Author: Aloisio Ponti Lopes
Date: 18:42:17 01/03/03
Go up one level in this thread
On January 03, 2003 at 18:22:07, Stephen Ham wrote: >Dear Aloisio, > >Thanks for the helpful reply. > >What is your experience with the "triple brain" feature, Aloisio? Does it indeed >make the resulting moves stronger? I imagine that it must divide the total >allocated RAM by 1/3 each, so that means that there won't be many hashtables >available for long searches. Does that then mean it's only effective for faster >time controls? After I bought WM-Paket 2001 (actually an old friend of mine bought it in Munich for me), I asked Stefan Meyer-Kahlen by e-mail about what to expect from Triple Brain's analysis, as I was thinking of using it to analyse my own games. He told me that if I had a Dual machine it would be better, because each engine could use 1 processor. He did not mention any memory or hash tables issues (I think that if someone is using a Dual he's got plenty of memory too...). I used to run Triple Brain at my local FICS chess server, jogo.cex.org.br. Scores were not bad, but unfortunately I gave up because of auto-232 and time management (could not play games with fast time limits, because Triple Brain would loose on time - I had to set a high increment, i.e., 7 seconds or more, but no one wants to play an internet game with these increments! ). Most used set-up that time was Shredder 5 + WbNimzo2000b: Blitz: 2213 (RD 242.7) (win 2) (loose 1) (draw 0) total= 3 games Standard: 2334 (RD 164.7) (win 12) (loose 0) (draw 2) total= 14 games I think that now with a Dual machine, Shredder 7 + Ruffian would be a very tough opponent, though I don't have any games to prove what I'm saying... :-( A. Ponti
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.