Author: Robert Hyatt
Date: 07:22:29 08/14/01
Go up one level in this thread
On August 14, 2001 at 04:47:43, Uri Blass wrote: >On August 13, 2001 at 23:55:33, Robert Hyatt wrote: > >>On August 13, 2001 at 21:42:20, Manuel Rodriguez Blanco wrote: >> >>>Hi, i had read a lot messages about K-F match and had see some wrong statement. >>>i want from my humble point of viw give some clarify. >>> >>>1) Krammnik HAVE access to fritz 7 in a 8 1GH procesors machine. Do you know >>>how?? Of course, you can only put fritz 7 in 1 procesors machine and simpli >>>give more time to fritz 7 until its hit the sames nps it hit in 8-BOX, you can >>>ask how much time? there are siples express that can level 2 computers with >>>diferent procesors power. its not? >> >> >>This is correct. Probably giving a single-cpu version 5X longer to think would >>be a very good approximation of the 8-way box. > >I doubt if it is true >It is possible that the program says use evaluation A >if you have one processor and use evaluation B >if you have 8 processors. It is _possible_ to write a program that uses a different evaluation starting on a specific date. Or one which turns off most search extensions until after a specific date. Or one that behaves in some different way before and after a specific date. It is possible. But so unlikely as to be something that really would not be done. The opportunity for bugs is simply enormous. > >I hope that kramnik will be surprised in the match and that he is not >going to get the exact weaknesses that he expects from Fritz. > >What is he going to do in this case. >Is he going to complain about cheating? > >I think that the public is not going to buy the excuse >that the program did not play the moves that kramnik expected. > >Uri I don't believe it will happen, but we will see. A chess engine is already far too complicated. Adding SMP adds bugs that take years to eliminate. Going this far is just asking for trouble.
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.