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: Paul Engel who wrote (42144)	11/24/1998 1:58:00 AM
From: Scumbria	Read Replies (2) \| Respond to of 1571864

Paul, Available benchmarks - try Tom Uberclockermeister - show that the Mendocino - a 128 K full CPU speed L2 Cache Celeron - performs almost as good as a Pentium II with 512K half-CPU speed L2 cache, when the Mendocino and Pentium II are clocked at the same speeds. Conclusion - SMALL FULL SPEED caches can provide enough performance as a half speed cache FOUR TIMES the size ! The Mendocino cache is located on chip, and has about 1/2 the latency of the PII cache- which is located on a separate die. Do you remember when I was querying you for details about Mendocino cache latency, early in the year? Now you can see why I was so interested. To paraphrase James Carwell: "It's the latency ......" Had you attended Haavaad instead of MIT, you would have said "almost as well" instead of "almost as good". Scumbria

To: Paul Engel who wrote (42144)	11/24/1998 9:33:00 AM
From: Ali Chen	Read Replies (1) \| Respond to of 1571864

Good verbal "engineering", Pal:) Let's try this one: Available benchmarks - try Tom Uberclockermeister - show that the Xeon - a 512 K FULL CPU speed L2 Cache - performs almost the SAME as a Pentium II with 512K half-CPU speed L2 cache, when the Xeon and Pentium II are clocked at the same speeds. Conclusion - FULL SPEED caches can provide almost NO BENEFITS in performance as a HALF SPEED cache of the same size ! ...but costs much more :)