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: Process Boy who wrote (66071)	7/19/1999 1:41:00 PM
From: Ali Chen	Read Replies (1) \| Respond to of 1576297

Boy, <and extrapolating that Coppermine will exhibit the same lackluster performance gains. This is a dangerous assumption, IMHO.> Yes, your "assumption" is truly dangerous:) For you. First, the gain from 256k L2 on K6 is TREMENDOUS by any contemporary metrics. The CuLattermine will exhibit MUCH LESS PERFORMANCE GAIN than K62->K63. We saw this effect with Celeon-128, and with Xeon too. Please do not try to discuss matters you have little clue about. In simple terms for you to understand, the speed of L2 does not matter too much. What does matter is the size/associativity of L2, or, more accurately, the amount of L2 cache misses. The main memory (which serves these misses) is the major source of processors delays and overall performance therefore. The major performance gain on Dixons is not because of faster L2 but due to some enhancements to write buffering of victim lines, or something like that (as per some conference publication). Happy process farming.