Author: Mark Young
Date: 14:47:45 01/02/99
Go up one level in this thread
On January 02, 1999 at 17:23:46, Robert Hyatt wrote: >On January 02, 1999 at 14:15:47, Mark Young wrote: > >>On January 02, 1999 at 11:33:17, blass uri wrote: >> >>> >>>On January 02, 1999 at 11:13:00, Dan Kiski wrote: >>> >>>>On January 02, 1999 at 03:57:37, blass uri wrote: >>>> >>>>>Fritz5 could win game 113 by 50.Ndc6 >>>>>My Fritz5.16 finds this move in 59 seconds on pentium200MMX >>>>> >>>>>In the game Fritz5 had 81 seconds to use its permanent brain and to find the >>>>>move on pentium400 but it could not find the move. >>>>> >>>>>Fritz5 used the permanent brain in this game(otherwise it would not play in 0 >>>>>seconds) >>>>> >>>>>I know that Fritz5 is optimized for pentium200MMX but I think it does not say >>>>>that Fritz5 is slower on pentium400 but only that it does not earn speed from >>>>>pentium400 like other programs. >>>>> >>>>>The reason for this strange behavior of Fritz5 can be one of the following: >>>>> >>>>>1)Fritz5 was slowed down by another application or a virus >>>>> >>>>>I think that programs should tell the user if something like this happens. >>>>>They can feel it if there is something strange in the number of nodes per second >>>>> >>>>>2)I have a different version of Fritz5(16 bit) >>>>> >>>>>3)Fritz5 got tired during the game because of a bug in Fritz5(It may be the >>>>>case) >>>>> >>>>>I found that there were cases when Fritz5 was slowed down for some moves during >>>>>a game against another engine. >>>>>In most of the games it does not happen but it can happen. >>>>> >>>>>4)I used different hash tables(I do not think that this is the case because I >>>>>tried different hash tables and it did not change the result of finding Ndc6 in >>>>>1 minute) >>>>> >>>>>Uri >>>> >>>>Please post the fen or state which game this was. >>>1n1N4/r1p5/5k2/3nN3/pp1P1P2/P5P1/3R1K1P/8 w - - 0 1 >>>The last move of black was b5-b4 >>>Fritz5 replied 50.Ra2 instead of Ndc6 >>> >>>The game can be found in Enrique's games (this is the 3th game between Fritz5 >>>and tiger 11.7.5) >> >>I ran the position on my P II 400 256MB ram. I set the hash at 120 mb. >> >>5.03, Found Ndc6 45 sec. >> >>5.16 in Junior 5, Found Ndc6 in 30 sec. >> >>I let the position run 3 plus mins, it stayed with Ndc6. >> >>Did the move have a time stamp and ply number, that would help clear things up. >> >>> >>>Uri > > >This appears to be a gross tactical blunder that might suggest something >is wrong with the hardware or the version of fritz being used. This may be it, I just ran the same version of Fritz 5 and ran it on the same hardware, P II 400 256mb ram. It finds Ra2 at ply 10, 15 of 29 moves searched but at ply 10 and 18 of 29 moves searched it finds Ndc6. The times match up, I found Ra2 at 33 sec. but I also find Ndc6 at about 45 sec. Which his computer did not find. Looks like a bug of some kind. On my notebook >crafty says this is .55 after a second or so. While Nc6 is almost +3.0...
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.