Author: Helmut Conrady
Date: 01:33:14 09/23/01
Go up one level in this thread
On September 22, 2001 at 22:17:06, Uri Blass wrote: >On September 22, 2001 at 19:08:06, Torstein Hall wrote: > >>On September 22, 2001 at 18:29:46, Andreas De Troy wrote: >> >>>When I see results for the so called "Fritzmarks", I notice that the actual >>>numbers decrease with increasing hashtables. Is this an artefact of the >>>measurement? In other words, are larger hashtables always better? I suppose it >>>depends of the speed of the processor. If you have, for instance an Athlon 1 Ghz >>>(or a 1.5 Ghz or...), does it -in general- make sense to increase the size of >>>the hashtables to 256 Mb, 512 Mb or more? >>> >>>Thanks in advance for any help! >> >>It also depends on your time settings. For long time analysis I think very big >>Hashtables are always a bonus. >> >>Torstein > >I am not sure about it. >There is a danger of hash collision and this problem can become more important >if the hash tables are bigger and it is possible that part of the programmers >did not care about very big hash tables+ very long time control+future hardware. > >The problem is that programs can assume that 2 different positions are the same >because their 64 bit number is the same when it is not truth(I also believe that >there are programs that use less than 64 bits and in this case they may have >problems earlier). > >If the hash tables are very big then the probability for hash collision can >increase and if there are enough hash collisions the result can be a bad move. > >Uri BTW: Stefan Meyer-Kahlen said, he had never seen a hash collision during his many years work in programming. Helmut
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.