Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: AMD athlon 64 and Crafty (the 3400+ or the 3500+) and request

Author: Michel Langeveld

Date: 04:02:16 10/27/04

Go up one level in this thread


On October 26, 2004 at 21:00:11, Mike Byrne wrote:

>On October 26, 2004 at 18:54:42, Michel Langeveld wrote:
>
>>Incase someone else is also interested hereby two benchmarks done with the AMD
>>Athlon 64 3400+ and the AMD Athlon 64 3500+. I found the Athlon 64 3400+ on the
>>internet and let the Athlon 64 3500+ do by someone who was very helpfull on ICC.
>>
>>The 3400+ has 2.4Ghz - 1 memory controller  - L1 128KB - L2 512KB - socket 754
>>The 3500+ has 2.2Ghz - 1 memory controllers - L1 128KB - L2 512KB - socket 939
>>
>>I was not sure which was the fastest ....
>>For chess the result is that the +3400 is around 4.9% faster as the +3500.
>>
>>AMD Athlon 3400+ 2.4Ghz - no crafty rc
>>======================================
>>Crafty v19.17
>>Total nodes: 90925937
>>Raw nodes per second: 1567688
>>Total elapsed time: 58
>>
>>AMD Athlon 3500+ 2.2Ghz - no crafty rc
>>======================================
>>Crafty 19.17
>>Total nodes: 90925937
>>Raw nodes per second: 1490589
>>Total elapsed time: 61
>>SMP time-to-ply measurement: 10.491803
>>
>>Request
>>=======
>>Does someone else has an Athlon 64 / Opteron? I am very interested in the
>>benchmarks results of crafty without hash. The we can compare what is the
>>influence of the socket / cache / etc.
>>
>>You can find an executable here:
>>
>>http://wbec-ridderkerk.nl/html/download/crafty/crafty_1917_DC.zip
>
>Question -- is the exe without futility enabled here?  Otherwise , total nodes
>searched is not correct I believe for 19.17 with futility enabled.

I think the executable is a default compile by Dann Corbit



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.