Author: Robert Hyatt
Date: 07:45:07 02/16/04
Go up one level in this thread
On February 16, 2004 at 10:34:54, Peter Skinner wrote: >On February 16, 2004 at 10:27:24, Ingo Bauer wrote: > >>Hi >> >>>If you didn't change hashing sizes, something is broken in the compiler >>>options... apparently a bad optimization somewhere. >>> >>>node counts can not vary... >> >>Something is broken! I have post one below Bryan's and my nummbers differ in >>node counts as well. It seems the MSVC6_non-SMP is broken. >> >>Ingo > >MSVC 6.0 /w sp5 /wo /dfast > >EPD Kit revision date: 1996.04.21 >unable to open book file [./book.bin]. >book is disabled >unable to open book file [./books.bin]. > >Crafty v19.10 > >White(1): bench >Running benchmark. . . >...... >Total nodes: 89942714 >Raw nodes per second: 708210 >Total elapsed time: 127 >SMP time-to-ply measurement: 5.039370 >White(1): > >MSVC 6.0 /w sp5 /w /dfast > >EPD Kit revision date: 1996.04.21 >unable to open book file [./book.bin]. >book is disabled >unable to open book file [./books.bin]. > >Crafty v19.10 > >White(1): bench >Running benchmark. . . >...... >Total nodes: 89942714 >Raw nodes per second: 686585 >Total elapsed time: 131 >SMP time-to-ply measurement: 4.885496 >White(1): > >I get exactly the same node code on two different "versions" of the non-smp >build of the MSVC compile. > >In fact I get the exact same node count on all of the versions. > >What I do notice is if the /DFUTILITY option is not set the nodes are different. >I compile all of mine with it _enabled_. > >Peter. FUTILITY certainly makes a difference. In fact, other -D options can make a difference. But if all the -D stuff is the same, except for enabling SMP, and the node counts vary, it's a clue something bad is happening.
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.