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 (26906)	12/18/1997 7:48:00 PM
From: Kashish King	Read Replies (2) \| Respond to of 1573924

A few months from now AMD investors will look back and say "geez, it sure would be nice to get back to the high teens again." Intel will been able to keep revenues pouring in and earnings piling up with the introduction of 3D graphics and sound as the modern day analogy to the math coprocessor. Considering the amount of Intel silicon which will surround their core product, there's every reason to believe Intel will see another doubling in price from these levels. I'm looking for a sub-$10 AMD before the end of next quarter. This reminds me so much of SIII it's spooky.

To: kash johal who wrote (26906)	12/18/1997 8:03:00 PM
From: FJB	Read Replies (1) \| Respond to of 1573924

Kash, Thanks for the information about TSMC and UMC. Do you know what the design rules are for those processes? I think the single greatest factor effecting all the chip companies at this point is demand. A 20% savings rate in some SEA countries really hurts us. They were going to be an important factor in PC growth before the debacle, now the Japanese market will actually decline and of course spending won't be high in some of the Tigers. Sorry for stating the obvious, Bob

To: kash johal who wrote (26906)	12/18/1997 8:46:00 PM
From: Petz	Read Replies (1) \| Respond to of 1573924

Kash, two counterpoints to your super-overcapacity doomsday scenario wafer capacity is doubling in terms of units shippable in 98 vs 97 and the marketplace is due to see unit growth of 10-20% depending upon you're favorite analyst. Actually, if you're assuming the Intel's capacity for units shippable is doubling, think again. Their die sizes in 0.25 micron are not coming down 50% but only about 25%. This is partly because the '98 CPU's such as Deschutes are more powerful and have more transistors than the '97 CPU's such as the Pentium II. (Katmai and Willamette will be even bigger.) Second, a die size reduction of 25% does not equal a 33% increase in units shippable because 1)only half the production will be 0.25 micron and 2)the yield for 0.25 micron (new process) will be less than the yield for the mature 0.35 micron process. Putting all this together, Intel could probably increase output no more than 15%. If AMD can really make 15M CPU's, thats the entire growth for '98 and Intel will be running at 85% capacity compared to '97. In '98 AMD, despite my theoretical calculation of 50M CPU's will not be processing 5000 wafers a week while ramping up 0.25. Secondly, the K6+3D will be coming out second half and its a 135 mm^2 chip. So the following four chips will be produced in '98 along with some guesses for % of production: K6 0.35 162 mm^2 45% K6 0.25 68 mm^2 10%(mostly notebooks) K6-3D 81 mm^2 30% K6+3D 135 mm^2 15% The blended die size is 124 mm, allowing at least a 30% increase in production compared to Q4, all other things being equal. They can probably process 20% more wafers in Fab 25 so now we're up to +56% production. If they get 50% higher yields than the abysmall 4Q number, they can get a production increase of (162/124)1.21.5(1.6M Q4 CPU's)4 quarters = 15.05M CPU's. This is very tough, but doable, and according to game theory, we should see what my opponent does if I make my "best move." Intel's best move is probably to drop the price of the non-P2 chips close to their marginal cost of production (around $50) since that is probably the only chip where their manufacturing cost for a given level of performance is below AMD's. This will hurt Intel's profits immensely because they only expect at 50-50 P2 mix by the middle of '98. AMD's ability to turn a profit would then depend solely on how fast it can ramp up production of the 0.25 chips and such profits would not happen until Q3'98. I don't think Intel will go as far as $50 though. I still like my strategy of holding long term AMD calls and short term Intel puts for the time being. Petz

To: kash johal who wrote (26906)	12/18/1997 10:23:00 PM
From: Brian Hutcheson	Read Replies (3) \| Respond to of 1573924

Kash , there is one thing missing in your scenario.. Intel has made a business in advancing MPU speed by selling the high end at exorbitant prices , this is still their plan . The only difference now is that AMD and Cyrix can play the same game . Not only does .25 give AMD almost 3x the capacity of .35 , but it also gives them higher speed ramp and also the ability to cram more on the chip . Hence the K6-3d 1st or 2nd quarter and K6+ in 3rd or 4th quarter . What this will do for AMD should be obvious as far as sales/earnings are concerned . So can you imagine them making as much as 5M K6s a quarter while introducing higher end products with higher ASPs . As long as they stay competitive the profit growth will be there . regards , Brian