Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Killer and history

Author: Ernst A. Heinz

Date: 02:22:29 06/26/98

Go up one level in this thread


On June 25, 1998 at 17:46:56, John Stanback wrote:
>
>As for history heuristic, I'm still using a variant of this idea
>that I came up with to save a little memory under DOS.  I have a
>table like this:  Bonus[side][piece][square] which gets incremented
>when a cutoff occurs.  For the best move, the piece and destination
>square form the indexes to the array.  As I recall this was
>slightly worse (< 1 percent?) than the standard 64 x 64 history
>table.
>
>John

"DarkThought" also does this so-called "piece-to-square" history because
it is extremely cache-friendly on modern CPUs.

As a literature reference see:

Uiterwijk, J.W.H.M. (1992).
Memory efficiency in some heuristics.
ICCA Journal, Vol. 15, No. 2, pp. 67-74.

=Ernst=



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.