Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: The need to unmake move

Author: Anthony Cozzie

Date: 14:13:42 09/04/03

Go up one level in this thread


On September 04, 2003 at 15:08:12, Gian-Carlo Pascutto wrote:

>On September 04, 2003 at 09:56:21, Anthony Cozzie wrote:
>
>>also note that Diep & Sjeng probably have better move ordering than crafty, not
>>to mention searching more slowly; its not suprising that it works for them to
>>split lower in the tree.
>
>Uh, that's a tricky comparison. You don't know how I or Vincent order
>our moves

Crafty is GPL, ergo you should be at least as good or you need to read the
crafty source again :)

 and I fail to see why you 'know' that we are searching slower
>(hint: NPS is not a good indicator).

why not? if it takes 1ms to evaluate a node, then split overhead is relatively
smaller. admittedly I've never written a parallel search, but this seems logical
to me.

>Also, for splitting it's not so much move ordering itself that matters -
>you can use additional heuristics to estimate whether you are at al ALL
>node. That's one of the nice tricks in DTS.
>
>There's much more to it as well. For example the overhead in the splitting
>itself.
>
>--
>GCP





This page took 0.01 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.