Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: More Crafty 19.10 bench marks

Author: Mike Byrne

Date: 09:25:48 02/07/04

Go up one level in this thread


On February 07, 2004 at 12:20:59, Drexel,Michael wrote:

>>I just sent you exe's -- one is based on PGO and one is based on Aaron Gorodn's
>>make file.
>>
>>Both versions are without detectdraw.
>>
>>Best,
>>
>>Michael
>
>The one based on Aaron Gordons make file is the fastest here but PGO is also
>fast.
>
>New Ranking for AMD:
>
>Byrne AG 7.2.2004:
>
>Crafty v19.10 SE - 2004 CCT Champion (1 cpus)
>Krafty Personality == 0
>SE code disabled
>
>White(1): bench
>Running benchmark. . .
>......
>Total nodes: 71227364
>Raw nodes per second: 1003202
>Total elapsed time: 71
>SMP time-to-ply measurement: 9.014085
>White(1):
>
>Byrne PGO 6.2.2004:
>
>Crafty v19.10 SE - 2004 CCT Champion (1 cpus)
>Krafty Personality == 0
>SE code disabled
>
>White(1): bench
>Running benchmark. . .
>......
>Total nodes: 71227364
>Raw nodes per second: 975717
>Total elapsed time: 73
>SMP time-to-ply measurement: 8.767123
>White(1):
>
>Skinner 3.2.2004:
>
>Crafty v19.10
>
>White(1): bench
>Running benchmark. . .
>......
>Total nodes: 71227364
>Raw nodes per second: 962531
>Total elapsed time: 74
>SMP time-to-ply measurement: 8.648649
>White(1):
>
>My test with another AMD computer confirmed this result.
>
>Michael


Your node count is not correct.  You must be using a hash setting that is not
the default.  Can you run the tests without Crafty.rc or setting  - that way I
can it to my log from other uesers.  The correct node count is  89942714.

So far, other users have confirmed the AG and the PGO versions run exactly the
same on their machines with defualt hash settings and using non-P4 processors.



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.