Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Dann's Crafty vs Standard Crafty - A bench test.

Author: Robert Hyatt

Date: 20:54:07 06/19/01

Go up one level in this thread


On June 19, 2001 at 08:46:50, Gian-Carlo Pascutto wrote:

>On June 19, 2001 at 08:41:10, Dieter Buerssner wrote:
>
>>On June 19, 2001 at 05:43:00, Gian-Carlo Pascutto wrote:
>>
>>>On June 18, 2001 at 21:07:19, Slater Wold wrote:
>>>
>>>>These tests were run on a Dual Pentium III 1,000mhz, 1028MB RAM.
>>                            ^^^^
>>[...]
>>>BTW. Your comparisation is horribly broken. If you bench you should
>>>get the EXACT SAME nodecount for all versions.
>>
>>I think, this is not true for SMP programs.
>
>Yikes I missed that.
>
>This doesn't change the fact that the comparisation doesn't work
>if they are searching different trees.
>
>--
>GCP


The way to run the benchmark here is to use _one_ cpu.  Then, if the 2cpu number
is wanted, that can be calculated by running the two cpu test and dividing the
two times.  But don't use SMP times to compare two different programs, or two
different compilers, because the SMP stuff itself adds enough variability to
make things look odd...



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.