Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Perft benchmark

Author: Dieter Buerssner

Date: 12:56:37 12/13/03

Go up one level in this thread


On December 13, 2003 at 03:04:23, Albert Bertilsson wrote:

>Your program actually seems a bit faster than mine. If perft(12) becomes real
>would you consider letting the project use your exe? (It would only require a
>known commandline interface).

Sure I would. But I think, the honor belongs to Sharper! Also, before wasting
lots of CPU time, I should review and test my perft code very carefully. The
hash table stuff for perft is just a bit of an hack done in perhaps an hour
without any good tests. Also, I changed my perft code to avoid many Incheck
calls - so it is not a very simple 10 liner perft function anymore. With more
possibilities for bugs ...

I did not change the perft code to make perft faster, but rather to do something
very similar, that the real search code does.

Regards,
Dieter



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.