Author: Stuart Cracraft
Date: 17:05:45 01/22/98
Go up one level in this thread
On January 21, 1998 at 21:49:52, Robert Hyatt wrote: >If you read Beal's paper, which produced results very similar to those >we found in Cray Blitz years ago, the idea of probing multiple entries >doesn't have much effect until the hash table becomes over-subscribed. > >Don found significant improvement when searching 10X the nodes that the >hash table would hold. But for less than 3x or so, there was no real >benefit. This was in a recent ICCA journal. > >You might not be testing long enough for the hash size you are using... What I could do to test the success of rehashing is experiment with a smaller hash table (in order to keep the test times the same as now.) This should have the same effect. Thanks for the tip.
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.