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 (45122)	1/7/1999 4:45:00 PM
From: Kevin K. Spurway	Read Replies (1) \| Respond to of 1571911

I don't suppose you have a link to the slides?

To: kash johal who wrote (45122)	1/7/1999 6:16:00 PM
From: Yousef	Read Replies (1) \| Respond to of 1571911

Kash, Re: "The 100nm gates look tighter than what Intel was talking about at IEDM presentation." Kash, Before you go "jumping for joy", you might want to investigate the Gate length numbers. I believe these could well be Leff (Effective gate lengths extracted from electrical data) numbers. Depending on the FET architecture (Implants, thermal cycles ...) you need to add 50 - 80nm to AMD's numbers to get to the Lgate (Physical gate length) quoted by Intel. For example, Intel's .25um process has an Lgate of ~200nm with an Leff of ~140nm. The reason I question AMD's Gate lengths is because putting 2.2V across an FET with an Lgate of 170nm would give reliability problems. Also Kash, this explanation of AMD's Gate lengths would also explain why their CPU's don't run as fast as Intel's (Pay attention Ali). Make It So, Yousef

To: kash johal who wrote (45122)	1/7/1999 6:23:00 PM
From: Yousef	Read Replies (1) \| Respond to of 1571911

Kash, Re: "Just got through going over the AMD analyst conference call slides. Fab: 1. AMD looking at $10k / wfr revenue." This is very interesting, Kash ... We can get a lot of info from this "little" piece of data. Making some assumptions ... ASP = $100/die and possible die per wafer ...One can calculate an estimated yield. My calculation would put it at about 30%. Not good, IMHO. What do you think, Kash ?? Make It So, Yousef

To: kash johal who wrote (45122)	1/7/1999 6:30:00 PM
From: Yousef	Read Replies (1) \| Respond to of 1571911

Kash, Re: "Just got through going over the AMD analyst conference call slides. Fab: 3. 98 wafers out are 100,000" Kash, Is 9000 wafers out per month ... or looking at it another way ... Is 2200 wafers out per week the definition of a MegaFab. Question ... How much of MegaFlop25 (tm PREngel) capacity was utilized in '98 ?? My estimates are around 50% ... How about your estimate ?? Make It So, Yousef

To: kash johal who wrote (45122)	1/23/2000 10:52:00 AM
From: milo_morai	Read Replies (1) \| Respond to of 1571911

Kash said "Process. Cs44e6/7 are 2.2V 0.25 micron with 170nm Gates 5 layer metal. cs50 for 99 is 1.5V 0.18 micron with 120nm gates with 6 layer metal. Hip6l for 2000 is 1.5V 0.18 micron with 100nm Gates with 6 layer copper ( copper only in Fab 30/ 25 will run with AL). I gave Dan3 the credit for similar info he posted in November of 99, but You are the MAN for finding the 100nm gate info before anyone else. Thanks Milo