Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: 80 KN/s in a celeron 466 (UNIX) and 4 KN/s in a k6-2 450 (WIndows)

Author: Dann Corbit

Date: 15:33:21 06/18/02

Go up one level in this thread


On June 18, 2002 at 18:30:21, Adriano Bedeschi de Souza wrote:

>On June 18, 2002 at 18:19:21, Dann Corbit wrote:
>
>>On June 18, 2002 at 18:13:56, Dann Corbit wrote:
>>
>>>On June 18, 2002 at 17:48:46, Adriano Bedeschi de Souza wrote:
>>>
>>>>Using mingw for windows
>>>>with -O3
>>>>and other minor optimizations
>>>>what can it be?
>>>
>>>Run another profile and show us the output.
>>>
>>>Are you doing incremental hash updates?
>>
>>Your subject line makes me suspect something else also.
>>
>>What functions are you using for timing on Win32?
>>
>>A factor of 20 can easily be explained by a savagely broken timer.  If anything,
>>Win32 binaries tend to be faster, not slower.
>>
>>If you are not nervous about someone looking at your code, I will take a look at
>>it and run it through the Intel profiler.
>
>My code is nothing =)
>how can i send it to u? email?

Email is fine...
dcorbit@connx.com

I won't promise any miracles.

>about the hash, i disabled it an still getting 4,5 KN/s =) incredible

That the speed went up is not surprising.  If you get more depth with hash off,
then your hashing is severely broken.



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.