Peraso Inc (PRSO) 2003 Q4 法說會逐字稿

Good afternoon ladies and gentlemen. We are now ready to begin the fourth quarter 2003 earnings conference call of Monolithic System Technology. I'll now turn the call to Beverly Cling [ph] of Shelton, the investor relations firm of record for MoSys. Beverly, over to you.

Thank you Jean. By now everyone should have received our press release. However, if you haven't, it is available on MoSys' website, www.mosys.com. Before we begin the discussion of the fourth quarter and yearend results, I would like to acquaint you with our forward looking statement. The discussion during this conference call will contain forward looking statements within the meaning of the Private Securities Litigation Reform Act of 1995, which include without limitation statements about the market for the MoSys technologies, benefits and performance expected from use of the 1T-SRAM technologies, licensees of 1T-SRAM technologies and their strategy, the development and production of products that use Mo-Sys' license technologies, license fees, and royalties attributed to 1T-SRAM technologies, and the company's anticipated or perspective financial performance. Any forward looking statements made during this call are subject to risks and uncertainties that could actual results to differ materially from those projected. Additional information concerning factors that could cause actual results to differ materially from any forward looking statements made during this call are contained in the company's most recent annual report on form 10K filed with the Securities and Exchange Commission, in particular in the section titled Risk Factors in the form 10K and in other reports that the company files from time-to-time with the Securities and Exchange Commission. Mo-Sys undertakes no obligation to publicly update any forward looking statements for any reason except as required by law even if new information becomes available or other events occur in the future.

Thank you for your attention. I would now like to turn the call over to Dr. Fu-Chieh Hsu, President and CEO of Mo-Sys.

Good afternoon everyone. Welcome to the Mo-Sys fourth quarter and yearend earnings conference call. With me today are Mark Voll, our CFO, and Mark-Eric Jones, our VP and General Manager of intellectual property. We will begin the call with Mark Voll presenting a brief overview of the financial results for the fourth quarter and the fiscal year ended 2003 as well as providing our business outlook for the first quarter of 2004 after which I will discuss business and technology highlights of the fourth quarter. We will have Mark-Eric Jones give a brief overview of our customers and the market. Finally we will conclude by answering any questions you may have. Now I will turn the call over to Mark.

Good afternoon. Today we reported financial results for the fourth quarter and year ended December 31, 2003. Here is a summary of the results. Net revenue in the fourth quarter was 3.4 million, total net revenue including licensing revenue of 1.9 million, royalty revenue of 1.1 million, and product revenue of 334,000. Licensing revenue in the fourth quarter was 1.9 million compared to 3.3 million in the fourth quarter of 2002. Licensing revenue was recognized from 21 different chip development projects this quarter compared to 20 chip development projects in the fourth quarter of the previous year. Six of the chip development projects were new this quarter. Royalty revenue in the fourth quarter was 1.1 million compared to 4 million in the fourth quarter of 2002. The decrease is attributable to the absence of GAMECUBE royalties that were approximately 3.2 million in the fourth quarter of last year. During the fourth quarter of this year we earned royalty revenues from eight different licensees most having multiple SoCs in production. We recognize royalty revenue from reports provided by licensees, which are typically received in the quarter following that in which the licensee has sold or manufactured products containing our 1T-SRAM technologies.

Product revenue was 334,000 in the fourth quarter compared to 599,000 in the fourth quarter of last year. As we have stated throughout this past year, we continue to see limited demand in the communication markets for our stand-alone memory chips. We believe that it is likely that we will continue to experience weakness in product revenues. For the year net revenues decreased by 31% from 27.8 million to 19.2 million. The decrease was attributable primarily to a decline in royalty revenues, which fell from 14.3 million to 6.9 million. The decrease in royalties is primarily due to the reduction in royalties received from GAMECUBE unit sales. The gross margin percentage in the fourth quarter was 80 percent of net revenues compared to 87% of the fourth quarter of 2002. This decrease was mainly attributable to lower revenues from royalties compared to the fourth quarter of last year.

Operating expenses including research and development were 3.6 million. Our fourth quarter operating expenses also included stock-based compensation charges of $38,000. Other income for the quarter was 461,000 with approximately 335,000 from interest income and 126,000 of other income, substantially all from Canadian research and development incentive tax credits. The net loss for the quarter was 415,000 for a loss of one cent diluted earnings per share compared to net income of 4.1 million or 13 cents fully diluted earnings per share in the same period last year. Fourth quarter diluted earnings per share were computed using 30,704,000 shares. Net income for the year was 2.5 million compared to 12.4 million recorded last year. Fully diluted earnings per share for the year were eight cents compared to the 40 cents earned in 2002. We used 30,998,000 shares to compute fully diluted earnings per share for 2003. Net income was calculated using an effective tax rate of 10 percent for both years.

Significant revenues from licensees included TSMC and Sony, each representing 17% of total revenues for the quarter. For the year significant revenues from licensees included Sony, NEC, and UMC representing 17%, 14%, and 11% respectively of total revenues for the year.

Prior to releasing our fourth quarter results we revised our revenue guidance. In both the third and fourth quarters we gave guidance based upon booked orders and forecasted new orders for which we had received strong indication from our customers concerning their intentions and commencement dates for new design projects. In the fourth quarter we, in fact, received some of these orders, however, too late to recognize significant revenues. We are limiting our revenue guidance in the first quarter of 2004 to business we have currently booked as of yearend. As a result we would anticipate total revenues will range from 4.5 million to 5.5 million. We would anticipate that the majority of the increase in the first quarter revenues would be realized from licensing activities. We expect operating expenses will range between 3.6 million and 3.8 million in the first quarter. Our tax rate for 2004 is expected to be 20%. Despite the net loss, we were still able to generate approximately $1.5m of net cash flow from operations in the fourth quarter. Cash, equivalents, both long- and short-term investments totaled 85.8 million at the end of the quarter. On December 31, 2003 accounts receivable totaled $1m of which 151,000 were attributable to product sales. Inventory at the end of the fourth quarter declined to 474,000 from 592,000 at the end of the third quarter. At yearend we had 92 employees of which 67 are engineers.

That concludes my prepared remarks about the financial results. Fu-Chieh will now discuss the business highlights of the fourth quarter.

This past quarter proved to be another challenging one for Mo-Sys. That is reflected in the quarter's financial results. However, we ended the year strong with new orders receiving $5m in licensing fees. We will recognize revenue from these new orders throughout 2004. We currently secure the majority of our business with our customers on a project-to-project basis. During the past year it has been difficult projecting exactly when we will be receiving new design projects from our customers. We still experience instances where the customer has indicated to us an approximate start date for the project. However some of those projects have been and continue to be delayed by our customers. We believe that many of the delays are caused by the high cost associated with designs in 0.13-micron and 90-nanometer designs causing many of our customers to be very cautious in initiating any new design projects.

During the fourth quarter we signed a technology license with Fujitsu in which we have licensed our 1T-SRAM-Q technology. Under the agreement Fujitsu will have the ability to create their own 1T-SRAM design for their customers. In return we will receive royalties from their sales of chips containing our technology. We will recognize revenue from this license agreement as we import our technology to their process, first the 0.13-micron, and train their engineers to design 1T-SRAM macro. We also announced that Sanyo Electric and Agilent Technology were signed as licensees during the quarter. At yearend we had 41 signed licensees, 10 of which were added during 2003. After the close of the quarter we announced the launch of our new single-port and dual-port 6T-SRAM-R compilers. Customers are very pleased with the quality and low-power advantages offered by our 1T-SRAM technologies for large memories and have asked Mo-Sys to help them extend these benefits to the small memories in their design. In response we developed our 6T-SRAM-R compilers, which deliver a new standard of quality combined with much lower leakage than other six-transistor embedded memories. In generic 0.13-micron logic processor replacing large six-transistor memories with 1T-SRAM-M technology can reduce leakage by more than a factor of four. In addition to the considerable die size reduction, using Mo-Sys' new 6T-SRAM-R technology compilers designers can also reduce the leakage of remaining six-transistor memories by more than a factor of two further optimizing the total SoC leakage without compromising die size or requiring special processors.

During the quarter we received a report that AEC had resumed shipment of chips for the GAMECUBE game console. Nintendo has announced that they can still achieve the six million GAMECUBE unit sales they had projected at the beginning of their fiscal year 2003. Although we do not know Nintendo's projected GAMECUBE sales for their 2004 fiscal year ending March 31, 2005, we do anticipate recognizing royalties from GAMECUBE unit sales in 2004.

Now I would like to have Mark-Eric Jones give some more detailed insight into our markets and customers. Mark-Eric.

Thank you Fu-Chieh. We have to acknowledge that 2003 was a very tough year for Mo-Sys. However, we are now seeing signs of recovery a little later than some commodity IT suppliers. I am encouraged by our fourth quarter bookings both in magnitude and also in quality of customers.

Our technology license with Fujitsu marks a very important milestone achievement for Mo-Sys. It marks the first integrated device manufacturer to take a full technology license from Mo-Sys. Fujitsu, the largest ASIC supplier in Japan, is a company with extensive memory experience. Despite having tremendous D-RAM expertise, Fujitsu nonetheless saw the distinct advantages that our 1T-SRAM-Q offers over embedded D-RAM. In its selection of 1T-Q, Fujitsu cited the advantages of our technology in manufacturability, performance, and cost over embedded D-RAM. Fujitsu has said it will use Mo-Sys' 1T-SRAM-Q to provide high-density embedded memory solutions to their SoC design customers in consumer applications including digital cameras and video camcorders.

During the quarter we also signed both Sanyo and Agilent as new licensees. The semiconductor group within Sanyo has revenues of $2.5b annually providing Sanyo with leading edge semiconductor solutions for [Mohawk] communications equipment, PC, and PC-related products, analog and digital audio and video equipment. The initial development project is an application for the consumer market.

During the quarter we also signed Agilent Technologies as a new licensee. The ASIC group within Agilent Technologies has annual revenues approaching $2b providing advanced semiconductor solutions to their customers including their largest customer Hewlett-Packard for which they provide ICs for HP's printer and scanner products.

We continue to see strong interest in our 1T-SRAM-Q technology as customer projects specify increasing amounts of embedded memories. In 2003 we ported our 1T-SRAM-Q technology to several foundry partners including TSMC, UMC, and Chartered Semiconductor. Following this excellent support of the major foundries, we look to 2004 for proliferating the licensing of our 1T-SRAM-Q to end customers requiring a cost-effective solution to increasing amounts of embedded memories.

In addition to the continuing trend of increased embedded memories we also are seeing customers requirements being driven by the growing need for low-power solutions. The release of our 6T compiler was made in conjunction with a joint marketing agreement with Virtual Silicon Technology. By jointly calling on customers we can offer the customer a set of low-power solutions designed to meet the customer's ever-increasing power requirements.

I am often asked by investors exactly how we identify 1T-SRAM opportunities with customers or within various markets. We typically work with customers where we can identify a design project that has one or more large memory blocks, which represent more than 20% of the total die area. It is within this range that we have the opportunity to create significant cost savings for the customer and secure a design win. However, with our new low-power 6-T compilers, Mo-Sys can now provide the complete set of memory blocks for our customers' designs. We now have the ability to offer our customers a range of embedded memory solutions that offer density, performance, and low power unequalled by competing product offerings.

Now I'll turn the call back to Fu-Chieh.

2003 was indeed a difficult year for our company. We saw delays in design starts by many of our customers. We also saw delays or cancellation of projects in which we had completed our design and had anticipated that those projects would begin production in 2003. However, we ended the year with a strong booking of new orders in the fourth quarter. We look forward to continuing improvement for our business in 2004. In our prepared comments we have outlined our business results and guidance for the first quarter of 2004. I would like to reiterate that, due to the confidential nature of our licensing agreement, we usually cannot provide detailed information on the specific projects or licensee. This is a contractual obligation that we must uphold in order to preserve our relationship with our customers. Aside from the limits of this nature, we intend to disclose relevant material information concerning our business and operating results as we are able to do so.

That concludes our prepared comments. We are now ready to take questions. Please introduce yourself including your firm's name when you ask your question.

Thank you sir. If you have a question or a comment, key star then one on your tone-dial phone. If you want to withdraw your question, key star two. Once again, key star one for questions. We'll pause for just a moment while your questions are collected. That question comes from the line of Dan Scovel of Needham & Company. Please proceed.

A couple of questions for you. On the GAMECUBE project, Mark Voll you said in the guidance for Q1 that most of the increase was coming from license revenue. Within your guidance are you assuming that there is some increase at least in royalties and that GAMECUBE will be part of that?

Some.

So, yes, there is some GAMECUBE in Q1?

Right. A modest amount as they begin production again.

The Fujitsu deal with regards to ASICs for their customers, you support other ASIC providers? Or is this your first ASIC deal?

We have supported other ASIC providers. This is our first deal of a complete technology license whereby their designers can create their own 1T-SRAM memory blocks for their customers.

Does that forgo you from potential licensing opportunities for other designs? What is the trade-off?

It is part of the proliferation of 1T-SRAM getting more engineers working on 1T-SRAM designs beyond what we can achieve with the engineering within Mo-Sys.

OK. Can we assume that, because of the technology nature of it that it is probably worth more to you than a standard licensing agreement?

Yes, that is a correct assumption.

The other issue - you said you booked about $5m in license revenue last quarter. Fu-Chieh, you seemed to indicate that that would take all year this year to recognize that. How long would you expect to take to recognize those particular bookings?

I think I mentioned the total amount covered many individual projects. Each project has a different duration. I think the rule-of-thumb is that it would probably cover the bulk of the year. It may or may not be [unclear] depending on the digital product mix and what kind of milestones and deliverables we are required to deliver.

What is typical? Is it typically two to three quarters? Or more or less?

I think typically-correct if I am wrong Mark - it is spread over a year.

Correct. We would assume that typically 50% to 70% would be within a couple of quarters.

OK. The pre-announcement of weakness is a replay from the third quarter. The $5m in bookings you received in the fourth quarter, how does that compare with bookings you received in the third and second quarters - to put some perspective on that?

I think that you point out that we were trying not to repeat the previous quarter's guidance. If you notice Mark comments for the forecast of the first quarter of 2004, we tried to be a little more conservative counting only the business that we have already signed. Rather than in the past we also included a normal projection of [turns] in the quarter.

What is a normal amount of turns during a quarter?

I think that in the last two quarters, it would be abnormally slow.

What I am trying to get at is how-give us some-I am asking for a little more granularity to give us a little more confidence in the guidance going into this quarter. If you can scale it with bookings-you told us what the bookings number was last quarter. Can you give us some perspective around that? Or, maybe it's total backlog. How much backlog did you have going into the second or third quarter? How did that help?

I think that in the previous quarter's conference call-if I remember we had similar questions. Our guidance at that time-the forecast for the revenue for the fourth quarter - at that time there was a significant portion of the business already booked. Although a significant fraction of it was expected to be booked in the quarter. Now at this conference, I think our current guidance-we have changed the method. We are only forecasting-based on guidance-based on business that is already booked. So hopefully----

So is that to say that your current forecast assumes zero turns and your current backlog holds for the aging on the quarter?

That is correct.

OK. The next prime question would be - you cited some push-outs and cancellations of other projects. What sort of a normal push-out and cancellation rate? Would that represent risk to your current forecast?

Currently we do not forecast beyond the current quarter. We stated earlier that the current quarter is all based on established contracts. We see that relatively low risk to the guidance, although that could have an effect on the following quarters when some of the deliverable activity has not yet been authorized.

Also, longer-term how should we be thinking about this year? Would you expect-what is the chance that revenues fall below the first quarter level? Do you consider that unlikely in that you're continuing to see growth this year?

I think first of all, we have no control in our licensee and customers' business. Judging from the general up-tick in the IT industry as a whole, we see that-we believe that our customers' business will be significantly more robust than in the past. Therefore, I think that theoretically such a possibility always exists. We think that it is progressively less likely.

Thank you. Good luck.

Again, ladies and gentlemen if you would like to ask a question, please key star one on your touch-tone phone. We have a question from Todd Allen [ph] of Kenny Securities.

Good afternoon, gentlemen. This is Todd Allen. I am calling in for Matthew Curtis. I have couple of quick questions on product timing. The Motorola product cell phone application, has that begun production? Do you anticipate it beginning in the near future? Similarly, the LSI logic project - has that already begun production?

We can't comment about specific licensees' products except where we've had permission to make public announcements.

OK. You can't comment even to-as to whether or not it has begun?

Right.

OK. Also, could you provide some estimate as to the size of potential royalties from GAMECUBE for 2004? I know that you said they might be somewhat slight in the first quarter.

I think that we were pleased that the inventory in the channel has been depleted and that Nintendo has resumed manufacturing. As we reported earlier, the resumption has started at a relatively modest rate. How much is the quantity? I think this is up to Nintendo in whatever the public announcements that may come in the months ahead. We are in no position to forecast or predict what the manufacturing or production quantity might be.

OK. That is what I had. Thank you.

You have a question from Gavin Duffy of A.G. Edwards. Please proceed.

I have a couple quick questions. One, with the deals with Agilent and Fujitsu you have been talking about both these customers for quite awhile before they actually signed on-officially signed licenses. Would it be fair to say that probably of these customers have got end projects in mind-probably already targeted or even probably themselves have customers in hand when they finally signed the license agreement?

Yes. It is highly unusual for a customer to sign a license agreement without a specific target project in mind. If you look at the magnitude of what both of those customers you mentioned business is, I would say there is scope for many projects.

OK. When you expect 90-nanometer at Fujitsu?

I think we just started the program as we reported at the 0.13-micron generation. There has been a lot of discussion. We would normally expect that the progression is that once we finish the first project then we would move onto the second project.

OK. Last question is that I know that before you talked about disk-drive controllers as a great opportunity. Can you give us any update on work there?

Yes. We believe it's a good opportunity.

Alright. Thanks.

Again, ladies and gentlemen that is star one for questions. You have a question from Chris Chaney of A.G. Edwards & Sons. Please proceed.

Hello gentlemen. Good to hear from you again. Mark, this is a question for you. I want to ask you, the licensing sign-ups that happened in the fourth quarter - $5m - can you talk to us or let us know a little bit about the mechanics of how that is going to impact cash in the next couple of quarters? As I understood that, cash would be received in chunks and increments. I am trying to figure out how that would impact the balance sheet.

Correct. First of all to speak on the timing of the license agreements, themselves. For the most part, we received many of the agreements towards the end of the quarter. We had stated there wasn't an opportunity to see significant revenues from those. We would anticipate that we would be achieving the milestone and both invoice and payment cycle as we go into 2004. In particular we had stated normally 50% to 70% of the revenue that we would recognize, somewhat related to the milestones, would be within the first couple quarters.

OK. The Fujitsu contract, since it is a technology license I am assuming it is relatively large in size on a revenue basis. Would it be fair to assume that a good chunk of cash could be received-should be showing up on the first quarter's balance sheet from that?

I would say in the first couple quarters.

OK, spread out over a couple quarters. Alright. The second thing is, I am trying to get an idea of the mix of your end customer base right now. Of the 21 projects in the license revenue stream, without quantitatively saying, maybe you could tell me on average how many of those are in the consumer segment versus the communications segment, which is where they used to be. Also, the same question for eight licensees in the royalty stream as well.

I think throughout I can answer all three questions with the same answer. The majority is now very definitely in the consumer space rather than the communications space.

OK - not surprising, I guess. A question on the licensees. Can you talk about-I know you have mentioned a number of projects you have and the number of licenses you have right now in revenue. Can you give us an idea of your pipeline? How many are you working on? I know you can't quantify the amount of each. Can you give us an idea of how many projects you are currently working on that aren't in the licensing streams to give us some sort of confidence as to the licensing build-out maybe through 2004?

We are seeing strong pipeline of sales opportunities - to answer your question, as distinct from projects that we are actually working on, better than we have seen in the past. That is why we made our positive comments about the fact that we're seeing an upturn in the semiconductor activity as well.

Sure. So by quality you are referring also to the size per project, right?

Yes [inaudible].

Or the revenue per----

Yes the projects and also the quality in terms of the revenue potential.

OK, great. With the revenue guidance for the first quarter, it sounds to me like you would also be assuming that gross margin could be back into the mid 80s. Is that a fair assumption?

That is correct, yes.

OK. One last question - this is a little arbitrary. If you were using the old method that you used to use when you would forecast revenue guidance for Q1 compared to your guidance now or the method that you are giving now, could you estimate what the revenue would have been? In other words, would you be estimating that revenue guidance could be six million to seven million if you were using the old methodology versus the 5.5 million or so using the new methodology?

We prefer not to be burned again.

I understand. I am trying to get an idea of how to compare it with the past few quarters and how the trend is going in terms of what you expect versus what you have in hand.

Typically we would range of having anywhere from 50% to 70% of the orders booked going into a quarter. It was much higher in 2002. 2003 our assumptions were that we would get these orders. In fact, we got the orders in Q4 that we had anticipated, however, too late in the quarter to recognize any significant revenues. Had those occurred earlier in the quarter, our results would have been much improved.

OK, great. That answered my question. Thank you.

You have a question from Dan Scovel of Needham & Company. Please proceed.

The GAMECUBE has a well-publicized price cut to spur additional sales. Because you collect revenue as a percent of the chip shipments into Nintendo, for modeling purposes is it not unfair to assume that their price cut is probably related to bill of materials as well and will probably trickle down in some way into your coffers?

Yes. In normal business arrangements, when there is a [pain] all the partners have to share it.

On the 6T-SRAM, that seems like a significant change, or a significant new product. Is that to say you've-how did you come about that? How are you able to offer that?

When you look at the work we've done in the past few years on low-power designs using 1T-SRAM, many of those techniques and technologies that Mo-Sys developed can also be applied to the smaller memories. We saw the demand from the customer. They could save a lot of power on the big memory in their design with 1T-SRAM, but didn't have a solution for the smaller memories. Hence our offering the 6T-SRAM as a complement-as a companion to the 1T-SRAM to help cover the smaller memories in SoC designs as well and stop the leakage there.

In terms of the deliverable-the delivery vehicle is primarily through a compiler so it is the automated layout [arena]. The customer has the flexibility to place a large number of distinct different size memories.

You are much closer to technology. Your whole story to date has been a 1T-SRAM story. Clearly you have a lot of technology surrounding the cell to do that. This is highlighting the point that you are not just a 1T house. You are the technology around the memory cell. Is that how we should look at this?

Exactly. Most of the technology was never tied to any particular [unintelligible]. In the past when we started doing the IT business, we had to be very focused to show the maximum benefit of our technology in all the various aspects. As Mark-Eric mentioned that today's SoCs are fairly complex. People need memories of all sizes and densities. Even though Mo-Sys provides a very compelling solution for some of the larger blocks, that does solve the customers' total issues. That is why it is very, very important for us, particularly for customers who can take advantage of our low-power capability. They do need to have a full suite of a memory solution that covers all their needs. That is what is driving our product offering, not just for the single-port, for the dual-port as well and with all the significance. We feel that our 6T-SRAM compiler product offering is a very, very compelling one for the customer.

Great. Thank you.

You have a question from Colin Williams [ph] of Williams Blair. [ph]

The Sony revenue at 17% in the quarter - can you comment on how much of that is royalties versus license revenue from Sony?

It's a mix of both.

It's a mix of both. Is the majority of that royalty?

Sorry. We can't answer that question. It is going to be a mix of both.

OK. Again on the royalties. As you look into the first quarter, backing out the GAMECUBE contribution would you expect a sequential up-tick in royalty revenue from Q4? That is absent the GAMECUBE contribution.

I would expect a modest increase in Q1.

OK. You provided new orders for licenses in Q4 of five million. Could you comment on new orders for license revenue in Q1 so far? Perhaps you could quantify the quarter-to-date data. I am trying to figure out if the December order momentum is continuing into January.

The prospects are looking good. We can't comment on actual booking at this point.

OK. The last question I have is, how do you track the progress of your relationship with Sony since they have a universal license and it's not necessarily easy to track project-by-project what they are doing.

Are you referring to the Fujitsu contract?

The Sony Kyushu [ph] relationship.

Currently I think with Sony we have a multiple projects with them. That is not a technology license arrangement. We do have visibility on every single project.

OK. How is that progressing? Are the number of projects increasing?

Yes. They continue to use 1T-SRAM in all new projects [inaudible].

OK. Thank you.

You have a question from Steve Velesque [ph] of Semi Equity Partners. Please proceed.

Yes. I would like if you could answer regarding licensing what the breakdown is between your standard and crafted macros and compiler generated in terms of the breakout. Is it 90/10 or 80/20, etc.?

I would probably say it is closer to 90/10. I think that this question is asked quite often. We offer many different methods of using our memory including full custom and the standard, no-standard, the macros as well as the compiler-generated macros. Our experience has always been that the standard macro and the compiler-generated macro has been [as a marketing tool has allowed the customer to do floor planning, SoC planning.] In general once they go to deployment phase because one of the main objectives is cost-savings for high-volume applications, we do see that the majority of customers typically ask us to do some customization.

With your new products and targeting smaller blocks, do you see that percentage changing over time? Do you see it the same because of what you just said?

In terms of new products, the 6T-SRAM-R compilers, I think that memory is generally quite small, optimizing for density doesn't have much significance for the customer. The key is that a fully generated compiler generated [is the] maximum flexibility convenience for the customer to use. The delivery technology, the distribution technology - the methods are slightly different to maximize the customers' benefit.

OK, but do you see the-I understand the benefits are different. Do you see the percentage using the compilers for these new products changing the mix from licensing?

Yes, 6T, I expect the vast majority will be compiled rather than custom [inaudible].

So if I were to project total mix between compiler generated and hand-crafted a year from now, what would be the overall mix?

We are talking the new product offering now. We just announced our 6T-SRAM-R and the single-port/dual-port family a few days ago. We would love to see the business take off instantaneously. Typically in the IT business, it will probably take some time for the business to develop.

Alright. Thank you.

You have a question from Brian Grinahan [ph] of Grinahan Investment. Please proceed.

Can you go over for me again the payment-the cash versus the revenue recognition. I am trying to reconcile the five million in bookings you got in Q4. Do you expect-I thought it was 50% upfront and then further on milestones. Could you go over that for me again.

Correct. We would bill according to the milestones that are in the agreement. Then we would recognize revenue based on our estimated work to complete that project as we complete that work.

Isn't there an upfront?

I think that one thing, yes. As Mark stated, we may bill the customer for upfront payment. The customer usually takes some-whatever 15 days, 30 days, 45 days cycle - depending on the size of the company to make that payment.

What is the percentage upfront? Did it used to be 50%?

50% mostly in the macro agreements. Some of the other agreements, such as technology, it would be somewhat different.

What are they?

Usually the larger agreement has multiple pieces of it. The digital pieces, the convention is that a portion needs to be upfront.

I am looking for some gauging of deferred revenue in the first quarter. I am trying to get a sense of how much of the five million you will see.

In deferred revenue?

Yes.

Deferred revenue under our business model is difficult to forecast. We record differed revenue as we receive cash. You have the aspect first of achieving the milestone, then (2) of payment on that milestone.

Are you referring to the fourth quarter deferred revenue or the first quarter deferred revenue?

In the quarter that we're in - the first quarter. I would think that you would get a deposit before the milestones hit. Before you recognize revenue you would have money-you would have cash - some percentage of the five million you booked in the fourth quarter. That would hit the balance sheet regardless of how much of the revenue you booked. I am trying to get a sense----

That would be when we collect cash payment.

Alright. The royalty decline - in the past three quarters we haven't had any Nintendo royalties, yet we've seen a decline. Can you break it out in terms of the unit? Is that 1-8 to 1-2 to 1-1, does that track the units? Does it reflect the unit production? Does it reflect some pricing or some dropping out of products out of production? To follow on, I think we're seeing strength in the semis now. Will we see some catch up or a lag effect? Would we expect to see that trend up reflecting the current situation?

You will see a lag effect because our royalties always come a quarter after the shipment to our customers. You will definitely see at least that three-month lag effect.

OK. Will the products that are in production that are generating the royalty - do you see them being benefited in the current environment? That is what I am looking for. Or are they not being benefited right now. There is a trend set-a downward trend that will continue?

We hope that our customers-that businesses will recover and become booming. I think during the-when business slows down in general not only the unit volume may be under pressure. I think our customers pricing or ASP may be under pressure. That is not something that we have any control over or can forecast. We merely get the report from the customer.

OK. To follow-up on that. The number of products that are generating royalty revenue, I believe has stayed fairly constant through the year. Does that reflect a turnover? Are those the same products or the same licensees that are generating that royalty?

Over time there is a transition.

So the number that have generated royalty in the year 2003 is some number greater than eight?

We have eight licensees.

Generating royalty.

OK, so the eight licensees have not changed?

There has been some fluctuation - new projects coming on and older projects discontinuing in [unclear] production.

OK. That is what I am getting at. OK, alright. Thanks.

Sir, we have no questions at this time. I apologize. We have a follow-up question from Gavin Duffy of A.G. Edwards.

I want to follow on what Brian was getting at. How many projects are generating royalty since you can have more than one project at a customer. You have eight customers generating royalties.

Most of those have multiple projects in production.

So what kind of ballpark are we looking at in terms of how many things are generating royalties? Is this in the 20s - 20 different products or 30?

I would say approximately 20.

About 20, alright. Thank you.

I have no questions at this time. I'd like to turn it over to you for closing remarks.

Thank you again for participating in our call. Before we close I would like to reiterate the key highlights and achievements of the past quarter. We ended the first quarter with new orders exceeding five million in licensing fees. We signed a technology licensing agreement with Fujitsu, the first IDM that is licensed to create 1T-SRAM designs for the customers. We signed licensing agreements with Sanyo and Agilent. We launched our first low-power 6T memory compilers available for the leading foundries. Thank you again for your participation in this call.

Ladies and gentlemen thank you for joining us in today's conference. You may now disconnect.