Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Shredder Triple Brain ???

Author: Manfred Meiler

Date: 08:33:41 05/25/03

Go up one level in this thread


On May 25, 2003 at 10:55:55, Michael P. Nance Sr. wrote:

>Let Me ask a question? Wouldn't using the resources for 3 Engine tie up the
>Processors capabilites? Wouldn't You think that Shredder would be able to come
>up with a better move by It's Self ,using the full potenial of the Processor ,
>rather than a 3 Engine sharing resources? I says this because I have tied the
>triple Engine function and They didn't impress Me. Let Me hear from You.>>>>Mike


In Triple Brain there are only 2 engines combined - plus a "decision modul"
which has to decide which of the two lines is the best by examining the output
of both engines to find out which has a better understanding of the current
position.
Of course these 2 engines and the "decision instance" consume more processor
capacity than 1 engine alone - similar to 2 or more engines in parallel
analysing a position (or multe variants mode of an engine).
Important are the results of this Triple Brain feature - compared to the results
of the engines "stand alone"; maybe you wanna have a look at my mentioned ECXEL
sheet in test suite "WM-Test".

IMO it's very important for the efficiency of Triple Brain which engines you're
using as parts of the Triple Brain: the 2 engines should be rather equal strong
and with different playing styles (e.g. one strong tactical and one more
positional engine).
I hope that Shredder author Stefan Meyer-Kahlen will further develop the Triple
Brain feature.

Manfred



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.