Author: Robert Hyatt
Date: 05:16:22 01/15/98
Go up one level in this thread
On January 15, 1998 at 01:35:03, Don Dailey wrote: >Hi Bob, > >Do you consider fractional plys in your hash table implementation? >Is a depth of 3 3/4 better than 3 1/2? > >- Don > > yes. I simply store the usual "draft", which is "depth". In my case, in units of 1/60th of a ply. I tried "truncation" as well, but it made no real difference, and not truncating is simpler. (truncation was simply depth/60 before storing it. And I don't store q-search nodes at all, nor do I do hash probes in the q-search. It was almost a wash when I did, but the few extra nodes I search by not hashing is way more than offset by the speed gain by having a simpler q-search that is more likely to fit into L2 cache. And of course, there is the issue that I can run with a smaller hash table without worrying about too many things getting lost. Helps at longer time controls even more on machines without a gig of memory...
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.