Author: Gerd Isenberg
Date: 12:15:57 03/03/06
Go up one level in this thread
>> >>From a clock cycle performance point of view I would agree, locking is to >>expensive. From a search performance pov, I would rather use a "global" table. >> >>Tony > > >You could be right. But there is "local" data in the history values. I'll try >a global table to see what impact it has, as it certainly simplifies things a >bit size-wise... My guess is that thread local tables are better, if you don't index with saturated depth left or ply. Local tables may avoid saturation, if cleared after starting a thread with a new position. Otoh History table is erroneous anyway. You may reduce a "reasonable" move in "this" position, because it was often tried earlier, while a later move failed high. A futile move in "this" position, like putting a piece en prise, may not become reduced. Because due to move sorting it was seldom an early move in the "local" tree, but had some first cut-hits in distant subtrees (of other threads). May be the from-ambiguity using <piece><to> has some positive impact as well - and is even favorable over unique quite move indices implying <piece><from><to>. And it avoids possible "small" <from><to>-errors, if several pieces may have same from-to moves.
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.