超微半導體 (AMD) 2001 Q2 法說會逐字稿

Ladies and gentlemen, thank you for standing by. Welcome to the AMD second quarter earnings release conference call. During the presentation, all participants will be in a listen-only mode. Afterwards you will be invited to participate in a question and answer session. At that time, if you have a question, you will need to press 1 followed by 4 on your telephone. As a reminder, this conference is being recorded, Thursday, July 12, 2001. I would now wish to turn the conference over to Mr. Robert J. Rivet, Senior Vice President, and Chief Financial Officer with AMD. Please go ahead sir.

Thank you Robert. Good afternoon and welcome to AMD's second quarter earnings teleconference call. With me here today in Sunnyvale and hosting this call is Mr. W. J. Sanders, our Chairman and Chief Executive Officer. Also present are Mr. Hector de J. Ruiz, our President and Chief Operating Officer, Mr. Robert R. Herb, Executive Vice President and Chief Marketing Executive, Mr. Walid Maghribi, Junior Vice President and President of our Memory Group and Mr. Benjamin M. Anixter, Vice President of External Affairs. I will begin by summarizing our second quarter highlights, and then turn over to Hector, and finally to Jerry. Today, we have representatives from the media listening in, but since this call is for financial analysts, we will only take questions from members of the financial community. Please hold all questions until we conclude our prepared remarks. This conference call is a live broadcast and we will replay via the Internet at www.streetfusion.com and www.amd.com. The telephone replay number for North America is 800-633-8284, and outside the US, dial 858-587-5842. The code to access the call is the same for both and that number is 191-51-508. Before we begin the conference call, I would like to caution everyone that we will be making forward-looking statements about management's goals, plans, and expectations. As you know, it is generally volatile.

Our product and process technology development projects and our manufacturing process are complex. Current economic conditions make it especially difficult to forecast product demand. Because our actual results may differ materially from our plans and expectations today, I encourage you to review our filings with the Securities and Exchange Commission where we discuss in detail the risk factors in our business. You'll find detailed discussion on our most recent form 10K filings with the SEC. Lets begin. Revenue. Revenue for the second quarter was 985 million, down 11% from the second quarter of 2000 for continuing operations, and down 17% from the first quarter of 2001. For clarity, when I mention continuing operations, we are excluding the voice communications business, now called Legerity from our historical results. We sold this business in the third quarter of 2000. In our microprocessor product line, we shipped in the second quarter an excess of 7.7 million units, a new company record, and aiding this business to grow revenues by 1% from the second quarter of last year. However, our microprocessor revenues did decline by 11% from the record levels in the first quarter. Revenues in our memory business declined by 13% over the second quarter of 2000 and it was down 23% from the first quarter. With our top line declining both year-on-year and sequentially, our [bill years] data block was challenged, but I am pleased to report that we made money in the quarter. ProForma operating income was 34 million or 3.5% revenue for the quarter as shown in our press release as our non-gap consolidate statement of operations, which includes the impact of FASL, our manufacturing joint venture with Fujitsu.

This represents a decline of a 155 million from the first quarter, driven principally by lower revenues in our memory product line and greater than 50% ASP decline in our microprocessor product line. Net income for the quarter was 70 million or 1.8% of revenue. Fully diluted earnings per share was $0.5 for the quarter. Gross margin on a ProForma basis was 36.7% for the quarter, down 11% points from the second quarter of 2000 from continuing operations. As stated previously, margins were significantly impacted by the revenue declines in our memory product line and the ASP pressure in our microprocessor business. Factory execution in both output and new productivities were better than the first quarter. Research and Development study was on plan at 171 million for the quarter, up 10% from the second quarter of last year. Marketing in general and administrative spending was also on plan at 156 million, up 3% from the second quarter of last year. Our cash balance ended the quarter at 1.1 billion, down 200 billion from the yearend, over half of this decrease, which were non-recurring investments in our FASL joint venture. We completed the quarter with $401 million of inventory, up 46 million from the first quarter. Inventories for both our microprocessors and Flash product lines are well positioned for our plans for the balance of the year. Capital spending for the second quarter was 215 million, about half of which was for Dresden and depreciation expense was 159 million for the quarter.

EBITDA was 181 million for the quarter and finally, we are pleased to report that we retired 517 million of convertible debt in the quarter, which brings our debt-to-debt plus equity ratio to 20%. For your modeling for 2001, please consider the following. Assume that our tax rate will be 27% for the year. We have reduced our capital spending to 900 million for the year, down 100 million from our previous guidance as we moderate our back end capacity needs. We will complete Fab 30 and instal 130-nanometer capacity on schedule this year. Depreciation expense for the year has been reduced to about 630 million for the year, down from our previous guidance by 40 million. We project our cash balances to remain at $1 billion for the balance of the year. For your EPS calculations, use 350 million shares for the third quarter. Finally, R&D investment will continue with the second quarter spend rate for the balance of the year investing in both our future technologies and products. I would now like to turn it over to Hector to start the business segment discussions. Hector.

Thank you Bob. I would like to make a few comments about our operations. Our execution across our factories continues to be excellent. We continued our ramp of Fab 30 in Dresden, Germany and expect to achieve full capacity in terms of wafer stocks by the end of the year. The demand for our Athlon products has being strong and we've sold out the output of Fab 30 in the second quarter. We have successfully made several product transitions as evidenced by nearly a quarter of a million 1.4-gigahertz parts that were shipped in the second quarter. Our plans continue on target for the beginning of the 0.13-micron conversion in the fourth quarter for both bulk and SOI based technologies. We are very encouraged by the progress of this transition as evidenced by the continued grade results in the fabrication of SDRAMs as precursor to the conversion. Before the end of the year, we will begin the conversion of Fab 25 to a Flash factory. The speed of the conversion will be dictated by the return of demand for our Flash products. This along with our manufacturing and partnering strategy and our current planned backhand and test capability will result in having ample capacity for both Flash and microprocessors for us to continue to gain share. Now, I am going to talk about our product lines.

Memory Group sales declined 13% from the second quarter of 2000, and 23% from the previous quarter. Networking and communication segments of some of the market remained very weak with some apparent stability in the set-top box and automotive markets. These conditions exist across all regions of the world. Flash memory density increased from the first to the second quarter by 17.4% or to 12.8 megabits and as customers move towards feature rich products in a lower unit environment. Our capability to continue to offer our customers increased density in Flash will be further enhanced by the recently announced technology innovation called [_____] bit architecture. The continued weakness in most market segments of these product lines make it very difficult to accurately forecast demand going forward. We currently expect that sales in the third quarter could be down as much as 30%. In the microprocessor business, virtually all of our sales this quarter were our seventh generation products. Both Athlon and Duron product lines established record unit sales in the quarter with the aggregate shipment being in excess of 7.7 million units. We expanded our portfolio there in the quarter with the introduction of the AMD Athlon IV processor for mobile PCs, which is the segment with great opportunity for us because our products offer unequal performance. We also introduced the AMD Athlon MP with its companion AMD-760 MP chipset for workstation and server applications. Additionally, higher performance Athlon and Duron processors were also introduced during the quarter. The 950 MHz Duron is the industry's highest performance processor for the value segment.

The 1.4 GHz AMD Athlon offers more delivered performance than any other processor as evidenced by industry benchmarks. Pricing pressures in the quarter resulted in ASPs declining significantly from the first quarter to the second quarter. While we do not expect the rate of decline to continue, we don't see an environment where flights and pressures will continue and ASPs could be lower sequentially in the second half of the year. We are prepared to adjust our cost structure and continue offering the best value for delivery performance so that we continue to gain share. I would now like to turn the discussion over to Jerry Sanders.

Thanks Hector. Well, today AMD is all about PC processors, which are 60% of our business and Flash memory, which was 32% of our business in the quarter. Accordingly, I will spend the time discussing our other IC businesses or the Foundry services we provide to some businesses, which we sold, and which are now operating as separate unaffiliated companies. Those businesses in the aggregate are less than $100 million a quarter and decline with the industry. That will continue. In the just completed quarter, our biggest disappointment was the absence of any recovery in the Flash memory market. Our revenue decline was twice of what we had expected. We were premature in believing that the second quarter was the tough quarter for Flash memories. Q3 Flash sales declined in dollars, as expected to approximate the decline we just experienced in Q2. On the strength of seasonal effects and new product cycles, we are confident that our Flash memory sale will resume growth in Q4. We continue to outperform in the market and are gaining market share.

The second part of our strategy of defend and expand in Flash memory will be evident in 2002 with the advent of [____] bit technology and greater participation in the enhanced sector. Our customer relationships are excellent and we continue to win awards, the most recent coming from the Robert Bausch Co. We are very well positioned, both customer wise and product wise to benefit disproportionately in an upturn. In PC processors as reported, we shipped a record 7.7 million Athlon and Duron processors with each setting individual records. Latest industry market research indicates we continue to gain market share reaching a level in excess of 22%. An unprecedented number of price moves in the quarter by Intel in an attempt to stall our market share progress was successfully countered, but impacted ASPs severely. ASPs for microprocessors, PC processors were down to about $75 in change. I remind you that on 7.7 million units, the $15 drop in ASP compared to the prior quarter cost us a $115 million in revenues most of which would have dropped to our bottom-line. In today's PC environment, price is particularly an important determinant in product choice. We plan to continue to offer superior delivery performance at a competitive price to that of Intel. In the value desktop space, our Duron products are clearly superior to the alternative by any measure. In the performance space, our superior delivery performance and substantially lower manufacturing costs have us well positioned to compete against Intel. Demand for our new mobile Athlon IV processors is strong and we are maintaining our maximum ramp in the Dresden sale.

In a weak PC market in Q3, we expect AMD processor unit sales to remain at record levels in the face of continuing aggressive price competition. Our working hypothesis is that ASPs will be no better than that in Q2 even on the richer mix. In Q4, we expect seasonal effects, the advent of Microsoft Windows XP, and a greater supply of power managed Athlon and Duron from mobile applications to enable us to increase PC processor revenues significantly. We remain on track to commence production of 130 nanometer Athlon thoroughbred devices in Q4 with samples available at yearend. In summary, AMD overall is executing exceptionally well in an exceptionally weak market. As indicated in the press release, if current business conditions prevail, our revenues could decline by more than $100 million in the current quarter, and we would report an operating loss. In this environment, we are intensifying our cost containment efforts where appropriate, in reducing or eliminating costs which will impact our near term outlook without affecting our strategy. We continue to make the necessary product and process technology investments that will position as well for the upturn. Our internal goal through a combination of cost control and aggressive selling is to breakeven in Q3 and return to solid profitability in Q4. Thank you and now I will turn it over to Bob for questions.

Thank you Jerry. Robert would you organize the Q&A please.

Thank you sir. Ladies and gentleman, if you wish to register for a question, you will need to press the 1 followed by the 4 on your telephone. You will hear a three-tone prompt to acknowledge your request. If your question has been answered and you wish to withdraw your polling request, you may do so by pressing the 1 followed by 3. If you are on a speakerphone, please pick up your handset before entering your request. One moment please for the first question.... the first question comes from Scott Randall with Wit Sound View. Please go ahead.

Great thank you. Two questions...first, on the transition of [alpha] towards Intel. Does that change any of the requirements on the bus licensing either currently or for the future roadmaps?

No, none whatsoever.

Okay. Future roadmaps, how would that work is that?

Well, our strategy for processors, it is to differentiate and plan. As you probably know, we do not have a compatible bus with Intel Pentium IV. We have our own bus and we intend to continue to improve the performance of that bus and offer an alternative platform to the Intel platform. When we go to our next generation, eighth generation Hammer family, again we have a new proprietary AMD technology, hyper transport, which again is the differentiated solution. It is our intention to compete with Intel not on a pin-per-pin basis, but as a superior alternative platform.

I understand and secondly, relative to the pricing and also given that you folks have done a good job of defending market share, but at a pretty big cost impact there. Is the flossy still to defend market shares sort of independent of pricing?

Pretty much. We are not going to be pushed out of the ring by a sumo wrestler. Basically, we've got a lower cost structure on our products than they have on theirs. When it comes to Duron, we've got a clearly superior solution, so the only differentiator is that we are compared to meet the pricing, we get the business. So, it is in our intention that whatever it takes in Duron. As Hector mentioned, we are selling out the production of the Dresden pay up. We are ramping at the maximum rate and so, what we are doing there is to negotiate on a customer-by-customer basis to optimize our product line. We are frankly a little surprised by the aggressiveness and the number of price changes that Intel have implemented last quarter, because they are really not doing nearly as well with Pentium IV as they had hoped. So, they are trying to push the Pentium IV down into the mainstream space where AMD is prepared to offer a competitive alternative. We've been giving the business, but there is no question about, it puts some pricing pressures on. The pricing though on Athlon is still in excess of a 100 bucks. The Duron pricing is substantially lower.

Great thank you.

Tim Mahon with Credit Suisse First Boston. Please go ahead with your question.

Thanks. Congratulations on the market share gains, Jerry, I was curious if you could give us a rough break up of Athlon versus Duron on a unit basis during the quarter. Was there a significant difference in the two?

We are reluctant to do that for competitive reasons and so, as much as I would like to give you everything you want, I can't really give you that. I can say that they are pretty evenly split at this point in time. Going forward, we expect obviously to sell more Durons and especially when we now have the power-managed, the mobile Duron, which we expect is going to be driving. Our target there is at 50% share of the US retail mobile market by yearend.

Great and then may be, Jerry you if you could help us out here, maybe some thoughts on, you know, what percentage of the process of shipments in Q3 you would expect roughly once again to be mobile processors?

Well, that's not something that I am prepared to give, may be Robert Rivet could help me out there. Basically, I believe that virtually all of our palominos are still being directed towards mobile, although we will be introducing a 1.5 GHz desktop palomino. That is the new version. So, the bulk of our units shipments from palomino will be for mobile and as far as Morgans go, which is the mobile Duron, I don't have a good answer. Rob can you help me here.

Yeah overall on the question of what percentage will be in mobile, less than 10% of our total shipment in Q3 will be for mobile market.

Great. Thanks Robin and finally, I do not know if you guys can give this out just kind of the linearity of the quarter for processor unit shipments? Thank you very much.

It was awful. It was the most nonlinear quarter in my memory. The customers were in a constant state of negotiation, re-negotiation, reallocation, redistribution, when, where, what and if to ship. It was much more than 50% of the year that we shipped in the final for a month.

Okay. Thanks Jerry.

Eric Rothdeutsch with Robertson and Stephens. Please go ahead with your question.

Can you give an idea of your speed roadmap on Athlon's review at the end of the year. I was wondering if you will be shipping the 1.7 GHz Athlon and faster versions?

As I said, we reduced the 1.5 GHz in the current quarter, the 1.733 is the next major speed rate and we are targeting the end of the fourth quarter for that product.

Thank you.

Next question comes from Mark Edelstone of Morgan Stanley Dean Witter. Please go ahead.

Yeah Jerry. A couple of questions, If I could, first one is the overall census to the health of the PC market, both as you went through Q2 and then kind of your observations here going into Q3, and then based on your comments about the non-linearity of Q2. How do you think Q3 looks from a linearity point of view versus the way we just had in Q2?

I will let Rob answer that since I am holding him responsible for these aggressive sales figures.

Yeah Mark. As far as the overall PC market, you know, we are to believe now that the PC market for 2001 can be roughly flat than what it was in 2000. I think most of data vendors are still looking at 2% or 3% growth that continued to deteriorate you know month after month here. So, I think what we are looking at right now is a roughly flat PC market for the year. Relative to linearity for the quarter, Q3 is typically a fairly backhand loaded quarter because you obviously get ready for a very strong Q4 PC selling season and a lot of processors are shipped in September into that. In addition with the Window XP timings, I expect that to be even more sold this year. So, I think we would have another quarter where we are fighting the linearity well and we are going to be relatively backhand loaded.

Robert, as you read that based on the rate of shipments running in Q2 that channel inventories are in relatively good shape here going into the third quarter?

Yeah, I think that the channel has got very sensitive on our managing inventories. I also made the channel while watching Q3 very closely because of course they also want to make sure that they are prepared with [___] based on XP as early as possible so that means monitoring closely what happened this quarter. I believe if you look at the US retail, the largest single segment to look at from a consumer standpoint you know, I think that we are planning Q3 year-on-year to be down probably in the 30-35% range which I think is a little bit pessimistic, but I will say that I think there is a lot to be returned to managed inventories prior to the XP launch.

Thanks a lot.

Next Question comes from [Tay Kim from Welch Capital]. Please go ahead

Mr. Kim, your line is now open for questions.

My question is up on the geographic breakout, rising in Europe related to the US in terms of enduser demand?

Yeah, basically Europe was, and you have got to put into both businesses... so I will first talk primarily about key business. So Europe was actually strong as it went through Q1. In Q2, you know, clearly I think European economy slowed a little bit. I know that the professional part of European market was down. I think it has also caught up in consumer market a little bit. So, clearly have seen a slowing in Europe, although I still don't think it's a nearly as bad as what we've seen in the US. Relative to the Flash business, I think it's safe to say that we still are slowing in all segments to the world and in all market segments.

Yeah... Great thanks.

Hans Mossesmann with Prudential Securities. Please go ahead with your question.

Yeah, thanks. Can you, Jerry, give us an update on your 300 mm strategy? Thanks.

Well, as I said many times what counts is the die cost or the ultimate cost. Our view is with our very small guide sizes and moving to a 130-nanometer technology, we have adequate volume for the next several years to achieve our 30% world market share of units. So, our 300 mm strategy really ties with the comment that Hector made earlier today when he said our manufacturing partnering strategy. We have a number of discussions that are ongoing so we are not at liberty to talk about it at this time, which will also going to result, I am sure, in our 300 mm announcement, but for the near term be it in the next year or two we think that we are going to have a lower cost on our current 200 mm wafers with our die sizes, with our technology, than our competitors, and that's what really counts.

Okay, and as a followup, you mentioned [NAND] Flash in 2002. Is that mean that you are coming out the [NAND] architecture or is it your near bid that is going to allow you to go up to those guys?

I do not, but Walid Maghribi will the answer that but those are two very different market opportunities that Walid will already discuss.

Actually we've had [NAND] products for a long time and the reason we did not push them in the market it is because we were totally sold out in the year 2000 and even in the first quarter of 2001. Now, what we are coming out with is a more advanced technology and higher density and the first product will be introduced very, very soon, we thought shipping probably in Q4. The [NAND] product is geared towards data storage where the [____] bit architecture, which also gave us a very high density, is geared towards cost reduction of node where the code shadowing is utilized such as networking. Then, we will have three products that address three different segments to execute in place, the code shadowing, and in the data storage.

Okay, great thanks a lot.

Hans, there is one thing I would like to add to that is up until now we've really had the most effective counter to [Strata] Flash. [_____] bit directly attacks [Strata] Flash and that it is competitive cost and there is no compromise in reliability or endurance. The customer has to make a big compromise and use [Strata] flash and he does not get the one-million-cycle endurance that AMD guarantees. [ _____] bit will give him that. So, we think that you might know the 28% Intel's Flash business [Strata] Flash. So, there is a 28% of market that we haven't been able to go for. So, [___] bit gives us a part of our extension strategy.

Thanks.

Dan Niles with Lehman Brothers. Please go ahead with your question.

Great. Thank you... couple of questions, if I could.... the first one Jerry, is to try and understand or maybe, for you, Rob, what your views are on inventory and I know you said that you thought that they were in good shape. It gets a little bit confusing, at times like you shipped about 50% of your microprocessors in the last month and it sounds like Intel's quarter was pretty back end-loaded as well, from what I can tell, yet on the PC shipment side, it seems as though shipments kind of fell off pretty, or they were fairly non-linear in terms of, third month didn't ramp that well for names like Compaq and other companies who are yet to report. So it would seem as though inventories have actually essentially increased and I am trying to figure out when it might be just connected there?

Yeah. Couple of comments, first of all on the AMD front which I can speak a little more confident on, because I do quite know what Intel did at the end of their quarter and what they did, but from an AMD front, if you will remember, we are still fairly heavily rated in the consumers space, and if you take a look at what was traditionally the refreshers in the string, we were pretty much nonexistent in US retailing, and we did almost no business there. So, it is not unusual that our early part of the quarter was lower than normally would have been expected by the factories out in that business. We think we're very strongly positioned in the back school starting seasons for a lot of our shipments, again of CPUs happen in June and going into July, August, and September selling season. So for us it was a big month because we went from being virtually not at all engaged to being in a very strong position in this upcoming PC zoning season. That's the case of processors linking PC sales from an AMD perspective and very good involvement. Besides that I feel a little less confident, let us go on to the commercial world and see what are the CPU shipments into the enterprise space versus the PC shipments look like. It has been clear to me that our competitive ships roughly have been able to do a, say roughly, 26 million PCs a quarter; that is 2 millions a week, and you know, shuffling around pulling here and pushing out a couple of week worth inventories, in CPUs. So we would regard our efficiency depending on your strategy on what number we are trying to make out there.

Okay another question, I guess, I don't know if there is an ASP associated with this but the commentary I think you made, Jerry, is that may potentially revenues could decline 100 million in Q3 relative to Q2. How much of that is driven by Flash versus processors and tied into that are you anticipating in that number losing share to Intel processors, I guess?

I am glad you asked that question. Let me clarify. Our press release said 10-15% decline in revenues, my safe bet would be 10% of $985 million is 98 million. Therefore, I was giving you a range of $100 million. Flash declined $95 million Q1-Q2 and I said we expected the same kind of decline in Q3 over Q2. Therefore, the first 95 to 100 million comes from Flash. That maybe the only decline we get if we can maintain our revenue base on microprocessors. Since we plan to hold our units at record levels that would say that any decline in revenues coming from processors would come from a decline in ASPs. I am not prepared to commit that we will have a decline in ASP. We've got a great mix and it depends a lot on what the market place is like. We are trying to be very, very conservative there. We are going to sell our heads off, if you will, to optimize our pricing. We've got a good line up for the Mobile markets that we didn't have in the past. We are ramping the Palomino, that is the power line of Athlon as fast as we can. We will be introducing a 1 GHz Duron this quarter. So we expect to get some shipments of that at the end of the quarter. So I think that the answer to your question is the softest market out there is the Flash market. That is just the top. We got blind sided last quarter; it just didn't recover at all. We are just assuming it is not going to recover this quarter and sales could be down at another 95 million.

Okay and a final question, I may?

You will not lose market share.

Okay. Congratulations on the job you have done, by the way. I guess the final question is just sort of on a margin question. Your big competitor out there is, and you know, obviously this quarter, they sort of guided margins for Q2, they were, sort of reported somewhat in the 48-49% range, but the guidance for the full year is 50. So, there is a sort of, anticipating an upward trend in gross margins at least if you believe the last guidance they gave. Your margins have gone from about 48% middle of last year to 40 in Q1 to, I don't know to what it got to....

35, I think in this quarter. Yes it is pretty ugly.

Right. So I guess, how do we do we sort of view some of the discussion here around pricing versus you are saying you have a lower cost processor and then sort of the margin differential potentially going in opposite directions where Intel potentially gets better and for you guys, I guess, it probably is a little bit worse.

I don't see how Intel can get better, frankly, so lets see how the results actually commit. It seems to me that if Intel is going to try to take the Pentium IV with its cost structure down and compete with the Athlon, they are going to get hurt in their margins.

How much more cost structure do you think is in Pentium IV versus Athlon.

I'd rather not answer that off the cuff, I can tell you, die cost is almost twice as large. So the difference in cost is pretty huge.

And even when you go to 0.13, what is the difference?

Yes, even when we got to 0.13 it is still 70% larger.

Thank you very much.

Your next question comes from Doug Lee with Banc of American Securities. Please go ahead.

Thanks. Two questions. First, with respect to relative profitability of Flash between Flash and end-microprocessors, is it safe to assume that Flash is now operating at a loss?

Not yet.

Is the profitability about the same between the two at this point?

That's more information than we are prepared to reveal.

Okay, that's fine. Thank you for that. And then on the operating expense side, just maybe a question for Bob there. Bob, I think you mentioned in your remarks that R&D should be about flat sequentially. Did you make a comment on SG&A spending cut to think about that for mileage purposes?

I did not specifically comment, but in this environment you can assume kind of the same.

Okay about flat, on a dollar basis?

Yes.

Okay. Thank you very much.

[Jim Tentori] with IBC. Please go ahead with your question.

Yes. Could you further explain just how [_______] bit compares to [Strata] Flash, or is it really just the same type of product with two bits per cell?

It is two bits per cell, but it is a very different product, Walid will you...

Okay. So, in [Strata] Flash, you get one fixed charge divided into two bits and that's the affected liability in the speed of the product. We are using the [____] bit architecture. We have two individual bits in one cell. Each one has the same charge. So, we are able to ship twice as many bits per unit of silicon. So our cost structure is at least as competitive to [Strata] Flash, and in addition to that there is a process simplification with this architecture that adds somewhere in the 15-20% of process cost reduction. This is a major difference between the [_____] bits and MLC or [Strata] Flash is that in [____] bit architecture there is no compromise on product performance or reliability, it is therefore pin-compatible and software compatible with existing node product.

Thank you.

[Vincent Bernadette] with [_____]. Please go ahead with your question.

Thanks. [Eric Rudolf Vincent]. If you can provide any followup to Mark Edelstone's on general conditions in the PC market. Can you provide any anecdotal information that would lead us to see any uptake for further September, back to school selling season?

I cannot, maybe Rob can.

Is he still there? Hello?

Sir, I am sorry. He just disconnected his line. We will go ahead and proceed with the next question. [Steve Allen] with [Hobartech Pomeranz]. Please go ahead with your question.

Yes. My question is regards to your investment in SOI. Is AMD going to develop their own internal SOI fabrication capability or they are going to rely on the likes of IBM?

This is Hector answering the question. We have had now for sometime the technology alliance with Motorola where these two companies have jointly developed the SOI technology and what we are primarily putting in production in Fab 30, which will start production at the end of the year, is based on this technology that was done between these two companies. In addition to that, you may remember that in the last six months or so we made a public announcement that we also license the design kits from IBM for SOI. So, we can have a combination of capability on the design side coming from IBM, some help and on the technology side the manufacturing fees is actually coming from the alliance between Motorola and AMD and that is what is being implemented in Dresden and we are currently running pilot production from that technology.

Is this likely to be a large-scale production or is this likely to be just a small niche?

100% of our production will be going to SOI.

A 100%?

100%.

Is this Jerry?

Yes.

Jerry that's freaking huge.

That's freaking huge guys. We are going for 30% market share. We are going to get there by offering a superior value. Our HiP-7 technology is the 130-nanometer SOI, that's where we are spending the money in Fab 30. That's why we built Fab 30 and that will be 100% capacity in Fab 30 by the end of this year. We are ramping to a 100% SOI capacity as soon as we can, in 2002 and in 2003.

Well, I am glad I stuck around long enough to ask that question.

Well, we believe SOI technology is crucial for high performance microprocessors. We don't see how you could get to where we need to without it.

Thank you very much.

Dan Scovel with Needham and Company. Please go ahead with your question.

Yeah, a couple of questions. You've done very well in the market share side now. Obviously, there is, the selling prices are sort of denying a savory win here. What are you doing on a platform basis, I guess to either save costs or combat any price aggression? Is there anything you can do in the short or mid term?

Yeah, clearly, we continue to work the platform issue. If you look at the low end space and the value space of Duron, we now have a wide availability of unified memory architecture platform, which are the lowest cost platforms to do that in. If you look on the Athlon side, the DDR prices that continue to have DDR Memory prices, continue to come down to the point there are virtually at parity or very slight premium to sequence the round today and all of our Athlon products are capable of running on DDR platforms. So not only do you get a very high performance platform, but you get a performing space, you get a very cost competitive platform. And as Jerry mentioned earlier, the course on the silicon is probably the most directly controlled. We believe that our die sizes are more than competitive and can give us an advantage in both the value and the performance space.

Can you expect to see incremental advantage on the UMA and the DDR in the next few months over what we've seen in the last few months?

Yeah. The DDR memory prices in the year are down at the platform levels. That is an incremental gain, you know, month in and month out as we go forward. On the UMA stage, you know, there were limited amount of motherboards available in the beginning of the year, we pretty much have a full range of motherboards available now, and I expect that the Duron sales will ramp considerably as a result of the competitiveness of our products and the platform competitiveness from the cost standpoint as we go forward.

Are you constrained by Durons sold by motherboards at this point?

No, I wouldn't say we are constrained by motherboards at this point.

Okay. Moving to cost cutting internally. Can you give us a few anecdotes that make us comfortable that you are going to hold the rein on internal costs or cuts in cost? I know that the head count went up quarter to quarter.

Let me first of all say that it is the time latency involved in any of the decisions we make. We began actually containing costs late last year in the fourth quarter and we have intensified, as Jerry pointed out in his remarks at the end, we have intensified our cost containment measures. And we are, whether it makes sense, either canceling or totally delaying things that we think we can live with without sacrificing our strategy and long-term vision. We are basically not hiring, except for key things that we need to keep position. The one thing that we have alive and we are strongly in pursuing is our college program because we think that is the key to our futures. So, we are not sacrificing that. As Bob pointed out in his introductory remarks, we are cutting capital of $100 million from the previous guidance that we gave you, and in addition to that, we have an internal goal of cutting capital further if we can. So, we may go even further than that. There are a number of things that we believe are going to have an impact in helping us get through this difficult quarter coming up and as also Jerry pointed out in his remarks, we don't have any internal goal of accomplishing enough of those and perhaps it is doing slightly better on the ASPs and sales to have an internal goal of breaking even in third quarter.

Finally, can you talk about progress in developing the commercial markets?

I guess the first thing I would say is that in an environment like we have today, you don't see too many people sticking their heads up and trying to do something new and aggressive. So when you say the commercial market, you mean Dell, Compaq, HP, and IBM were making little progress, and people like Fujitsu and Siemens are making progress, people like NEC were making progress, and people outside the United States are making progress. And that's part of our differentiate-and-flank strategy. Basically, if everybody is offering the same product, they have only got one weapon and that's price; and that's basically what is going to be happening to these commercial providers if they don't offer some alternative. We believe that we can offer better solutions through these flanking strategies, certainly outside the United States initially, and then making progress in the United States. We do have one or two commercial design wins, which we think will go into production, but our experience has been that our enemy continues to make them an offer they cannot refuse and when profitability is so difficult to obtain for the PC maker, Intel has been subsidizing a few of those deals, but I think that if pricing pressures continue on Intel, there is going to be a limit to how long they are going to support the entire infrastructure.

One comment I would make on the commercial space. As we mentioned earlier, we think we just have a superior product offering in the mobile space and even HP has recognized that. HP has announced a commercial mobile product for sale in Japan. We are hoping to extend it to other parts of the world of course, but I clearly think the mobile products, the Athlon IV and the Duron for mobile are going to be our best chance to break down the barriers relative to the enterprise space.

I guess can you just, sort of, itemized what those barriers are and when you would expect that we should be able to see some you know meaningful design wins there?

The barriers are called Intel Monopolistic Practices. Packaging deals, buying deals, advertising collaborative deals, and the Intel Inside Program, that's the only barrier. Intel uses us a lot of money to throw at any place where our products are getting close to a win.

We were there, we held our first formal CIO conference this past quarter and was very successful conference, a very well received by all of the IT folks who have joined us from [____] and almost unanimously acceptance that they certainly are willing to both look at and if we live up to our promises, purchase AMD PCs. So now it is a matter of getting our OEM partners to a step up to the place and put those products in the market.

Have you imagined NEC does have a server, which is for the professional market that uses the AMD processors and the reasoning primarily, we have an opportunity with our flanking strategy is that they can't effectively gain any market share against Intel's favorite customers if they don't have an alternative and a differentiated solution. So I believe that the pressure that we put on by offering an alternative is going to cost some of these other guys to move more toward us and as Rob pointed, out the mobile space, we have clearly the superior solution.

Again thank you and congratulations on your market share and good luck.

Thank you.

Next question is from Adam Horowitz from Ulysses Management. Please go ahead.

Good afternoon, two quick questions. One is, can you just give us some heads up what you intend to do with head counts and another two was up than the three. It is actually very difficult. The second one is, I don't know if I missed this earlier, but what is the utilization of your plans currently.

Well, let me try to answer the first part and then let Hector ponder the second part on utilization if he wants to answer that. Our head count is going up because we are ramping production in Dresden. We haven't reduced head count elsewhere because this is a very short-term phenomenon that we are currently undergoing. We are obviously going to be looking at all of our call up centers but the realities are, we need the people in Dresden, and people in Dresden are not intangible with people in Texas or California, kind of a long commute. So we are going to certainly freeze hiring. We certainly are not going to be adding to personnel and I think that you will definitely see a flattening out of the hiring situation. Obviously, we are of the opinion that the five or six rate reductions of the FED, the tax relief programs that are in are going to have some beneficial macro effects, as that with the Christmas selling season for personal digital systems, improvement in cell phone demand, new products GPRS that what we've got here is one or two quarter phenomenon with difficulty and 2002 should be a much more normal year. So, we don't think that we have a disproportionate excess of employees, but we are probably little ahead of ourselves.

Back to utilization, FAB 30 in Germany is 100% utilized as per the ramp. What I mean by that is, we are ramping it as we fast physically and possibly we can. And as we've said in the call earlier, we are selling out all the output of FAB 30 and we believe that we will continue and so we are ramping that as we are able to do in a prudent manner with high yields, good productivity, and good cost. That we will continue and for that factor it is fully utilized and it will stay that way from as far as we can tell for a long time. In Fab 25, where we build practically totally all our Durons and will be the Internet Flash conversion is that were slowing down and traveling back somewhat, as a matter of fact, we are asking our employees to take one week per quarter of vacation between last quarter and the end of the year, a total three weeks, roughly speaking that will help us throttle back that factor at this point in time. It is still running at fairly high efficiency because it was a full-loaded factory and so one week out of thirteen gives you nineteen other, and we will continue to do that if necessary we will throttle even further. Other than that we are very confident that our strategy for utilizing our factory is manageable as the market recovered since flash, and the CPU production continues.

You know our objective is to outperform the industry, I would like to point out that our revenues in the first half of this year versus the first half of last year in microprocessors were up 5.9% and our memory revenues were 5.5% over last year, while both of those sectors were down through the industry. So, we think we have got the right strategies. We are going through a weak market environment right now. We think we are very well positioned for the upturn, which we think is forthcoming in the fourth quarter in to next year.

Thank you very much.

The next question comes from David Blue with David Blue Associates. Please go ahead.

Yes. I just wanted to clarify two things. #1. When you talk about the fourth quarter/recovery, is that based on the assumption that OEM customers have finished the elementary liquidation by ahead Q3 and that GPRs stones actually get going by Q4 in Europe at least. Second part is on the situation regarding the back-to-school season, shouldn't AMD get better unit demand than flat sequentially, given your high exposure to the consumer segment. I know this is a funny year where Windows XP might delay some business from Q3 to Q4, but I was little surprised about the flat unit volume.

Well, I think, I will let Walid Maghribi answer the Flash question. Perhaps, he will comment on the processor business.

Well, what you mentioned is correct. First of all, our customer basically bought nothing in Q1; bought a little in Q2, and we are counting on that they are going to buy very little in Q3. That is what we have in our current forecast. So by Q4, we believe that some of the customer inventory will be depleted, that is #1. #2. There are some new products that are requiring new flat architecture such as GPRs, for example, which require totally different Flash. Thirdly, Q4 is typically a strong quarter for cellular, telephone, and for PDA. So because of all these reasons, we believe that Q4 should be significantly higher than Q3.

Let me answer the thing on, I think, as far as the processors, David you know, we didn't say flat, we said they will be at record levels. That means that they will be at least at 7.7 millions units presumably. I remind you that 5% of 7.7 will fetch you about 8.1 million units. This is certainly within the realm of possibility. So, when you look at what the market is likely to do, what is the seasonal difference between Q2 and Q3, when you consider the fact that lot of our businesses are also in Europe and Europe is traditionally a slow market in the September quarter, what we try to put together is the conservative view of what is going to happen in Q3 after the disappointment of pricing pressures in Q2. So, we think that if we can maintain ASPs no worse than in the range what they were in the last quarter and get the units at the record levels that will be of superior performance and we will gain market share.

Thank you.

John Joseph with Solomon Smith Barney. Please go ahead with your question.

Yeah, good afternoon. Sorry, if this has been asked, I got dropped, but with regard to the notebook win 60% of the US notebook market by the end of the year that should mean that you have got some pretty locked in design wins. Do you have a very high degree of confidence in those wins and then the second is, the Asian market, how are you penetrating in the Asian market in the notebooks space?

Okay, let me answer, may be I misheard you John but what we said was our goal was 50%, I thought I heard you say 60%. I am not ready as yet to commit to 60%. We are looking for 50% of the US retail mobile market by yearend. That is based on discussions with our customers and with the channel. Robert, you want to take it from there?

Yeah, obviously, we got a very successful launch of the mobile stuff in US with three very strong partners; selling with the first partner was Duron-based mobiles, both Compaq and HP has been selling mobile-based on the new Athlon core products and I think if we look at our wind up through the balance of the year, we have the opportunity to continue expand our presence and to grow our shares. So, I think it is a very aggressive goal, but we are able to gain market share, I certainly think it's something that we can accomplish.

We are going to introducing 1.1 GHz and 1.2 GHz Athlon for this quarter, which is in very good shape by yearend.

Jerry, I just got off the phone with one of your peers who has been in the business for 40 years. He said that this is the sharpest downturn that he has ever seen. Can you reflect a little bit on this downturn and 1985 downturn and where your position was and how fast or how slow the inventory and the return to growth might be?

Well, I am not old enough to have a peer within the business for many years, but I am close. In any event, this is the sharpest downturn I have ever seen. In my view, all is laid at the feet of the communications bubble. It is all about the communications bubble. All of the over investment in communications networks, all of the entire fiberoptics, all of that put tremendous demand on the semiconductor industry or that put demand for servers and PCS. All of this could be laid to that, and that is going to take the work its way through. I don't think it is the same kind of downturn as in 1985. I wouldn't say in 1985 revenues in the industry dropped, I think, about 30%. I think the industry revenues will drop in that range in this year. So, I guess in that regard, it is very similar, but the reasons are very different. I think that the communication scene is not going to get well fast and I would say, as you know, our strategy is the lowest cost solutions. For the market opportunities there, we are going to be the guys that are standing on the other end of it.

So, do you think that the impact on the PC market was a sort of collateral damage from the excesses in the comp space?

Totally, collateral damage. I think that the PC market got ahead of itself because of lot of companies buying PCs and servers that would not normally should even have in their mind to buy anything with, just say, the biggest, you know transfer well into the history of mankind. As a result, I think that we have got some top comparisons and so, we are still talking about a 150 million-unit PC market. That's nothing to be ashamed of, you know, 30% of that 45 million unit can make a lot of money. And then one thing that I would like to say is that we are not really happy with our current opportunity for profit in this quarter. Our long-term profitability, our profit making capability is unimpaired. We still have the leading edge technology, lowest cost die, great products lined up. There is a great product lined up there, just a market there and I think the reason is not just the inventory liquidation, it is a question of you know the people didn't start buying because there is no new opportunities those companies are going.

Do you think that the recent pricing pressure is mostly a weakness in the PC market or is it a change in relationship between Intel and AMD?

I think the reason for product pricing is that the P4 is a dud. I think that the P4 is a lousy product and they get a price at much lower that they have ever priced a new product before and there is no advantage to the user. On top of that, they made a lot of noise that they wont give up any market shares. So between these two things, having a second rate product and a wrong call on the memory technology for high-cost solution in a market place that wants lower cost solutions. Remember there are two things that always have recovered in the semiconductor industry. First and foremost is the new product cycle and second is economics, driving lower and lower cost. AMD is at a very good position with Duron to do that with our Athlon product line; Pentium 4 is a loser and the only hope they have with Pentium 4 is to invest the tremendous amount of money to build capacity for a 130 nanometer. So, they will only be marginally uncompetitive. They are basically, you know, it is the world's worse return on incremental capital since the communications we had for the last year. You have to tell them that. You have to send them home.

Okay I'll do that. Question, earlier in a conference call three months ago, you have said that the second half was going to be up from the first half year you are implying, sequential increase may be slight flat to slightly up in Q3, probably seasonally up in Q4. Are you reiterating that claim and what kind of growth do you think we are going to expect in 2002?

Well, it turns out with the industry, the industry just continues to get worse it seems, and so, I think currently you know the estimates are that the industry could be down 30% for the year. I don't think that anybody knows what the industry is going to do the next year at this point in time. I think that's how we should be in a much better position to answer at the next conference call, but I think at this point in time that anybody is making forecasts for 2002 is highly suspect.

Okay thanks.

Next question is from [Zaheer Monazadein] with Bear Sterns. Please go ahead with your question.

My question has been answered. Thanks.

Robert, two more questions please.

Thank you sir. Matthew Greg with [_____] Associates. Please go ahead with your question.

In looking at Q4, I am wondering what your assumptions are for Flash revenue growth and the microprocessor units and ASPs to get you to solidly profitable, which I think, the wording you used in the press release?

Well, I think as Walid had mentioned earlier, many of our customers are coming out with new products and are going to use more of our Flash. There is a seasonal upturn in both processors and in Flash memory in the case of Latin America, it is the PDAs and cell phones relative to our processors, as I said, in the fourth quarter we will have a much more abundant supply of our Dresden manufacturing, Athlon force with the mobile space, as well as a substantial amount of mobile Durons and we think that they could be substantial contributors. So, to answer that, Windows XP, we think that there is every indication that we have high degree of confidence that revenues will grow in Q4.

Just a followup, Jerry. Am I, in the ballpark when I am using roughly up 10% on microprocessor revenues and same thing on Flash as a normal seasonal upturn in Q4?

That's a good question. We are all trying to figure out if that's a good number. But days and days, we are talking about, and I don't feel uncomfortable with that number.

Okay could I, I am still getting a mild loss in Q4. I am just wondering what you guys assumptions are to say, to be profitable.

No what I think is probably you are underestimating our diligence at reducing our cost.

Okay and thanks.

Next question comes from Eric Ross with Thomas Weisel Partners. Please go ahead.

Thanks guys, actually a followup to Matt's question. As far as reducing cost you are not going to be reducing your R&D cost, but you will be reducing SG&A and gross margins. Is that an assumption I should be making?

This is probably a reasonable assumption. R&D, we said would continue at the current spend rate and we will moderate SG&A appropriately.

Okay, and as far as some more target operating margins in the next couple of couple of years, can we assume that you'll ever get back to the mid to high 40% gross margins in the next couple of years?

Absolutely. Our long term earning potential is undiminished.

Your mobile Athlon 4 margins. I am assuming that there is a much better than your desktop margins, even in the performance side.

Yes, in general that is true to say that.

Okay. Last question, depreciation, I didn't catch that.

For the current quarter?

For the current quarter and expectations for the rest of the year?

Current quarter was $159 million, it is in our press release, and the total year number is $640 million.

Thank you so much.

Thank you so much everyone for listening in and asking a lot of good questions.

Ladies and gentleman that concludes your conference for today. You may all disconnect and thank you all for participating.