Author: George C Dixon
Date: 16:17:37 10/14/97
Go up one level in this thread
I loaded the third game of the four games listed. (The one that ends in 27.Bxf8) I used the Hiarcs6 engine in my Fritz5 program on an AMD K6 200 mHZ machine with a 32 megabyte hashtable . The game left the fritz5 book when Hiarcs6 played 14.. Qb6 so this is where I started the Hiarcs6 engine to see if it could do any better than it did on Covax's Setup. It played identical moves until move #19...Nxa1 It "thought" for about 8-10 minutes (In TOURNAMENT time control of 3 mins per move average) and then showed that it "Saw" the entire rest of the losing line it had played against covax by listing it in the principal variation window. It calculated for about five more minutes and made the move 19...g6 which addressed the king side attack instead of trying to grab the knight. Now I don't know if that saves the game or not but it suggests that given optimum calculating ability, ie. faster chip the engine would play better. If this is indeed true perhaps when the performance of a given chess program is discussed some hardware performance benchmarks similar to the fritzmark should be noted to allow others to assess how much the particular hardware setup used affected the performance of the program. Perhaps one of the computer chess programmers could invent a program we could load and execute to tell us how well our computer would run a given type of chess program. George
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.