Author: Joshua Lee
Date: 10:54:06 02/24/01
Go up one level in this thread
Yea so much for that theory right? powers of 2 explain how 35MB is better i do realize how yes more memory is avialable for the computer itself so that would explain this being faster than with more memory for hash tables but ammount of Hash Tables when using 184 should make up for that speed difference in depth. I am not sure some but powers to the two maybe they should say and also not odd numbers but i am not sure 2^5 32 524FM 13.4sec 764Kn/s 2^6 64 524FM 13.4sec 756Kn/s 2^7 128 517FM 14.6sec 739Kn/s Someone who has a spare 256 for Hash Tables needs to check this because 2^8 is 256 but herein lies the problem Like in Hiarcs the best result on the FM is with not 32MB for Hash Tables but 35!? 528 13.3 764Kn/s I think Chessbase needs to double check this and fix their product.
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.