Quicklogic Corp (QUIK) 2003 Q3 法說會逐字稿

Good day, ladies and gentlemen. Welcome to the Q3, 2003 QuickLogic earnings conference call. My name is Carlo and I will be your conference coordinator today. At this time, all participants are in a listen-only mode and we will be facilitating a question and answer session towards the end of this conference. If at any time during this call you require assistance, please press *0, and a conference coordinator will be happy to assist you. A reminder this conference is being recorded today for replay purposes. I would like to turn this presentation over to your host for today's call, Mr. Tom Hart, Chairman, President and Chief Executive Officer. Please proceed, sir.

Thank you, Carlo. Good afternoon ladies and gentlemen and welcome to our Third Quarter 2003 Earning Conference Call. Thanks for taking time to join us today and hear about QuickLogic and our leadership of this new category of semiconductor devices we call embedded standard products. Carl Mills our CFO will take you through the Q3 financials, and then I will share my perspective on our business. Finally, Carl will detail our guidance for Q4, and for the year and then we'll take your questions. Carl?

Thank you, Tom. Before I get started, I'd like to read a short Safe Harbor statement. During this call we will make statements that are forward-looking. These include statements concerning future results, financial metrics, visibility, the competitive environment, acceptance our products going forward, economic conditions, expected actions by QuickLogic. The expected timing of our returning to profitability and market opportunities generally.

These forward-looking statements involve risks and uncertainties, and QuickLogic's future results could differ materially from such forward-looking statements. We refer you to the risk factors listed in our annual report on form 10-K, quarterly reports on 10-Q and prior press releases for description of these and other risks that could cause actual results to differ materially from our forward-looking statements. QuickLogic assumes no obligation to update any such forward-looking statements. For your information, this conference call is open to all, and is being Webcast live. It can be accessed from the investor relations area of the QuickLogic Web site located at www.quicklogic.com.

We had another fine quarter and our Q3 profit and loss results were stronger than the guidance we provided. Some of the highlights of our results are our Q3 revenue was $11.2 million. Revenue increased 34% compared to Q3 one year ago. Our sequential revenue increased 5.4%, and was at the high end of our guidance of a 1 to 5% increase. Our ESP products contributed 42% of revenues. Our largest customer which purchased ESP products contributed 11% of revenue in Q3 compared to 27% last quarter. We'll provide more insight on demand from this customer during the guidance portion of our call. Two other customers contributed 6% and 5% of revenue respectively in Q3. Operating expenses were at the low end of our guidance for the quarter. Our cash flow is positive for the third consecutive quarter.

Although we had an expected sequential decline in revenue associated with our largest customer, we offset this decline with strong sequential growth in our North American and European revenue. In previous quarters, Tom has spoken with you about the issues that can arise from trying to read too much into sequential comparisons. Suffice it to say that our year to date results are more indicative of our regional mix. Revenues from Asia-Pacific, primarily Japan and China has more than doubled in 2003 compared to 2002 and accounts for 36% of total revenue. Revenue from North America and Europe has grown 7% on a year the date basis and represents 64% of revenue.

We sell our products to distributors and directly to OEMs. Distributors have accounted for 73% of the revenue so far in 2003, compared with 70% in fiscal 2002. Our gross margin was 51.5% of revenue in the third quarter. Gross margin was at the high end of our guidance of 47 to 52%. The sale of previously reserved inventory reduced our cost of revenue by 320 basis points which was twice as beneficial as we had expected. As a company, we're focused on controlling our expenses. In general, if an incremental expense is not necessary for new product development, for generating revenue or required for compliance programs, we have not increased our spending.

Our Q3 research and development expenses increased by $300,000 to $2.7 million from $2.4 million in Q2. We are bringing our QuickBits 2 (ph) and Eclipse 2 products to market. We believe that features and the price points of these products will expand our share of the market, allow us to pursue higher volume applications and drive revenue growth for the company. These products are being built at Tower Semi-conductor, using advanced manufacturing technology. As expected we had significant preproduction research and development expenses in Q3 associated with these devices. Our SG&A expenses declined by approximately $71,000 dollars sequentially as a result of lower bad debt reserve expense.

To recap our sequential results, our revenue increased by 600,000 dollars, our gross profit improved by 200,000 dollars and our combined R&D and SG&A expenses decreased by 200,000 dollars. These combined factors produce a loss from operations of $700,000 in Q3. In total, our net loss in Q3 was $709,000, compared to a net loss of $24,000 in the second quarter. As you may recall, we had a $700,000 gain on the sale of Tower securities in Q2 and this gain was the biggest difference between our Q3 net income compared to Q2. Our net loss of $709,000 in Q3 of 2003 compares to a net loss of $6 million in the third quarter of 2002. Several factors contributed to this $5.3 million improvement. One, higher revenues of $2.9 million. Two, gross margin improved to 51.5% of sales from 21.5% of sales one year ago. This improvement is a result of lower subcontract loss, better overhead absorption and lower costs associated with inventory reserves.

Finally, operating expenses were $1.5 million lower than one year ago, including long-lived asset write-offs which were lower by $1 million. We had positive cash flow of $1.3 million this quarter. This is our third consecutive quarter of positive cash flow. Cash flow from operations provided us with $1.9 million of cash in Q3. And we have generated nearly $4.4 million dollars of cash flow from operations so far this year. This is very positive news for our company, and compares to an approximately $5.7 million dollar use of cash from operations during the first three quarters of 2002.

Let me provide you with some of the details of our Q3 cash flow from operations. Our net loss, adjusted through the effect of non cash charges such as depreciation and asset write-offs, was a positive $800,000. We reduced accounts receivable by $600,000 during the quarter even as our revenues increased by $600,000 sequentially. The reduction of inventory contributed $500,000 in cash in Q3. Essentially during Q3, we consumed inventory we built during the second quarter. Other balance sheet items including our reduction in accounts payable and our reduction in prepaid and other assets had no net effect on operating cash flow in Q3 as they offset each other in total. Our investing actives used $700,000 of cash in Q3. We purchased nearly $800,000 of capital equipment and sold the 1. – or 13,000 Tower shares which generated $68,000 of cash.

Our financing activity did not have a net impact on our Q3 cash flow. We did repay $1.2 million of debt during the quarter and we also received $1.2 million from the exercise of stock options by our employees. As a result of these factors our Q3 ending cash balance increased to $25.3 million from $24.1 million at the end of Q2. Our cash position less our interest-bearing debt was $18.8 million at September 30th. This is a $2.4 million improvement from June 2003. We had 161 employees at the end of the third quarter compared with 160 at the end of Q2. Now let me turn the call over to Tom.

Thank you, Carl. Well, you have already heard from Carl or seen on our Q3 earnings press release that we had another great quarter by virtually all measures. But we still are not where we want to be. We're getting a whole lot closer. Clearly, our trends are in the right direction, especially when you compare our progress against what is normally viewed as our competition. In comparison to what has classically been our direct competition, namely Altera (ph) -- Xylinks(ph) with Actel is a distant third, we grew over four times as fast sequentially in Q3 and over twice as fast compared to year ago as all of them combined.

In Q3, our four quarter rolling average revenue growth rate --this is my personal favorite when it comes to meaningful comparisons, shows that we have outgrown the total of our competition for eight out of the last eight quarters. Our strong growth rate is a direct result of our embedded standard products driving our growth picture. In Q3, ESP has accounted for 42% of total revenue and grew 59% from the same quarter a year ago. From a bookings perspective this Q3 was 11% ahead of Q3 last year. However, we were behind as expected from the booking rate of Q2, primarily owing to a sequential decline in Asia-Pacific. The bright spot for Q3 sequential new orders growth was North America and in Europe, which together grew 19%.

New orders were led by strength from OEMs in all three markets and distributor orders from Europe. As we expected, new orders lagged in Asia-Pacific, owning to the significant new order rates we experienced in Q1 and Q2. Turns in the quarter were 51% of total revenue. Down significantly from 68 in Q2 and clearly less than our typical 75 to 80%. Of course, this is largely a direct result of having large orders on the books from our largest customer in Asia-Pac. Owing to our fundamental business model of high velocity and high turns, we fully expect to return to our typical turns rates.

One clear effect of our model in contrast to our competition's model is that we finished last quarter with less than four days of inventory in the channel. Yes, you heard me correctly, I did say less than four days of total inventory in the channel. This has been fairly typical for the last several quarters as we basically build to order out of a die bank and continued to delight our customers with three-week lead times which we have done now by the way for well over three years. This three-week lead time includes our programming devices for customers with their custom code.

So if you think about this for a minute, to our customers, we look like an ASIC supplier with several positive exceptions. Namely, there's no large non-recurring engineering expenses, there's no mass costs, and there's virtually no inventory liability. Perhaps the best of all aspects of this is that our customers can easily make changes to their designs, and quickly see those changes reflected in the product we ship to them. By the way, if the customer's willing to give us their run rate visibility, we will shorten their lead times to one week. Just for the record, and to put this in perspective for you, we now program over 80% of the units that we ship.

Of course, the results of our offering short lead times to our customers is that we don't get much visibility on coming quarters shipments and subsequent revenue. And reduced revenue visibility makes you folks who are concerned with sequential revenue numbers uncomfortable or at least nervous. We can't change your concerns, only suggest you look at how conservatively we run this business, look at the wealth of new products we have invested in and observe the reality of embedded standard products revenue growth.

As we have said in the past using the sports analogy, we view this business as a marathon not a sprint. The time constants are long, the customer base diverse and we're driving a new paradigm. And our results show that our vision of embedded standard products is turning into reality as have we moved from ESP's being 12% of total revenue in 2000 to 42% of this year’s revenue to date. A big part of the good news here is that all of the revenue growth we have enjoyed to date still does not include any meaningful revenue from our newest products. The Eclipse II and QuickMips II (ph) families, both of which began sampling in Q3. The Eclipse II family is our very low-power, high performance FPGA family. In fact, Eclipse II offers the very lowest power of any comparable density FPGA in the marketplace. Eclipse II is the lowest in all three dimensions of power. First, the power required when the device is powered up. Sometimes also called the start-up or end rush power. Second, the static power when the device has voltage is applied but its clock is not running. And third, the dynamic power consumed when the device is fully functioning at design-intended speed. We fully expect that Eclipse II will successfully challenge the larger density, low-power CPLDs currently on the market. Eclipse II family is fully pin (ph) compatible with the older Eclipse and Eclipse-plus family of devices, and therefore, we have customers in the queue ready for product as soon as they finish their qualification process.

We expect to complete our internal qual (sic) process and therefore ship Eclipse II product for revenue in Q4. Our QuickMIPS II family of devices continues the customer design wins and some truly exciting applications. I would like to share two of those with you today. First as everyone is well aware, the world is going digital. What that really means is the world is headed towards packet-based communication systems. Voice, video and data converted to packets, and communicated over the same network. Of course, the challenge here is to make these networks able to provide different levels of priority for the different packet types. This management of quality of service, QOS as it is called has been one of the major challenges of this desire to merge data, voice and video. Quick MIPS II is ideal for this kind of application in that the onboard MIPS microprocessor along with its other tasks manages the quality of service or priority -- resource priority allocation, and the programmable fabric enables the packet inspection. You see microprocessors to do a really good job of managing and manipulating bites. While, packets are more bit oriented. Bit oriented operations are ideal to be done on the programmable fabric. Hence, a very cost-effective solution is being designed to interface legacy systems including voice, video and data to the newer packet-based networks. This particular Tier one customer has also integrated compression and encryption of the packets using the programmable fab for an even higher value-added solution.

The second application I wanted to share with you today is primarily a high-speed video challenge. Gigabits per second of video coming from multiple sources being compressed an impressive 100 times and then transmitted over category five wiring or wireless at Ethernet rates. For 3-D graphics and test, the compression rate is a phenomenal 1500. To put this in perspective for you, Mpeg-2 which is the industry standard for full motion video compression offers a compression of about 20 times. So our customer has used virtually all the resources on the Quick MIPS II device to turn the compression world on its ear. For example, they're using 99.7% of all the logic cells in the Quick MIPS II device. That's 2010 out of the total of 2016 cells available.

Be careful though, don't try this level of logic utilization from your SRAM-based devices. They don't have the massive on-Chip interconnect required to utilize this much logic and still route reliably. Our on-Chip interconnect is enabled by Via Link, a truly remarkable invention and the innovation that makes our devices the lowest power and the highest security devices on the market.

And speaking of high security, we performed some tests during this quarter for a Tier one customer that knows firsthand the heart break and the negative P&L impact of having your intellectual property purloined and then used against you. Very ugly indeed. We have talked before how it is virtually impossible to discover what has been programmed into our Via Link-based programmable devices.

This major use of FPGAa asked us to demonstrate this -- not in a marketing sense, but in the lab. To do this, we went to a local lab that has a focused ion beam milling machine. This is a machine that costs well over $1 million and uses an ion beam to dig a trench in a silicone, so that with a scanning electron microscope, you can try to find which switches have been programmed to enable the logic desired. Our folks knew exactly which switches had been programmed, and it still took them over two hours of milling to find one physical spot where a connection had been made. Now, the thing you need to know is that on the Quick MIPS II device we just talked about, there are over 4.3 million potential switches. Only 59,000 of which were programmed to use the 99.7% of the logic cells. At less than 2% of the switches programmed, I think this qualifies as a search for a needle in a hay stack.

Needless to say, the customer is now a believer that it is virtually impossible to reverse engineer a ViaLink-based programmable logic device. In contrast, an SRAM-based FPGA is almost like an open book. An SRAM based device is volatile, which means it loses its program when the power is turned off. Its must therefore be reprogrammed each time power is turned on. The bit stream that is used to program the SRAM-based device is easy to capture and therefore copy or to reverse engineer. In fact, there was a company here in Silicon Valley that made a business out of supplying ASIC devices built from the programming bit stream of FPGAs.

That is they had a business, until a large FPGA supplier sued them for copyright infringement of the bit stream. The point here is that there is virtually no design security afforded by SRAM-based devices. Now this has actually always been the case. So what's the big deal now? Well, more customers are using the FPGAs in place of ASICs. And the size of FPGAs increased and as the size of FPGAs increased, customers began putting more of their proprietary design elements into them. For example, going back to our Quick MIPS II customer above, their compression algorithm is highly proprietary and would be very valuable to their competition.

Putting this high value intellectual property into our secure programmable fabric ensures it will continue to be proprietary. Protecting valuable IP is vital to many customers, and ViaLink-based devices from QuickLogic are the answer. We can prove it. Okay. Thank you, and now Carl will give you the details of the Q4 guidance. Then I'll come back and wrap up.

Thanks again, Tom. Tom mentioned our low turns in Q3. On our last call we said we'd return to the historic turns rate of 75% to 80% of revenue in Q4 or Q1. We also mentioned that our bookings rate declined in Q3. As a result, we entered in Q4 with less backlog for Q4 shipment than desired and we need to have a turns rate of 70 to 75% of revenue in Q4 to have flat revenue. The good news is that our weekly booking rate has improved significantly since the end of Q3. However, with lower visibility for Q4 and lower beginning backlog, we are providing guidance that revenue will be flat to down single digits sequentially in Q4. Based on this guidance, we expect fourth quarter revenue to be 21% or more higher than the fourth quarter of 2002, and expect our 2003 revenue growth to be 27% or more compared to 2002.

One customer contributed 11% of revenue in Q3 and has contributed 17% of our revenue so far in 2003. Its appears that revenue from this application will be winding down over the next few quarters. We are currently engaged with this customer to provide a complete systems solution using our ESP products in a new system. Gross margin which was higher than planned in Q3 is planned to be between 46 and 51% in Q4. We expect that our sequential gross margin may decline due to higher overhead costs and due to our reduced benefit from the use of reserved inventory.

Our company's manufacturing strategy is to reduce the cost of producing our ESPs and FPGAs so we can pursue higher volume sales opportunities. We believe that fabricating products at Tower will reduce our unit costs. We are also actively reducing the other costs of sales by reducing the time and cost needed to test and program our products and by negotiating lower assembly costs. As Tom mentioned, we made significant progress with Tower Semiconductor since our last conference call. We shipped QuickMIPS II and Eclipse II engineering samples produced by Tower to customers in Q3. We expect to ship production orders to our customers in Q4 and to sample the balance of our Eclipse II product family.

We expect our research and development expenses to be sequentially higher in Q4 primarily due to the purchase of preproduction material and qualification expenses. Q4 research and development expenses could increase $300,000 to $500,000 over Q3 levels. We expect that selling general and administrative expenses will be $200,000 to $400,000 higher sequentially due to marketing costs associated with the new products, compliance costs and new sales hires in Asia-Pacific. Interest income and other net includes interest income on invested cash, foreign exchange, gains and losses and interest expense on borrowings.

During the third quarter, the net expense was $4,000. We expect that interest income and other net will be an expense of less than $50,000 in Q4. We plan to use up to $1.5 million of cash during the fourth quarter. Inventory levels may be within $300,000 of Q3 levels. And we do not expect a significant net benefit from account receivable, accounts payable, prepaid expenses or other balance sheet items in the fourth quarter.

Tower is an important partner of ours providing us with advanced manufacturing technology. We have a strategic investment in Tower, and are currently renegotiating the final $3.7 million investment. As a result of these negotiations, we may make an investment in Tower during Q4. This would increase our planned use of cash in the quarter.

Let me take a moment to outline our financial strategy. Our primary financial goal is to return to profitability by increasing revenues and gross margin dollars. We believe the products in development at Tower will play a significant role in our revenue growth, and that significant sales of these products will occur in the middle of 2004. Our sales team is enthusiastic about the new QuickMIPS II and Eclispse II products and we're generating significant sales opportunities for these products. Customers appreciate the time to market, security and low total cost of ownership delivered by our QuickMIPS II product and the security and industry-leading low power of our FPGAs. Customers have specified their system designs with our products in mind, customers are designing systems today to work with these devices and some customers have started to qualify systems built with our new products.

As a result of our focus on profitability, we continue to carefully manage our expenses. While we expect an increase in project expenses in the short term, we are only selectively adding head count and carefully managing new hires. In general, we must believe new hires are essential to meet our new product or revenue objectives.

We believe our QuickMIPS II embedded standard product and our quick – Eclipse II FPGAs which are among the most secure and lowest-power devices available today are particularly attractive to the growing market for semiconductors in Asia-Pacific. We recently hired a managing director of Asia-Pacific sales and we are actively recruiting field applications engineers in order to pursue additional opportunities in this region of the world. Our financial model based on $25 million of quarterly revenue and stated as a percent of revenue is to generate a gross margin of 60% to 62% to have research and development expenses of 17% to 19%, and to have SG&A expenses of 19 to 21% of revenue.

This would result in income from operations of 20% to 26% of revenue and EPS of at least 20 cents based on 25 million shares outstanding. Our current expectation is to break even from operations in the second half of 2004 at $13.5 to $14 million of revenue per quarter. We expect that new products built at Tower will be a significant source of this revenue growth. We expect to be able to maintain break even operating cash flow at approximately $12.5 million of revenue per quarter.

Please remember, that we have made forward-looking statements in our presentation and we will likely make others in the question and answer period following our prepared remarks. Our actual results could differ materially. Please review our SEC filings for specific information on the risks and uncertainties that we face. Now, let me turn the call back over to Tom.

Thank you, Carl. I'd like to wrap up our presentation here today with my own perspective on the guidance for next quarter. We've guided that our Q4 revenue will be flat to down single digits sequentially. And our competition is guided to be up in the low single digits in Q4 over Q3. So why the difference? Good question. On a year to date basis, our total revenue was up 29%, while the sum of our competition is up only 12%. This basically says we're growing our revenue over twice as fast as our competition. For the same period by the way ESP revenue is up 58%. In Q2, we had a large Asia-Pac customer account for 27% of total revenue. And in Q3, the revenue was only 11% of our total revenue.

In spite of this significant drop expect as it was we increased total revenue in Q3 over Q2 by over 5%. Well ahead of the competition's 1% sequential growth. So what's my point? Well, we have had a large customer that we have benefited from, and we have managed to grow the rest of our business as well. Actually, our other business grew almost 30% sequentially in Q3 over Q2. Now, the large customer’s program is coming to an end over the next several quarters. We are working with them, by the way, on new designs that we hope could also bring us significant revenue in the future. So I view our Q4 guidance as being a function of the uncertain visibility associated with a high-turns business model.

For the first three weeks of this quarter, new orders have been very strong. Well ahead of the Q3 average weekly new order rate. Three weeks doesn't make a quarter in our business when you have a 70% new order turns requirement. So this is really a visibility issue. Or lack thereof. You should not read into our Q4 guidance any concern on our part about our long-term business. We are on course. Okay.

I hope that's enough about sequential guidance from me. And now for your scheduling purposes our Q4 2003 earnings conference call is scheduled for Wednesday, January 28, 2004 at 2:30 p.m. Pacific Standard Time. If you have the opportunity, we'll be presenting at the AEA financial conference in San Diego on November 4th and 5th. We're also going to be presenting at In-Stream partners 2003 financial conference in San Francisco on November 11th. We are also then expecting to present at the Needham annual growth conference in New York City the first week of January 2004. Details for all of these can be found on our web site or the respective web sites of the investment banks. Now let's open up the call for questions.

Thank you, sir. [operator instructions] Our first question comes from Apjit Walia from RBC Capital.

Thank you. I wanted to ask about this Asia-Pac out for the next quarter. You know obviously it has gone down substantially as part of your overall revenues. If you were to guestimate (ph) what percentage of revenues that it would be in the December quarter, what would that be?

It's going to range a little bit -- could be between 5 and 15%. We just don't have that visibility.

Wow. And essentially in the next few quarters following that, is it something which you think is going to stay around that or just go even lower?

Well, we think that we'll continue to have significant revenue over the next few quarters perhaps into Q2

Okay. So you see this range staying around 5% to 15% for the next few quarters then?

Yes.

Great. And in terms of you said there were two other customers 5% or 6%. Can you comment on who they are? Are they use based in the US.?

One is based in Japan (ind) our test equipment. That was our 5% customer. Our 6% customer is a big instrumentation test manufacturer who we ship to worldwide.

Based out of -- it is a US-based company?

Yes. It is

Great. And in terms of your -- I'm looking at to just the drop down from 42% to 18% for computing -- this is essentially where it breaks down for the Asian customer, because they have been seeing the biggest drop there for the quarter. Where do you see that the most?

The biggest drop is a $1.7 million decline in the revenue from that large customer.

That's why you're getting a big drop. If you can give me an idea for the next quarter where if you could if what kind of breakdown you think you might have in the by segments would it be similar to the current quarter or the previous quarters?

No, we haven't done that assessment yet, so it would be a guess.

Yes. I mean, maybe I should phrase it differently. Where would you see the most growth come from in the fourth quarter if you had to break it down by one of these segments? Where do you see the least growth come from?

Our instrumentation and test has been really good. We could see some increase in competing as well.

Military as well. [d1]

As you expect Military and (indiscernible)? And I was looking at equipment and Eclipse that's exactly the future in terms of what if you could say in terms of by end of '04 how much could equipment be part of overall revenues, on Eclipse?

For which quarter?

By end of '04, for example. If you think in terms of your --

By the end of '04?

Yes.

That's long-range planning. That would be just a -- I could give you goals, but that's different than giving you guidance.

Yes, Goals. I would say, where would you expect or hope QuickMIPS to be or Eclipse II to be by the end of '04?

Let me take a different perspective on that. We're just starting our planning process for 2004 so we're not ready to give guidance for the year. So, don't take what I'm saying as guidance. But as part of that planning process we expect to be up at least 20% in our revenue next year compared to 2003. And most of that growth is going to come in the second half from these new products out of Tower.

I can see so you are kind of back-end loaded. And you mentioned Asia. Do you have FAE's or Field Applications Engineers currently in Asia?

Yes, we do

And how many do you have, currently?

We have three.

Three, okay. Great. I might come back with a follow-up. I'll let other people have a chance to speak.

Our next question comes from Manaje Najkarni (ph) with Chip Investors.com

Congratulations on a good quarter.

Thank you.

Can you please give us color on the ESP revenues, how well you did with Quick PCI and Quick RAM?

I'll tell you it was a broad-based great quarter we had in ESPs when you consider that our ESP revenue from our biggest customer dropped by a million dollars. The rest of our customers made up for that and it was broad based both among Quick RAM, PCI, and V-3 products.

Okay and you mentioned that your sales in North America and Europe grew nicely sequentially more than offsetting sales to Asia. Can you provide more color on the sales in North America?

Again, it was very broadly based. We had numerous customers increase revenue 1 or 200,000. We had one customer that was up more even more than that. So it was really just a very good uptick in our business. Both in our FPGA business, and in our ESPs.

OK. It is fair to say excluding this one customer from China, you would see sales increase in the fourth quarter compared to the third quarter?

That could certainly happen.

But certainly expected it, actually.

Okay

The real source of visibility challenge for us is our largest customer.

Okay. Now, talking about your new products, when will you start shipping chips made at Tower Semiconductor? In production volumes.

This quarter.

Current quarter?

First I like this quarter the rest, the following devices next quarter.

Okay. And you announced earlier that you had started post production lots. How are you doing with the use, etc?

Well, you know, it’s a new production process at a new facility, so there are always engineering challenges associated with that. And this is no exception. But we are very pleased with our progress. We think we're on track. The customers who have seen the products are very positive on them, so we think we're going forward very nicely with it.

Okay.

Let me give you a little different flavor on it. The base level process is yielding very nicely. And the only issues now are the fuse and, you know, we've brought this fuse module up in probably seven different fabs now and five or six different technologies. So we have no real concern about our ability to do this. And the especially in conjunction with tower. So we're this is not a major concern for us.

Okay. Can you talk more about design wins activity for the Eclipse II and QuickMIPS II, how many different customers are evaluating these chips?

Well, we typically don't characterize design -- number of design wins for folks in these calls. Only because everybody uses a different criteria for what's a design win. I can tell you that let's take Eclipse II. The beauty of Eclipse II is that it is pin for pin compatible with Eclipse I or Eclipse and Eclipse Plus. And so because it is built in a smaller or a finer geometry, we can offer cost savings to customers who want to make those conversions.

So we have people who are in the process of doing evaluations, their own evaluations right now, their own qualification testing, and actually have orders on the books for devices from them for the Eclipse II products. I don't know exactly how many there are there, but I would suspect it is in fact across a broad range, including gaming, test and instrumentation customers as well as computing customers. So this is not a one or two customer deal. This is people that have been queued up and waiting for product.

So the -- so -- yeah, go ahead.

Quick MIPS II is little different situation. There we’ve had Quick MIPS I out there and people have used this to begin designs and have used it successfully, actually to go into –- in some cases in preproduction. And the challenge there though is that it just takes longer to get those products to market, primarily because you've got both the software and the hardware challenge. But we think we'll see some revenue -- significant revenue certainly by Q1 perhaps a little revenue in Q4.

And any Tier one customers potentially among them?

Yes. As a matter of fact, Tier one customers for both Eclipse II and for QuickMIPS II. The one I told you about that was doing the interface of legacy systems to packet-based networks is in fact a Tier one customer.

Okay. And what are your plans for the transferring more products to Tower Semiconductor?

We haven't talked about those publicly at this point. And we're in the process of formulating those plans. And when that formulation is done, then we'll share that with the outside world.

Okay. All right. Thank you.

Thank you.

Our next question comes from Larry Boardman with Cantor Weist.

Hello Tom. I wanted a little clarification on gross margins for next quarter. I wasn't -- I didn't quite get the range. Then I wanted to understand if they were a function --excluding that is whatever product you might sell that had been written off, if they were a function of product mix, volume or yields?

Let me take that one. Our forecast is 46 to 51% of revenue for gross margin in the fourth quarter. And really that's a function of lower expected benefit from the sale of reserved inventory. Reserved inventory provided 3.2% of our margin last quarter. It will probably be between 1 and 1.5% of our margin in Q4. So it was a significant dip there. You know, our production activity could be down in Q4 and as a result, we would have less absorption that's the other driver that could take our gross margin down sequentially. Otherwise, you know, we're very happy with our standard margins of our products. We have been selling, they have been very healthy and the rest of the margin indicators have been good.

Would you say your standard margins are pretty much fixed at this point or what yields you're getting on the ESP products, or is there still a learning curve there that you expect to benefit from next year?

Everything we have been shipping now has been performing very nicely in terms of yields and costs. In fact, we have had dramatic back-end cost reductions. Our engineering team has reduced in some case of the cost of programming and testing our products by 70% or more. So the product costs are looking great from our standard products. Our products we're selling today. We expect to have, you know, maybe some impact from low yields on the new products. But not until the first or second quarter next year.

And the good news is of course that while you're building -- while you’re supplying them in smaller volume if your yields are poorer than they're going to be obviously they don't have the big impact on you.

Okay. Thanks.

Thank you.

Our next question comes from Darryl Todd (ph) with SouthWest Securities.

Hello, Tom?

Yes. Good afternoon.

Is there any way you can give me some idea of the cash needs that you may need for the investment in Tower if you so chose to do that?

3.7 is what the final tranche is. If we elect to do that.

Okay. Thanks.

[operator instructions] Our next question is from Apjit Walia with RBC Capital.

Hello, I want to try to look at the gross margin question and if you could give me a break down in terms of the different segments if you have, what is higher than the corporate average or what is lower historically for computing instrumentation, data and military? Could you give me some color on that, please?

We haven't actually taken a look at that. We take a look at it more by product family. And again, those have been healthy. Generally, the standard cost of producing our products has been 43% or lower than our revenue. So it indicates a, if you will, a contribution margin of at least 57%.

Right. But let's go to the product and mature and new and imbedded, so if you could me a rough -- what would be the typical gross margin in the embedded standard product division right now?

We don't disclose those publicly. We keep those internal. We don't break them out.

Okay. You're looking -- if you could just guestimate gross margins for -- I know what you're guiding for the next, you have like a 500 basis points GAAP here, but would that something that you think will narrow going forward or do you think that would stay roughly the same?

No. I think in the short term the guidance will be that wide and hopefully by the end of next year we can narrow it down for you.

Okay. Great. Thanks.

Thank you. Apjit.

Our next question is from Mike McCormick, Gifp (ph).

Hello. You had given some competitive dynamics about the some of your new products and how they compared to the SRAM-based product sets out there. The smaller competitor, Actel has some flash-based products. Do those competitive advantages still apply under that situation -- under the flash-based products as well?

Well, certainly, the flash-based products stand a better chance of -- after their programs stand a better chance from a security perspective of not being read out, if you will. But the problem with flash-based products is they're very slow. So it's a performance issue then with flash rather than a security issue.

Now, in terms of utilization, I haven't seen the utilization rates. I would not expect their utilization to be as high, only because each switch is a transistor plus it's fundamentally a transistor –- a pass device, and therefore that takes -- more switches means more dye area means more cost. And so that's -- their switches cost more than a ViaLink-based switch which is up between the metal layers. So I would not expect that their routing would be as good either. So if their routing isn't as good and they don't have the number of switches then their utilization isn't going to be as good.

So those are the two areas. Power probably is not going to be as good either, owing to the fact that the impedance of the switches is higher. So from a security perspective, I don't know enough. My gut sense is that it may very well be secure. But from a power, logic utilization and routing, my sense is that they will not be as good.

And one question which is -- I'm relatively new to the Company. You talk about having a turns business which is 70% to 80% of any quarter, and at any one point in time. You're getting designed into these products, but if your product --your model is to not require lead time? I'm trying to understand if you're designed into the products you’d usually have some visibility. You just may not have a booking.

Well, yes. You have a visibility of where you’re designed in, certainly. You certainly have -- and we pay a lot of attention to where we are designed in and where we are being designed in. The challenge is that when you give the customer very short lead times, they order at the last minute.

Sure.

And what that says then is you don't know, you don't have big backlogs that you're operating out of. So you're at the mercy of what the customers are doing. And when people focus on sequential comparisons, the way our industry is prone to do, just take this last quarter's results and move $300,000 in sales from one quarter to the next and look at the impact of that on sequential growth rates, and then ask yourself is that really relevant to the business?

Right.

My position is it is not. But obviously I'm shoveling it against the tide here because the financial community wants to look at sequential numbers.

I understand. So I guess what I'm trying to think about when you look at your customer base, if you kind of forget what kind of happened in the last down cycle what kind of standard deviation from ordering rates and so forth do you see? Is it across the board wild fluctuations between your customer mix? You guys can look at it and look at your customer base and say with reasonable certainty these guys are going to go X and Y and so forth and so on, so maybe the visibility is a lot better than you may be leading on. It just may not be an order in hand.

Well, I certainly -- you know, you have the law of larger numbers going for you. You know, if you have lots of small accounts, you probably wind up with better visibility than you do when you start having certainly -- certainly when you start having a 27% customer.

Yes, that’s usually a problem.

And even 10% customers. And now you know for the first quarter we have three that are over 5%. So we're having bigger customers and actually that -- you would say, well, that improves your visibility. No, it really doesn't improve your visibility. I mean, we have our largest customer, we were booking and billing in the same quarter a significant amount of revenue. So that's the nature of having short lead times. Now, the argument could be, well why don't you go to the longer lead times? Guess what? The customers don't want that.

Right.

That's of significant value to them to not have to order six weeks in advance, for them to be able to wait three weeks.

Okay. Thank you very much.

Our next question is from Manaje Najkarni (ph) of Chip Investor.com.

Hello, just a follow-up question about your guidance for the fourth quarter. You are saying that you need about 70 to 75% turns business -- is that to get at the high end of the guidance of flat sales?

Yes. That's correct.

Okay. And historically, Tom, what is the highest level of turns business have you had?

81 or 82% I think.

OK. So you have a system in place, you can support up to 80% turns business?

Definitely. We could even support more than that if it was in the right mix. You just never know.

Okay

The real deal is we don't maintain a finished goods inventory, so we build to order and the question then is as you get closer to the end of the quarter if you book something where somebody wants it in that quarter, you run out of time to get it built. So that's the only challenge.

Okay. Thank you.

You're welcome.

Gentlemen, we have no further questions at this time.

Okay then. Well, then let me bring our conference call to a close. Thank you very kindly for your interest and what we're up to. We believe as we have all along -- that embedded standard products are the answer. It’s based on simple economics for our customers and that's what we're driving on a forward-going basis. Thank you.

Ladies and gentlemen, thank you for your participation in today's conference. This does conclude the presentation and you may now disconnect your line at this time. Have a good day.

[d1]Switched these lines to more correctly follow chronological order.