Author: Martin Müller
Date: 03:53:44 07/26/01
Go up one level in this thread
On July 25, 2001 at 17:15:30, Uri Blass wrote: >On July 25, 2001 at 16:08:32, Martin Müller wrote: > >>Hi all, >> >>Please check yourselves the behaviour of DJ7 against S5.32. I better do not >>comment this game and leave it up to everyone's own thoughts. Just check the >>jump from 0.51 to 6.44 at depth 27. I used 650 MB RAM for HT on a Dual 1 GHz >>PIII. Amazing. > >I suspect that you use too much RAM and your computer is too fast. >Maybe if you downgrade your hardware you can get better results. > >Here is an analysis of Junior7(p200 64 Mbytes) that cannot see something close >to the 0.51 at the same depth or at smaller depthes. > >New game - Junior 7 >8/2p5/3p1p1p/k2P1p2/P1PP1PBP/2K3P1/8/8 b - - 0 1 > >Analysis by Junior 7: > >58...fxg4 59.Kb3 h5 60.c5 f5 > +- (1.81) Depth: 3 00:00:00 >58...fxg4 59.Kb3 h5 60.f5 Kb6 61.Kb4 > +- (1.85) Depth: 6 00:00:00 1kN >58...fxg4 59.Kb3 h5 60.f5 Kb6 61.Kb4 Ka6 62.c5 > +- (1.94) Depth: 9 00:00:00 6kN >58...fxg4 59.Kb3 h5 60.f5 Ka6 61.Kb4 Kb6 62.c5+ Ka6 63.c6 Kb6 > +- (1.97) Depth: 12 00:00:01 28kN >58...fxg4 59.Kb3 h5 60.f5 Ka6 61.Kb4 Kb6 62.c5+ Ka6 63.a5 Kb7 64.Kb5 Kb8 > +- (2.09) Depth: 15 00:00:01 94kN >58...fxg4 59.Kb3 Ka6 60.c5 Ka7 61.Kc4 Kb7 62.Kb5 h5 63.f5 > +- (2.25) Depth: 18 00:00:07 724kN >58...fxg4 59.Kb3 Ka6 60.c5 Ka7 61.Kc4 Kb8 62.Kb5 Kb7 63.h5 f5 > +- (2.26) Depth: 20 00:00:13 1234kN >58...fxg4 59.Kb3 Ka6 60.c5 h5 61.Ka3 Kb7 62.Kb4 Kc8 63.c6 Kd8 64.Ka5 Ke7 65.Kb5 > +- (2.26) Depth: 22 00:00:29 2921kN >58...fxg4 59.Kb3 Ka6 60.c5 h5 61.Ka3 Kb7 62.Kb4 Ka7 63.Kb5 Kb7 > +- (2.28) Depth: 23 00:00:51 5121kN >58...fxg4 59.Kb3 Ka6 60.c5 Kb7 61.f5 h5 62.Kc4 Ka8 63.Kb4 Kb8 64.c6 > +- (2.31) Depth: 25 00:01:47 10815kN >58...fxg4 59.Kb3 Ka6 60.c5 Kb7 61.Kc3 Ka7 62.Kc4 Kb8 > +- (2.31) Depth: 26 00:03:04 17901kN >58...fxg4 59.Kb3 Ka6 60.c5 Kb7 61.Kc4 Ka6 62.Kb4 h5 63.f5 > +- (2.33) Depth: 27 00:05:41 32046kN, tb=8 > >(blass, tel-aviv 26.07.2001) > > >Uri Hi Uri, But is not exactly DJ7 a program which requires high performance computers and which is designed to calculate a high number of kns together with huge hashtables? I cannot find any such problem on the very same hardware with Deep Fritz. It is not really a malfunction (the move is accordance to the rules) but a move to which lacks sence. When having the so called "sprechstunde" (live talk with Amir Ban and Shay Bushinsky) I mentioned that the program acts strange but did not really get an appropriate reply responding to the unsound sacrifices. I am at a loss to understand all that. Kind regards Martin
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.