Author: Joseph Ciarrochi
Date: 13:31:22 02/05/06
I thought i'd post these results now, because it sets the stage for a question i have about the interaction between hardware and engine pentium 447 ghz. 32 mb hash. noomen test set 40/4 repeated time control rybka beta 12 (default) 292.5/437 (44.5-42.5 against fritz9) fritz9 274.5/437 fruit 2.2.1 235/437 toga II.1 216.5/ 437 spike 1.0a 151/437 ruffian 2.1 141.5/437 Rybka beta 12 is not doing better than what i would espect beta 1 to do (IMO). Also, fruit 2.2.1 seems to be doing worse than expected (based on cegt blitz ratings). Maybe this will change when the tournement finishes but..... Is it possible for engines to improve their relative strength on faster hardware (excluding 64 bit and multiple processor comparisons). this seems to happen for fruit 2.2.1. Perhaps Rybka also needs a fast processor to show its full advantage? To put what i am asking more concretely, if i had run this same tournment on really fast harderware, would rybka beta 12 and fruit 2.2. have won more games (meaning the other engines lost more games)? All speculations welcome best joseph
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.