, some time in December, it would seem that both a drop-off in gain share and the environment caught up with you guys in December a lot more, above and beyond this deal. I guess what I’m trying to say is, then we were probably seeing your first quarter guidance for revenues pretty much stay intact.

This is John, I’ll answer that. As we said on the Q3 conference call, as we guided we anticipated a drop-off in gain share, I think it was drop-off that we did eventually see was in the range, maybe a little bit more extreme than we expected. But close to where we expected it to be. This contract was signed in the quarter, this had been something that we had been working with this customer for quite a while on, and the activity for this contract was begun earlier in the quarter and would have, on the percent completion accounting, would have been recognized over that entire quarter’s activity.

I think for sure if the gain share had come in, we are always beholden to how game share comes in. And if it comes in stronger, it certainly would have made up for a bigger slice. This contract was in this quarter a pretty substantial, it’s going to be a pretty substantial amount of revenue. When we look into Q1, if you look at Q1, our anticipation on Q1, as we’ve guided Q1’s gain share we’ll be down over Q4. What that basically says is, if you look at the Q3, Q4 and Q1 D2SY numbers, they’re basically flat. And if you were to factor back in this contract, they would be trending up.

So I think that the growth in the business, or the adoption of the customer is actually still going on. We are for sure caught up, the length of the gain share contract, coming to the end of the line for the gain share contracts, and some of them just, the volumes or the yields coming to their ceilings or the ends. We are of course seeing the gain share piece affecting us in Q4 and Q1. I think you’re correct in seeing that.

We thought we had anticipated that when we had guided in Q3, because of what we saw as substantial adoption, substantial adoption in that quarter from this contract as well as the other new contract.

Erach, when I spoke to that specifically in my monologue there about design to silicon yield solutions and I talked to a 9 and 8 percent decreases in the comparable period last year and last quarter, I had indicated, had we concentrated solely on short term revenue as opposed to long term business objectives, the design to silicon yield solution or the adoption of our infrastructures in fact could have been up over both those periods.

Gentlemen, don’t get me wrong. I’m all for being conservative on the revenue recognition side. I’m just trying to understand just the overall macro environment, sort of a little bit about your pipeline and just, you know, the activity level. So that’s a fair answer, and again, I appreciate the fact that you want to be careful rather than give away the future.

Let me ask a separate question on the deal. And that is, are you, John, is PDF free to pursue sort of an incremental, per unit, per performance deal with a particular established customer who has a ramp going through this foundry, in that that’s where you might be able to capture something that is more tied to how you deliver performance? Can you do that? Is that something that you’ve let yourselves structure room for, and is it possible to structure?

It’s a good question, Erach, and let me answer that as well as talk about the macro environment. I want to make sure that you and I have a, I’ve communicated well to you about that point.

Yes, we have left ourselves, and in fact, I’ve openly discussed with the customer the ability to go back and do that. And in fact, as we structure a number, we talked to, we want to structure a number of contracts for foundries. This is something that we believe is in everybody’s best interests, them as well as their customer. And of course, as well as us. We would like to have endorsements of the factories for our solution. And that’s something that at this point we don’t have, as we have not delivered our solution, we are in the process of delivering it, not completed delivery of our solution yet. We will look to establish that as we go out into the future, which I think will make the ability to add on business on top easier, although we are allowed to do that and can always do that today. We can sell to folks who use those factories to bring their parts in.

Now, as far as the macro, I think whereas you look at it from, we’re looking at it from the what did we think about when we gave guidance in Q3 and what did we see when we provided our results at the end of Q4. So from our perspective, we had anticipated a significant amount of that gain share down, roll-off, and we anticipated where we would see build-up of contracts. We think if you look at the numbers, the gain share did do, unfortunately, what we thought it would do. And the D2SY, in terms of bringing in the business, did do what we thought it would do as well, it did not contribute to the revenue. In fact, D2SY is, you know, modestly down because of the way we recognized the revenue. That’s what we anticipate in Q1 as well. If I said it in a more tight fashion, then.

Thank you.

You have a follow-up question from Bill Frerichs.

Steve, on your more normal contract accounting deals, what percentage of the total value of the deal do you typically take the revenue up front?

What percentage? I think the design to silicon yield solution from a fixed V perspective is, you know, it has historically been 75 percent of our overall revenue. But you know, I guess I’m getting stuck here a little bit between the other contingency of the gain share and the fixed nature of the – I guess I’m not –

This is John. Can I ask for a clarifying further question? What you’re saying is, what percentage of the fixed fee do we typically recognize up front? When you say upon signing the contract?

Yes.

If that’s what your question is, let me see if Steve can answer.

Oh, yes, I’m sorry. We don’t recognize any revenue upon signing a contract. The infrastructure implementation, you know, the definition of percentage of completion accounting is as you deliver the infrastructure and the support services, you recognize the revenue as a proportion of that technology and services as you deliver it. So we don’t recognize any revenue simply for signing the contract.

And when is the first milestone, typically?

Well, the first milestone from a billing perspective, or the first milestone from a revenue recognition perspective?

How about both?

Okay. The billing perspective will depend on the contract. Some customers like a monthly billing, some customers like a quarterly billing. From a percentage of completion basis, it’s every month as you deliver against the contract.

I guess is there anything about this $2.1 million, just to be blunt, that was not recognized in this quarter that could have been that gives us a clue as to overall size of the contract, other than it’s more than $2.1 million?

I think the contract parameters of the infrastructure implementation are typical with PDF’s normal infrastructure implementations. The opportunities at the back-end relative to usage across multiple fabrication facilities certainly offers a very broad opportunity.

And you seem to be implying that you elected to take this accounting treatment in the interest of a long term relationship. But aren’t the rules just the rules?

Bill, I think the issue was, we made a business decision to include in the overall offering non-incidental software. Therefore, first, it took us from contract accounting to I guess what’s now, I guess I can speak on this, SOP 97.2 software recognition. Then we took extensive analysis of what does that mean in terms of how we did that. When we made that decision, we decided that that was the right thing to do for the business. Because they needed, the fundamentally need the software component. We knew that that would, we would help you go back in and analyze the contract in more detail.

The scope at which we did the analysis and the depth of that had to go on was, I think, unanticipated.

And on the accounting front, the rules are rules, we abide by the rules. The business judgment is where we have the obvious flexibility.

Okay. Well, actually, congratulations. No one has congratulated you. It’s a big deal.

Thanks, Bill. We thought so, too.

You have a follow-up question from Dennis Wassung:

Hi, just another quick question to follow up on Bill’s line of questioning. Back to the $2.1 million, I guess, and John, you also mentioned that under the standard treatment of this contract, if it was treated in your typical way, you would have sort of been in line to put up D2SY revenue that would have kept you at a fairly flat revenue trajectory. Is that relatively a safe assumption here?

I think what I said is it would be up. The exact percentage I don’t know how much it would be up, because I haven’t really looked at it, frankly. We’ve been looking at the revenue in Q4 as it actually, you know, on a gap basis. When we gave the guidance in Q3, with the business that we, when we gave the guidance at the end of Q3, the beginning of Q4, with the anticipation of the business that we saw in Q4, which included this contract as well as other new contracts that were signed in the quarter, we had an anticipation of it being the D2SY revenue being up, the gain share revenue being down, and for us being mostly flat. And what ended up happening is the gain share revenue was down, the D2SY revenue turned out to be down, and that being based on how we, as we sort through the contracts, how we ended up interpreting them.

Okay. And I guess one more question along the line of this contract. Did it sort of fit the same parameters of your normal, I guess 9 to 12 month implementation process?

The implementation process is actually a little bit shorter than the typical. The typical are 9 to 12 months, and we timed closer to the 12. This is probably turning closer towards the 9. Because our obligation is to deliver the technology and help the customer apply it. We are not, as I said at the beginning, on the hook for yield, as we are in our typically engagement. So the activity, a lot of the activity is developing the vehicles, installing the vehicles, helping the customer use the characterization vehicle. And doing some amount of interpretation, but doing far less than we typically do in terms of making improvements in the technology. That’s why we’re not, our license, our back-end of the contract is not dependent on the actual yields achieved.

Right. So, but it’s safe to say, I guess if this had been treated along the lines of that $2.1 million, if that was treated in Q4 similarly, we’d have some amount, whether or not it’s $2.1 million or some number close to it or not close to it, you’d have some amount that would be recognized in Q1, just as another section of the contract was completed?

That’s correct.

Okay. And last quick question here, in terms of the delivery of the software that sort of kicks off the revenue recognition for this contract, what is the anticipation there? Can you talk about that in terms of when will that software be delivered? And then when will you be able to start recognizing revenue?

We anticipated, I think, I think the implementation phase of this contract was anticipated ending in the Q2 time frame. I think it will be installed somewhere in Q2. I think at that point the revenue would be recognized ratably on it, ratable quarter basis over the remaining period of the contract, which we said is multiple years.

Okay, so the revenue recognition is hinging upon the software installation of the yield ramp simulator, essentially?

Right.

Very good. Thank you again.

Thank you.

Your next question comes from Gary [Shinaro].

Hi, Steve, Hi, John.

Hi, Gary.

Congratulations, seems like a pretty watershed event for you. And I was wondering, in selling additional customers, clearly they want results that are going to make the big difference. But does just this event make a difference to customers? Or do they really want to see results in nine months to a year to be convinced?

You know, I always joke with customers that the physics are the same all the way around the world, so if it works here, why doesn’t it work there. But you find that folks need to see things for themselves. So I think every time you have another significant customer, especially in the future when we can make press announcements about all of our customers, it does have a positive impact and will get people to think, if this technology makes a difference for them, it must make a difference for us as well. But they’re going to want to see that it really does make a difference for them as well. And I think we always anticipate that, and always expect that through our selling process and through our delivery process, and in fact, for the entire relationship with the customers. It’s always, what did you do for me lately, basically. I think that’s what the business is all about.

In terms of your design to silicon yield solution revenue, it declined a little bit this quarter and your guidance is for it possibly to decline a little bit again. Could you just give a little color about engagements ending and that sort of thing?

Yes, the specifics of the engagements ending, that’s a very good question, Gary. We always have in the quarter some ending. I know that there are some ending in Q1 that would then go into a mode where we would be waiting for, anticipating gain share. We see about the normal number in every quarter. There may be a little more found in Q1 because of the Japan fiscal year’s roll-around on April. But I don’t think it’s anything particularly unusual.

I think I would add that inherent in our cautious outlook is the fact that our typical contract is percentage of completion accounting. And like Bill asked before, we don’t sign a contract and recognize revenue. We need to deliver the technology to provide our services. So any protraction of the selling cycle delays a contract signature and delays our ability to implement our infrastructure. So it is inherent in our cautious outlook that our customers are watching their pennies and dollars very, very carefully.

And looking at balance sheet, other assets jumped quarter over quarter. What’s in that number?

It’s a deferred tax asset in there. And now there’s actually a short and long term deferred tax asset.

Okay, thank you very much.

Thanks, Gary.

Your next question comes from Greg Weaver.

Evening, guys.

Evening, Greg.

Just a follow-up on the last question. How many deals do you need to close for D2SY in a given quarter to keep your revenue at this level? Obviously they vary in size, but as a general yardstick.

I think it’s on the order of, you know, couple of deals per quarter, typically. And I think that’s about where it’s been.

Okay, so you’re on par for this past quarter then?

I think Q4 was, maybe there were three significant deals that closed that I can think of. One existing customer.

Okay. And how does this revenue recognition issue affect the Tower semiconductor deal? Do you take away the yield ramp simulator piece of software from those guys every time they do one design?

That’s a good question. I think that we delivered and signed up Tower in 2001, that contract, there was a lot smaller software component to it. I think we are in discussion with them about things in the future. And I think there is a general need. Most of the foundries are going after a small number of larger accounts. When we sold it, we may have anticipated that, as opposed to the more established ones where there’s a broader number of customers.

But I think it’s a good question. I think it is something that we are looking at as we expand our business.

Okay. I guess the last question was, on the gain share being down 50 percent sequentially, is this strictly a volume thing, or is there a per unit revenue decrease as production process matures?

I think there’s a couple of components here. One is a contract coming to the end of life. The number of contracts in Q4 I believe is slightly lower than the number in Q3. Second, the volumes that customers do vary on a quarter by quarter basis. Third, the yield against baseline, baselines go up especially toward the end of the contracts at times. And hence, although the yield that the customer is performing at may be flat, the delta between the baseline and that number may come down and you get it the same way for volume if the dollars change. All those things will wiggle around the gain share number. The contract is coming to the end, of course you have the best visibility on the volumes and the yields. Frankly, we have day by day visibility on it.

Okay. That’s helpful. Thank you.

At this time there are no more questions. I will now turn the call back over to Mr. Kibarian for closing remarks.

Again, we are excited to see more significant interest from customers early in the adoption cycle of new technologies. We continue to experience valuable success at customer sites. We also clearly see the long term opportunity for PDF, the process design integration company. Thank you for participating in our fourth quarter 2002 conference call. Goodbye, everyone.