SI
SI
discoversearch

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.
Technology Stocks : Advanced Micro Devices - Moderated (AMD) -- Ignore unavailable to you. Want to Upgrade?


To: Joe NYC who wrote (118263)4/3/2004 10:18:18 PM
From: Dan3Respond to of 275872
 
Re: all they need is some 10,000 parts for launch. It does not take a lot of wafers to make 10,000 90nm parts. 100 wafers, and even that is assuming bad yields.

Yes, and if yields are good, they'll be getting 200 winchesters per wafer. If they can devote 2,500 wspw, that's 6.5 million Athlon 64s per quarter, from half the FAB, leaving the rest available for Opterons, FX's, some XPs, and more Athlon 64s....

Rivet also notes that the 90nm Athlon 64's die size comes in at under 100mm² - below Intel's 90nm Pentium 4, which has a 112mm² die size. Of course, Winchester has half the L2 cache of Prescott.
theregister.co.uk



To: Joe NYC who wrote (118263)4/4/2004 1:31:08 AM
From: dougSF30Respond to of 275872
 
I don't think 939 is launching on 90nm. And I don't think it's a problem for CG to support only the 1 GHz level of HT 2.0, with future revisions (which occur roughly every 6 months, per Kevin McGrath) to support higher speeds. Analogous to the way DDR support evolves from revision to revision.

I thought HT 2.0 was backwards compatible with HT 1.0.

Doug