Author: Robert Hyatt
Date: 19:52:40 03/19/03
Go up one level in this thread
On March 19, 2003 at 22:42:09, Tom Kerrigan wrote: >On March 19, 2003 at 13:55:12, Robert Hyatt wrote: > >>a couple of journal papers behind it giving the details. 66.2 seems wrong, as I >>mentioned, >>because the Cray T90 can't break 100ns. And they are _known_ for memory speed >>bandwidth and no cache. > >Uhm, "memory speed bandwidth"? Aren't we talking about memory latency? > >-Tom Yes. I meant "memory speed and latency". Memory speed and bandwidth would be the same thing. However, most (except for Aaron's drastic overclocking) are reporting numbers that are right in line with my 100-120ns figure, although the high-performance boxes I have are actually slower than that. My quad xeon 700 clocks in at 130ns latency and a box of dual xeons from 2.4 to 3.06 ghz are clocking in at right around 150ns.
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.