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: Brian Hutcheson who wrote (39966)	10/24/1998 11:55:00 AM
From: Jim McMannis	Read Replies (1) \| Respond to of 1574005

Brian, RE:"An additional problem for them (Intel) is K6-2 is getting closer in speed , it becomes increasing more difficult to ramp speed as clock speed gets higher , the result will be less of a gap between K6-2 and PII . AMD no doubt is finding K6-3 a problem because of the L2 cache just as Intel has with Celeron A , that is probably why its introduction was delayed." -------- From all indications the K6-2-400 has an enhanced core and run equal to the Pentium II-400. AMD will not push this difference so as not to create confusion. Seems like the max for the Pentium II/Celeron core at .25 is about 500MHz. That's probably Intels ace in the hole....if that. The K6-2 will be running at 450MHz by the end of the year and AMD would remain a speed grade behind but with the enhanced core at least the K6-2-450 would be equal to the Pentium II-400. If Intel doesn't push to 500MHz then the K6-2-450 would mean the speed grades are at a standoff. All this without the K6-3 entering into the picture. I don't think we know that AMD is having trouble with yielding the K6-3 based on fault L2 caches but lets assume that in any ramp-up of new technology that there is a problem. AMD apparently KNEW (they planned the K6-2+ well in advance) that this would be a problem so rather than waste a bunch of wafers by trash-cannning K6-3s that have failed L2 caches they simple choose to introduce the K6-2+, the K6-3 disabled L2 cache. Since these K6-2+s are at the highest megahertz rating and will sell at an ASP of over $200, AMD can make a good profit despite the 117mm die size and higher cost...even if AMD has to disable the L2 cache in "good" K6-3 chips to fill orders. When the yields on good K6-3s reaches a certain point, AMD will simply start to ship them. "Purest" process guys like Elmer and Yousef are really missing the mark, IMHO. AMD is working around their ramp up problems. Same for K7. Jim

To: Brian Hutcheson who wrote (39966)	10/24/1998 12:39:00 PM
From: Jim McMannis	Read Replies (1) \| Respond to of 1574005

Brian, To carry this roadmap speculation a little further, I expect the K6-3 to be faster than the current Pentium II by about 15%. Intel won't sit on their hands however and I wouldn't be surprised to see a "Dixon" chip packaged for a desktop socket. We've seen what the 128k in die cache does for the Pentium II core (aka Celeron A) so I'd expect the Dixon with 256k in die cache to literally blow away (in X86 terms) a Pentium II and make up any gap between the Pentium II and the K6-3. If Intel doesn't do this then AMD will eat more of their lunch. Jim