Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Politics : Formerly About Advanced Micro Devices -- Ignore unavailable to you. Want to Upgrade?

To: kash johal who wrote (81394)	11/29/1999 12:29:00 PM
From: Yougang Xiao	Respond to of 1572777

From Albert: 09:41am EST 29-Nov-99 SoundView Financial Group (Scott Randall 203-462-7246) AM Advanced Micro Devices (AMD) Company Update Advanced Micro Devices (AMD) Price: $27.69 Hold November 29, 1999 (FYE Dec.) F98 F99 F00 Curr. Last Yr. Ago Revenue ($M) 2542.1 2785 4246 896 595 686 EPS -.66 -2.74 0.94 -0.04 -1.10 .01 We expect AMD to introduce availability of a 750 MHz Athlon today. Originally planned for mid-December, today's introduction represents a slight pull in from earlier time tables. Importantly, we believe that the price point (1000 piece) will continue to be at levels comparable to INTC's PIII pricing at equivalent clock rates. For reference INTC's 733 MHz book price is $776; AMD's 750 MHz price is expected to be about $800. Volume price points for AMD's products have typically seen a meaningful discount from these levels. Although today's introduction will put AMD ahead of INTC in terms of bragging rights (INTC's current highest clock speed PIII is at 733 MHz), we do not see signs of a developing price war between the companies. At current levels of production for both AMD and INTC and positive unit demand trends we believe that the pricing environment should continue to be fairly benign. As we indicated prior to AMD's analyst meeting in late October, we believe that AMD could establish a trading range in the mid to upper $20s. Completion of the sale of its communications business, finding a partner for Dresden and establishing momentum in the corporate market are the next challenges in front of the company. Although AMD's introduction today of a 750 MHz Athlon could cause some investor nervousness on INTC, we believe that the current pricing environment continues to be benign: Historically, pricing wars in the microprocessor market have been caused by one of three factors: * INTC seeking to aggressively protect market share * AMD seeking to gain marketshare, regardless of price points or profitability * Other manufacturers (i.e. Cyrix or IDTI) establishing ultra-low price points that drag down the low end of the mainstream. Currently, we do not expect any one of these factors to trigger a price war between the vendors. * At AMD's current production levels- and given the strong unit demand- we do not expect INTC to use price as a major tool to protect its marketshare. Recall that currently AMD's marketshare is running at about 15%, a level that we believe falls short of causing INTC to take aggressive steps to regain marketshare vs AMD. * With AMD struggling to achieve breakeven (and move beyond) the company wants to maintain as high a price point for Athlon as possible. In addition, AMD's greatest challenge has been to establish Athlon as a premium level product; as such AMD is highly motivated to keep pricing at high levels. * With both Cyrix (NSM) and IDTI out of the microprocessor business- and with VIA not yet established- we don't see the kind of low price points (i.e. $25-$30) that would cause price pressure to develop from the low end. From Before: As we had previewed, AMD's 4Q has gotten off to a strong start for both the company's microprocessor and non-microprocessor business. Taken in total, AMD's non-microprocessor business could be up Q-Q in the high teens. AMD's confidence in being able to build 1 million K7's and ship at least 800k continues to be high. We have increased our Q-Q revenue assumptions for both AMD communications and Flash business to 20% and 21.4% respectively. We have decreased our K7 shipment assumptions slightly to 1 million from 1.1. million units. Recall that on Wednesday we had increased our unit assumptions to 1.1 million from 1.0 million. AMD' official K7 unit shipment guidance continues at 800k. For the first time, AMD showed air-cooled K7s running at 900 MHz. While the company has previously demonstrated 1 GHz K7's, this was accomplished using special cooling techniques. Interestingly AMD showed 900 MHz K7's based on 0.18 micron technology based both on aluminum technology from Fab 25 and based on copper 0.18 micron technology from Fab 30. We believe this suggests good head-room for AMD's 0.18 technology, with the Company's copper technology from Motorola likely to allow the K7 to achieve 1 GHz clock rates in 00. Strong demand for Flash continues, driven by a combination of strong wireless unit growth and increasing memory loading in digital and multi-mode cell phones. AMD has begun shipping 64 Meg Flash and expects to ramp volumes in 00. The Company expects to ramp significant additional volumes of Flash through a combination of increased production at Fasil (its JV with Fujitsu) additional capacity from foundry sources including Iwate and eventuall fab 25. We have increased our Flash contribution to $250 million from $225 million for 4Q. AMD's communications business is benefiting from the economic recovery in Asia with the Company indicated that its communications business could be up 20% Q-Q. AMD's core telecom communications products continue to address the analog line-card market. AMD's SLIC's and SLAC's (Subscriber Line Interface Circuits and Subscriber Line Audio Circuits) continue to see growth especially in developing countries. AMD continues to seek a buyer for its communications group, with closure likely, we believe, by mid 00. From before: AMD's longer term challenges continue to include: Management and employee turmoil: The unexpected announcement of Atiq Raza's resignation from AMD's board, its President, COO and CTO came at perhaps the worst possible time for AMD. With the launch of the K7 in process- and AMD's key challenge to convince OEM's to adopt the K7 into product lines targeted at corporate- the abrupt resignation of Raza has been one more concern for OEM's to wrestle with. In addition, we believe that Raza's exit has negatively effected both investor and employee confidence. Raza was well regarded internally to AMD and was the heir apparent to Sanders. AMD has, we believe, also lost some key engineers associated with the K8 program. While employee turnover at Silicon Valley companies is a fact of life, we remain concerned that Raza's departure could cause this trend to accelerate. Limited progress to date in moving into the corporate space: As we have noted previously, for AMD to gain marketshare and to move its blended ASPs to significantly higher levels, the Company needs to gain success in the corporate market. While Athlon's performance and its multiprocessor support positions the Company well from a technical perspective, we believe this challenge is still largely in front of the Company. Strategically, AMD's challenges may come down to one of scale. In a $35 billion market that requires $2 billion fab investments to be made every few years, the ability to fund these efforts is problematic. With AMD's microprocessor manufacturing efforts concentrated in only one fab, fab 25 in Texas, its difficulties are multiplied. Specifically, while Intel has the ability to migrate both processors and processes in multiple fabs, AMD has had to make all of these changes in a single location. Because of this, its manufacturing risks have been multiplied. Looking into 2000, when Dresden comes on line, we believe this could provide the company with in important benefit in that for the first time it will have two mainstream state-of-the-art fabs in which to migrate both processes and products simultaneously without having to concentrate its efforts all in one location. For AMD, the introduction of the K7 continues to represent a critical product transition for AMD. Success at positioning the K7 as a mainstream/high-end processor is critical to AMD's profitable participation in the PC/workstation/serer market. While we believe that the K7 is technically a very good product, we believe the challenge of AMD moving "up-market" will be significant with the burden of proof clearly on the company.

To: kash johal who wrote (81394)	11/29/1999 1:49:00 PM
From: vince doran	Read Replies (3) \| Respond to of 1572777

Kash - Congrats on your calls call; I bailed on my Jan 25s this am when Gateway said 'Athlons after Christmas', then kicked myself for leaving so much money on the table. Interesting how what seems old news here is seen by the market. I have a question about your reasoning re .18u bin splits: Since AMD did not announce the end of 550/600/650/700, and since we know that all Athlon wafers since Nov 1 have been .18u, won't those grades also be supplied from .18u die, thus leaving the point where 750 sits on the distribution open to question? It will be very interesting to see what the overclockers can do with the new chip; considering the amounts of money we are betting here perhaps we should buy one and do some research ourselves. regards, Vince

To: kash johal who wrote (81394)	11/29/1999 1:53:00 PM
From: Charles R	Read Replies (1) \| Respond to of 1572777

Kash, <The key is that they only introduced ONE speed grade. That means that from AMDs 0.18 micron process the parts are all 750 Mhz and above!!! > I have followed your thory with great interest to see how this is going to pan out. I am glad your hypothesis worked and matches with emperical data. Ignoring all the rumors anyone may have heard, it has been well known for a little while that AMD (since about Week 40 or so) is shipping pretty much all 650MHz chips for even 500s. This was on 0.25. Now, AMD would be shipping 700s or 750s in 0.18 as 500s. What irony! Great for overclocking but doesn't do much good for AMD. <Now there is usually a bell shaped curve around the speed mid-point. With good process control that speed distribution may be plus or minus 15%. What this means is that the bottom of the yield curve is yielding 750. That means that AMD can ship a decent yield with a 25% speed bump. And a 25% speed bump gets us over 900Mhz. > I am in agreement with this line of thinking. <I don't know about 1Ghz but I now believe that AMD should be able to ship 800/850/900Mhz speed grades in Q1 2000.> That sounds like a pretty safe bet. <Who would have thunk it??? > Actually a lot of people (including your truly) have been thunking that but given AMD's past most people did not want to believe that. Chuck