Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: EGTB disadvantage for solving probs quick

Author: Vincent Diepeveen

Date: 11:24:09 12/04/00

Go up one level in this thread


On December 04, 2000 at 14:16:21, Ernst A. Heinz wrote:

I don't 6 gigabyte of RAM, RAM very expensive.
If i would have 6 gigabyte of RAM i would use it for transposition
table however for the biggest part of it.

>>Hello here a nightmare scenario when using EGTB,
>>for DIEP that's position #4 from bs2830 testset.
>>
>>Wonder how other progs do with and without egtb on this
>>position. Of course the real problem is that this problem
>>is like a 17 ply problem as there are 5 ply to see then
>>12 checks and it's a stalemate.
>
>This is exactly why I like my combination of interior-node
>recognizers and RAM-based knowledgeable endgame databases
>so much!
>
>See my book on "Scalable Search in Computer Chess" for more
>details (http://supertech.lcs.mit.edu/~heinz/node1.html).
>
>"DarkThought WCCC'99" finds the solution move in 4 seconds
>(see htpp://supertech.lcs.mit.edu/~heinz/dt/) because its
>interior-node recognizers and RAM-based endgame databases
>quickly identify all other moves as far inferior.
>
>=Ernst=



This page took 0.04 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.