Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: A strange Crafty's bench results

Author: Dann Corbit

Date: 11:02:36 07/29/04

Go up one level in this thread


On July 29, 2004 at 06:10:32, Igor Gorelikov wrote:

>A group of testers (AEGT) use Crafty for bench. It happens to be Crafty
>19.11 with the same rc file (by Thomas Mayer).
>Most testers have Athlon and they have got more or less consistent results.
>But what I've got... You may look here:
>
>Celeron 1GHz 256Mb:	167 sec
>Celeron 2.6GHz 256Mb: 	281 sec!!!!!!
>
>I can't believe my eyes.
>
>Then I try Crafty 19.14 (Dann Corbit's compilation). The same story!
>
>Celeron 1GHz 256Mb:	139 sec
>Celeron 2.6GHz 256Mb: 	149 sec!!!!!
>
>
>How can it be?

Probably, a bad instruction mix for your compiler.  Do both chips have the same
first level cache size?


>For comparison, here are Axon bench results (which reflect a more common
>point of view):
>
>Celeron 1GHz, 256Mb,   [1334746 pps.]  PowX= 4.158
>Celeron 2.6GHz, 256Mb, [2126963 pps.]  PowX= 6.626
>
>Igor Gorelikov



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.