Author: Robert Hyatt
Date: 19:26:02 08/30/98
Go up one level in this thread
On August 29, 1998 at 13:09:49, Dave Gomboc wrote: >On August 28, 1998 at 10:22:43, Moritz Berger wrote: > >>After Dirk reported 80% scores for Junior 5 vs. Genius 5 on 2 P200MMX machines >>in regular games, I'm pleased to top this score with a 90% result vs. Crafty >>15.17 (engine) in a match with alternating colors from the 10 well known start >>positions by GM John Nunn on one P233MMX (internal Fritz autoplayer, no >>permanent brain, 32768 KB hash tables for both engines). >> >>Enjoy the games! > >If Junior 5.0 is really strong, that's great. A match against another program >running on the same machine isn't going to convince me though, whether permanent >brain is on or not. For all I know, Junior could be polling the message queue >just like Chessmaster xx00. > >In fact, I would go so far as to say that in most cases, posting results based >on a one-machine match is simply misleading and should be avoided. > >But please try with two separate machines if possible and tell me what happened. > If Junior again wins by such a margin, I will say "Cool!" :-) > >Dave Gomboc Personally, I think Junior is very strong against computers. It isn't playing any IM/GM players on ICC, so I haven't had a chance to see how it does there. But against computers (Crafty, Ferret, etc.) it is really doing well... But I agree with two programs on one machine...too many questions to ask.. does each get 100% of cpu when running? What about hash sizes? What about paging when a move is made since a program that sits idle generally sees many of its pages kicked out? etc...
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.