Author: Joseph Merolle
Date: 07:31:49 05/31/04
Go up one level in this thread
Mike how do you explain that my cpu in a game time setting of 40 in 2 finds the right move in 48sec flat with clearing hash and loading from scrach, however in infinent analisis it takes my cpu 3 minutes? Go figure. I guess I wont be doing serches in infinent any longer! P.S. from the demo version of shredder 8 classic can we exstract the opening that we need for Pocket Fritz 2. This would save me the money. If so can you give me the instructions again step by step how to exstract and rename file so I can load. However please check if the file is to big last time it was. Thanks! Regards Jospeh On May 31, 2004 at 07:35:00, Mike Byrne wrote: >On May 31, 2004 at 07:03:38, Mike Byrne wrote: > >>On May 31, 2004 at 00:55:28, Joseph Merolle wrote: >> >>>Mike what is your cpu speed and ram? 2.03 min is really fast on my lap top 3.06 >>>ghz 256 ram it finds in about 3 min. >>> >> >>I have a dual 1.7 Ghz with 512M ram, For the test below , Shredder was set to 2 >>threads with about 48M hash. The difference is probably the dual CPU machine. >>It somtimes can be more than twice as fast and for other positions can be less >>than twice as fast. I will run it with one thread and see what the result is. >> >> >>> >>> >>>On May 31, 2004 at 00:04:01, Mike Byrne wrote: >>> >>>>[d]Kasparov,G - Comp Deep Blue, Philadelphia m 1996 >>>>2rr2k1/pp1qnppp/2n1p3/3p4/1bPP3P/1P2RNP1/PB3P2/1BRQ2K1 w - - 0 1 >>>> >>>>Analysis by Shredder 8: >>>> >>>>20.Ng5 Ng6 21.Qh5 Bc5 22.dxc5 >>>> ± (1.13) Depth: 1/10 00:00:00 >>>>20.a3 Bxa3 21.Bxa3 dxc4 >>>> ± (0.85) Depth: 5/16 00:00:00 >>>>20.a3 Bxa3 21.Bxa3 dxc4 >>>> ± (0.85) Depth: 5/16 00:00:00 >>>>20.a3 Bxa3 21.Bxa3 dxc4 >>>> ± (0.85) Depth: 5/16 00:00:00 >>>>20.a3 Bxa3 21.Bxa3 dxc4 >>>> ± (0.85) Depth: 5/16 00:00:00 >>>>20.a3 Bxa3 21.Bxa3 dxc4 >>>> ± (0.85) Depth: 5/16 00:00:00 >>>>20.a3 Bxa3 21.Bxa3 dxc4 >>>> ± (0.85) Depth: 5/16 00:00:00 >>>>20.a3 Bxa3 21.Bxa3 dxc4 >>>> ± (0.85) Depth: 5/16 00:00:00 >>>>20.a3 Bxa3 21.Bxa3 dxc4 >>>> ± (0.85) Depth: 5/16 00:00:00 >>>>20.a3 Bxa3 21.Bxa3 dxc4 >>>> ± (0.85) Depth: 5/16 00:00:00 >>>>20.a3 Bd6 21.Ng5 Nf5 22.Bxf5 exf5 >>>> ± (0.94) Depth: 5/18 00:00:00 >>>>20.a3 Bd6 21.Ng5 h6 22.Bh7+ Kf8 23.Nf3 >>>> ± (0.94) Depth: 6/14 00:00:00 >>>>20.a3 Bd6 21.Ng5 h6 22.Nf3 Re8 23.Qc2 >>>> ± (0.75) Depth: 7/20 00:00:00 53kN >>>>20.a3 Bd6 21.Qd2 dxc4 22.bxc4 Bc7 23.h5 Nf5 24.Bxf5 exf5 >>>> ± (0.73) Depth: 8/20 00:00:00 150kN >>>>20.a3 Bd6 21.Qd2 Bc7 22.Qe2 dxc4 23.bxc4 Bb6 24.Ng5 Nxd4 25.Bxh7+ Kf8 >>>> ± (0.71) Depth: 9/22 00:00:01 265kN >>>>20.a3 Bd6 21.Qd2 Bc7 22.Re2 dxc4 23.bxc4 Bb6 24.d5 exd5 25.Qg5 >>>> ± (0.71) Depth: 10/25 00:00:02 632kN >>>>20.a3 Bd6 21.Qd3 g6 22.Qc3 Nf5 23.Bxf5 exf5 24.Qb4 >>>> ² (0.67) Depth: 11/30 00:00:04 1439kN >>>>20.a3 Ba5 21.Ng5 h6 22.Nf3 dxc4 23.bxc4 Bb6 24.c5 Ba5 25.Qb3 Qc7 26.Be4 >>>> ² (0.56) Depth: 12/30 00:00:10 3631kN >>>>20.a3 Ba5 21.g4 Ng6 22.h5 Nf4 23.Qc2 g6 24.Qf5 exf5 25.Re8+ Rxe8 >>>> ² (0.54) Depth: 13/34 00:00:24 8611kN >>>>20.a3 Ba5 21.Ng5 h6 22.Nf3 dxc4 23.bxc4 Bb6 24.c5 Ba5 25.Qb3 Nd5 26.Re2 Qc7 >>>>27.Qd3 >>>> ² (0.57) Depth: 14/31 00:00:47 16059kN >>>>20.a3 Ba5 21.Ng5 h6 22.Nf3 dxc4 23.bxc4 Bb6 24.c5 Ba5 25.Qb3 Nd5 26.Re2 Qc7 >>>>27.Bd3 >>>> ² (0.64) Depth: 15/40 00:01:25 28708kN >>>>20.Bxh7+ Kxh7 21.Ng5+ Kg8 22.Qh5 Qe8 23.Qh7+ Kf8 24.Qh8+ Ng8 25.Nxe6+ Qxe6 >>>>26.Rxe6 Nxd4 27.Bxd4 dxc4 >>>> ² (0.65) Depth: 15/45 00:02:01 42241kN >>>>20.Bxh7+ Kxh7 21.Ng5+ Kg8 22.Qh5 Qe8 23.Qh7+ Kf8 24.Qh8+ Ng8 25.Nxe6+ Qxe6 >>>>26.Rxe6 Nxd4 27.Bxd4 dxc4 28.bxc4 >>>> ± (1.00) Depth: 15/45 00:02:32 53632kN >>>>20.Bxh7+ Kxh7 21.Ng5+ Kg8 22.Qh5 Qe8 23.Qh7+ Kf8 24.Qh8+ Ng8 25.Nxe6+ fxe6 >>>>26.Rf3+ Qf7 27.Rxf7+ Kxf7 28.Qh5+ Kf8 29.a3 Bd6 30.Qg4 Re8 31.cxd5 Nf6 32.dxe6 >>>>Nxg4 33.Rc5 >>>> ± (1.15) Depth: 15/48 00:03:38 78706kN >>>> >>>>(Byrne, MyTown 31.05.2004) > >When I re-run test with cleared hash and position learning off. I get different >solution times depending on the hash but all of them seem to get faster. I have >noticed this before with Shredder , you only seem to get one crack at running a >position before it learns the best move even with position learning off. All of >my subsqequent re-run of the above positions are faster than above - varing the >hash size does vary the solution time.
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.