[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [gnugo-devel] GNU Go 3.3.23
From: |
Paul Pogonyshev |
Subject: |
Re: [gnugo-devel] GNU Go 3.3.23 |
Date: |
Wed, 16 Jul 2003 23:22:35 +0000 |
User-agent: |
KMail/1.5.9 |
Marco Scheurer wrote:
> More results:
>
> On 3.2 first_batch
> 3.3.23 1025.510u 7.310s 40:04.13 42.9% 0+0k 0+16io 0pf+0w
> 3.2 1249.210u 6.040s 24:16.30 86.1% 0+0k 0+23io 0pf+0w
>
> ...
if you have pike compiled, the best way to estimate speeds of
different gnu go versions (different gtp-supporting engines)
is to use interface/gtp_examples/twogtp.pike. most important,
this runs real games, not a collection of tests. and the
pike script computes real move generation time without any
support from engines. the only drawback is that it is _real_
time, not cpu time. therefore, to get reliable results, you
must leave the box alone.
Paul
- [gnugo-devel] GNU Go 3.3.23, bump, 2003/07/15
- RE: [gnugo-devel] GNU Go 3.3.23, Hellwig Geisse, 2003/07/16
- Re: [gnugo-devel] GNU Go 3.3.23, Marco Scheurer, 2003/07/16
- Re: [gnugo-devel] GNU Go 3.3.23, Marco Scheurer, 2003/07/17
- Re: [gnugo-devel] GNU Go 3.3.23, bump, 2003/07/17
- Re: [gnugo-devel] GNU Go 3.3.23, Paul Pogonyshev, 2003/07/17
- Re: [gnugo-devel] GNU Go 3.3.23, bump, 2003/07/17
Re: [gnugo-devel] GNU Go 3.3.23, Gunnar Farneback, 2003/07/17