Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Bitboard board representation

Author: Christophe Drieu

Date: 23:07:17 01/13/05

Go up one level in this thread


On January 13, 2005 at 17:36:01, Alessandro Scotti wrote:

>On January 13, 2005 at 17:18:46, Eric Oldre wrote:
>
>>so Prof. Hyatt can then use the following
>>
>>board->wPawns;
>>instead of the equiv. in my engine
>>board->bb_pieces[PIECEWPAWN];
>
>Hi Eric,
>I think this particular is not important for speed. In both cases the compiler
>knows the exact address of the variable at compile time, and should generate
>identical code.
>But generally speaking, it's getting increasingly difficult to predict what
>helps and what doesn't. I've had sometimes improved the speed of my program by
>shuffling code around trying to make the cache happy. On a fortunate day I got a
>27% speedup on a piece of the move generator by simply shifting a block of code
>a few lines (on a P4, it only gained 4% on a Athlon)... 8-O

Can you explain this more please ?

Regards



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.