Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Decrease in Crafty kNPS

Author: Dann Corbit

Date: 14:18:03 10/14/05

Go up one level in this thread


On October 14, 2005 at 13:56:03, Dann Corbit wrote:

>On October 14, 2005 at 13:42:59, F. Huber wrote:
>
>>On October 14, 2005 at 13:28:44, Dann Corbit wrote:
>>
>>>Try it now.
>>>EPD processing is back in, along with blockade detection.
>>
>>Many thanks, Dann!
>>
>>I´ve always prefered _your_ Crafty compilations, because they are working
>>on _all_ CPUs without any problems - and of course they are _the_ standard. :-)
>
>I will make a fast build later on today.

The builds can do EPD processing, but in a monumental tragedy, the new eval
function of crafty 20.1 does not contain the old DETECTDRAW stuff, so problems
like WAC.230 solve very slowly.  It probably makes for better overall play, but
will reduce the aggressiveness of problem solving for a large class of chess
problems.

I have :
Crafty-1921.exe (MS build) and
and
Crafty-1921i.exe (Intel build)

The Intel build is a little faster.

Intel build:

Crafty v20.1 (1 cpus)

White(1): bench
Running benchmark. . .
......
Total nodes: 63434418
Raw nodes per second: 946782
Total elapsed time: 67
SMP time-to-ply measurement: 9.552239
White(1): quit

MS build:
Crafty v20.1 (1 cpus)

White(1): bench
Running benchmark. . .
......
Total nodes: 63434418
Raw nodes per second: 919339
Total elapsed time: 69
SMP time-to-ply measurement: 9.275362
White(1): quit




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.