Author: Robert Hyatt
Date: 11:37:33 01/13/98
Go up one level in this thread
On January 13, 1998 at 13:24:26, Moritz Berger wrote: >On January 12, 1998 at 15:04:08, Robert Hyatt wrote: > >>I have just finished cleaning everything up, relative to the null-move >>mate extension and hash table refinement discussed here. As a sanity >>check, I decided to once again run win at chess at one minute per move >>to see how Crafty would do on my Pentium Pro 200. The best for an >>"official" version has been 297, with it typically staying around 295 >>or so. > ><snip> > >>Any other results to compare with? Be interesting to see who solves >>what quickly and who has trouble with what. I'm going to experiment >>with the chess middlegame suite Bruce mentioned. The only thing I don't >>like is the length of time it takes to run a suite that large... > >WAC results for Fritz 5 on P233MMX (98304 KB HT): > >< 5s: 292 positions correct > >< 10s: 293 positions correct >#250: 6s > >< 60s: 296 positions correct >#71: 13s >#235: 12s >#265: 31s > >< 90s: 297 positions correct >#2 - 88s > >Hard problems: >#100 - > 240s >#163 - > 240s >#230 - > 240s Interesting. 230 has been the hardest of the bunch for me. 163 is almost solvable by Crafty and only needs a couple of minutes more. 100 might be aided by some endgame specific code I have as Crafty gets this early and doesn't change. I think 100 is one of the ones that Ferret has problems with too... Bruce and I were comparing notes the other night...
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.