Author: David Blackman
Date: 02:23:38 08/17/99
Go up one level in this thread
On August 16, 1999 at 15:50:00, Robert Hyatt wrote: >Another place is in wild positions where 'lazy evaluation' becomes very >effective (ie in most variations lots of material has been won or lost, >which makes lazy eval take almost zero time and can easily double the nps >[at least in crafty]) > >Bob I doubt that Fritz uses lazy evaluation, or that it would help much if it did. To get the kind of speed it does, it must use a very simple eval that is already essentially zero cost. Piece-square tables, possibly augmented by a hashed pawn eval is my guess. You can do that stuff incrementally and very fast. For smarter and slower programs such as Crafty, lazy eval is a big help of course.
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.