Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Quad proc results

Author: Mridul Muralidharan

Date: 06:54:05 05/01/04

Go up one level in this thread


On May 01, 2004 at 06:36:21, Tom Kerrigan wrote:

>550MHz Pentium IIIs.
>
>-Tom

Interesting - I think I used an almost identical dell box :)
My observations were quiet similar with messchess , though my speedup in nps was
slightly higher (and my single proc nps must be much lower than yours I think).
More number of processors added more fail highs from one processor affecting
another's partial search and so my number of nodes required to be searched for
completing a search depth also went up marginally.

Not limiting the number of splits per position helped me increase the nps
speedup.

I will look up for the logs in office on monday.

Mridul



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.