Author: Robert Hyatt
Date: 07:05:41 08/29/05
Go up one level in this thread
On August 29, 2005 at 03:39:42, Gian-Carlo Pascutto wrote: >On August 28, 2005 at 12:58:44, Andreas Guettinger wrote: > >>A factor to consider is the speed hit that acessing tablebases causes. When Bob >>was testing crafty on ICC on his 8 way box I saw it dropping from 15 mln NPS to >>between 2-5 mln NPS during the endgame of blitz games. >> >>I think if the speed hit is really 3-4 times this could be a disadvantage in >>endgames where TB dosent help much and searchdepth is more critical. Especially >>if the opponent doesn't use TBs. But maybe I'm wrong. > >Crafty's speed hit is so large because the probing is so aggressive. > >You can probe EGTB with almost no speed hit with some tuning. > >-- >GCP The problem is that you then introduce errors. I only probe (on the quad 875) when a capture takes the board to 5 or less pieces, and I have all 5-piece tables on that box. I don't probe in the q-search at all, although some do and I used to. I can't see how you could reduce that without introducing significant errors. Years ago I used to limit the depth at which I probed, to keep the NPS from dropping too far. And saw too many mistakes where it would push the critical capture beyond the probe horizon, so that it could still think it was winning or whatever. And it would enter a forced line because of that, thinking it was winning, when it was forcing a draw...
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.