Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Test your program

Author: Dann Corbit

Date: 11:54:19 05/04/01

Go up one level in this thread


On May 04, 2001 at 14:48:14, Robert Hyatt wrote:

>On May 04, 2001 at 14:21:
[snip]
>No... they are just interested in seeing Crafty run on one of their SMP boxes
>using their UPC compiler.  If it works well, you will see it from time to time.
>They took a small system to the Linux expo and played chess against all comers.
>They seemed to have a good time doing it, and had planned on giving away
>T-shirts to anyone that could beat it.  After the first day, they had given away
>none, so I think they changed the rule to give away T-shirts to anyone willing
>to play it.  :)
>
>My 60M figure is "peak".  To compare that to DB you have to use 1000M nodes
>per second.  It would _still_ be a long way away.

On the other hand, they had ludicrous horsepower to play with, so they "played
it safe" in not using NULL move, and only using one-half of the cutoff (intead
of a high-low window) and lots of other safety features.  In so doing, they
would not be fooled by many anticomputer tactics and it gave them an enormous
margin of safety.  [I still marvel that someone advised Kasparov to train
against PC programs...].

Anyway, by using riskier algorithms, it might be possible to play on the same
level.  Of course, the Deep Blue team could probably re-architect their solution
if they so desired, since there was a C program that did the coordination of the
search.



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.