Author: Uri Blass
Date: 03:54:13 10/05/02
Go up one level in this thread
On October 04, 2002 at 11:33:45, Joshua Lee wrote: >Deep Blue Jr on a Single Chip performed around 300+ elo against the Top Programs >on a P200. Even though we have Hardware over 6 times faster you still won't see >Fritz 7 for example beat the old programs by so huge a margin. I have a different opinion. Chessfun rating list show that Deep Fritz is 200-300 elo better than old programs on the same hardware. If you give Deep Fritz faster hardware you can be sure that the difference is bigger. >Well on a 2ghz Athlon maybe 93% is possible. > I believe the pdf file that was floating around quoted 2 losses out of 30 games >and this with a chip that wasn't fully debugged. There is no evidence for this claim and I ignore events that I have no evidence that they happened. Logic dictates that untill >Something like Fritz can achieve 300+ elo against all those P200 programs on the >same hardware then software is as strong. Or Fritz getting a 3096! rating on the >current SSDF hardware. But therein lies the problem what if the SSDF ratings are >inaccurate? Funny it still doesn't matter as Fritz or any other Top Program will >not get %93 percent against the Top Programs of 1996 or 1997 on a Pentium 200. > >Ofcourse there are other ways to prove a point namely 2 or more wins by Deep >Fritz, I think that would prove it to be on par with Deep Blue. Deep Fritz gets better conditions than deeper blue so I expect Deep Fritz to lose. Uri
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.