Author: James Swafford
Date: 19:23:11 08/24/05
I'm sure most parallel searchers use algorithms that allow splitting at arbitrary nodes in the tree. I'm curious how good (or bad!) it would be to simply spawn threads at the root to search an entire move. In other words, if I have 4 processors, then I would have 4 search threads running, each searching a different root move. Yes, I know that's incredibly simplistic. I'm just curious as to how the naive approach compares to something more complex. Thanks, -- James
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.