Author: Chris Whittington
Date: 03:32:46 10/01/97
CSTal always used to have one depth-based hash table with five bucket attempts on clashes. After reading the threads in rgcc, I tried it out with two hash table, an always replace and a depth-priority table. Some test positions get faster, some get slower, but, on the whole, there's an improvement. Until, that is, the hash gets overloaded, when the old depth-only table starts to come out on top. Is this a generally observed phenomenon ? Chris
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.