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: Gary Ng who wrote (59734)	5/26/1999 11:18:00 PM
From: kash johal	Respond to of 1572120

Gary, Re: "Do you think even Intel seems to be hitting this problem about L2 cache ?" Obviously. Frankly knowing a little about SRAMs its amazes me they can even ship any Xeons. Designing and yielding Full speed SRAMs is non-trivial and very expensive. This is AMDs first attempt so maybe even 1/2 speed cache is difficult. It is much easier to meet the timing requirements with on chip RAM. So Coppermine is definately the way for Intel to go. They reduce costs by probably >>50% and up performance somewhat over an equivalent speed PIII and Xeon. AMD made the decision to go with 128Kb L1 and off chip cache. And that's probably the right decision for the K7 as they would need 4xL1 cache or 512Kb on chip cache to really hit its stride. So at 0.18 micron you will probably see the K7 really hits it stride next year when they incorporate on chip cache. However the 0.25 K7 should be very competitive with the 0.18 Coppermine. Should be interesting how the AMD 0.18 process ramp goes at Fab 25 and of course at dresden. Regards, Kash