Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Possible new idea for EGTB generation !! (Eugene-Bob-Dann) for openers

Author: Angrim

Date: 18:35:18 02/09/03

Go up one level in this thread


On February 08, 2003 at 23:17:53, Les Fernandez wrote:

>As we all know, generating egtb's for the 6 piece sets require alot of time and
>computer hardware to do.  I have an idea that appears to reduce the amount of
>work needed to generate these permutations and would like to hear follow-up on
>your thoughts.  What I propose to do is explained in the following example.
<snip>
>All of these appear to me to be legally sound and will provide us with 5 new
>solutions for the 7 piece EGTB for every 6 piece EGTB position that is
>generated.  If we take it one step further we can also say that we can permutate
>each position 4 ways based on symmetry which then would provide us with 20 new
>positions for the n+1 EGTB set as we generate the positions for the 6 piece
>EGTB.
>
>Please give some serious thought to this idea since it may allow us to generate
>EGTB's much more quickly.  I will look forward to hearing your thoughts.
>
>Thanks,
>
>Les
so while generating the 6 piece tables, you want to store 20 extra 7pc
positions for each 6pc position generated? we hardly have room to store
the 6pc tables as it is, and you want to use (at least) 21x as much room?
Also, random access to disk takes a significant amount of time, so those
extra 20 writes to disk would make the 6pc build much slower.  And
all this effort just to maybe gain a few percent on the time used to
build the 7piece tables when we are finally ready to do that?

Guess those weren't the thoughts you were looking forward to.

Better luck with your next idea.
Angrim



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.