Author: Vincent Diepeveen
Date: 03:37:50 09/17/01
Go up one level in this thread
On September 17, 2001 at 04:42:15, Gian-Carlo Pascutto wrote: >On September 17, 2001 at 03:59:11, Vincent Diepeveen wrote: > >>This is very easy to see. The statement that bsf/bsr are fast instructions >>is irrelevant because either of the conditional jumps receives 30 times the >>speed penalty of those instructions. > >The trick is that in crafty the bitboards are usually empty or >near-empty, and the number of mispredictions in minimal. you've at least 50% chance of a misprediction because chess acts as a worst case game for predictors: first a bit you get out of a bitboard for white and obviously chance is bigger that move is in a part X of bitboard. then after that black is to move and you get bit out of other part of bitboard. note that sliding pieces like Q,R,B already give a misprediction in advance as they always cover both sides of the board so getting bits out of bitboard there is a disaster. >-- >GCP
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.