Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: TSCP nps test (and question for Dann)

Author: James Swafford

Date: 14:50:38 07/24/99

Go up one level in this thread


On July 24, 1999 at 15:04:27, Peter Kappler wrote:

>On July 24, 1999 at 12:20:19, James Swafford wrote:
>
>>Using Dann's executable on my K6-2/400: 57,019 nps.
>>
>>Dann : what compiler did you use, and with what options?
>>That's over 50% faster than CygWin with the -O flag.
>>
>
>Now I'm very confused.
>
>Your K6-2/400 came in almost 10% faster than my K6-3/450.
>
>Even at equal clock speeds, the K6-3 should signifiantly outperform the K6-2,
>since its L2 cache runs at core speed, instead of 100MHz.
>
>It's true that I'm running Win98 and you're running NT, but I can't believe that
>could make such a big difference.
>
>Any ideas?
>
>--Peter
>

I just ran it again to be sure the first run wasn't a fluke,
and it did slightly better with 57,163.385827 nps, time=3.556 sec.

I'm a bit confused, too.  Below are some of my benchmark results from
WinTune98:  I'd be interested to see how your K6-3 compares.


CPU Integer: 895.7944 MIPS
CPU Floating Point: 483.7748 MFLOPS
Memory: 333.4171 MB/s

--
James


>
>>--
>>James
>>
>>
>>
>>
>>On July 23, 1999 at 14:49:20, Will Singleton wrote:
>>
>>>Results, so far, for the TSCP nps test:
>>>
>>>Platform      Compiler       nps
>>>
>>>Mac G3/300    CW 3.1 opt     40,519
>>>P2/300        msvc opt       38,223
>>>K6-2/400      cygwin opt     37,516
>>>P2/300        gcc opt        29,205
>>>K6-2/400      msvc no opt    28,745
>>>K6-2/400      cygwin no opt  23,885
>>>K5/116        gcc -0         17,200
>>>P6/200        watcom opt     16,018
>>>
>>>Can anyone beat the Mac G3?  I'm surprised the K6/400 isn't as fast as the
>>>P2/300 (though different compilers were used).
>>>
>>>I would really like to see a few more machines here, as well as get
>>>corroborating results from similar machines.  Please submit results here or by
>>>email.
>>>
>>>For those of you willing to participate, here's the method I followed:
>>>
>>>Download TSCP v1.3.  ( http://ucsu.colorado.edu/~kerrigat/ )
>>>Modify the procedure Think() in search.c to:
>>>   search 6 ply.
>>>   Calculate and display NPS.
>>>Compile using optimize on.
>>>Run TSCP, and type ON.
>>>
>>>Thanks.
>>>
>>>Will



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.