[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Octave vs Scilab
From: |
Tatsuro MATSUOKA |
Subject: |
Re: Octave vs Scilab |
Date: |
Thu, 7 Oct 2010 11:40:08 +0900 (JST) |
Hello
It is difficult to build octave on windows with MSVC since the MSVC version
octave has been
disappeared from the unfortunate license issue. Build tools specially prepared
for MSVC were also
deleted from the octave-forge web site.
Perhaps for Intel C for windows, it is not also easy to use to build octave on
windows.
BTW I have octave-3.0.x both on MinGW and MSVC, the speed of octave MInGW
version and MSVC version
are not so different as far as I remember.
I have changed the N for2000 to 500 to save time (my machine is old (HT
pentimum 3.4 GHz))
Octave-3.0.3 msvc
octave:3> xmpl
N = 500
10.139
Octave-3.0.5 MinGW
octave.exe:3> xmpl
N = 500
7.7969
The versions are not the same so that I cannot say that the MinGW is faster
than that of MSVC.
I found a benchmark for
http://www.pubbs.net/200811/ruby/27246-some-benchmarks-with-mingw-jruby-msvc.html
(Sorry it is rather old. As far as I observed, the speeds depend on kinds of
tests.)
BTW, the results of the same test for the octave-3.2.4 MinGW on the sourceforge
site and
octave-3.3.52+ build myself using MinGW complier are shown the below
As references,
octave:2> xmpl
N = 500
5.2188
octave:2> xmpl
N = 500
3.9531
The speed of octave interpreter surely has improved with increasing the octave
version.
Regards
Tatsuro
There are version difference so that it is not fair to compare the speed
strictly but I can say that
the binaries generated by the MinGW complier has lesser performance compared to
that of MSVC
--- Judd Storrs wrote:
> On Wed, Oct 6, 2010 at 6:29 PM, Fotios Kasolis <address@hidden> wrote:
> > What makes loops slower in one interpreter compared to another?
>
> Since the difference seems to be observed on Windows, I'll posit
> that the build method probably is very influential. Do you happen to
> know how SciLab is compiled? Possibly they use Visual Studio or the
> Intel compilers instead of MinGW.
>
>
> --judd
>
--------------------------------------
Learn more about breast cancer - Pink Ribbon Campaign 2010
http://yj.pn/JAy9L7
- Octave vs Scilab, Fotios Kasolis, 2010/10/06
- Re: Octave vs Scilab, Dmitri A. Sergatskov, 2010/10/06
- Octave vs Scilab, John W. Eaton, 2010/10/06
- Re: Octave vs Scilab, Dmitri A. Sergatskov, 2010/10/06
- Octave vs Scilab, John W. Eaton, 2010/10/06
- Re: Octave vs Scilab, Fotios Kasolis, 2010/10/06
- Re: Octave vs Scilab, John W. Eaton, 2010/10/06
- Re: Octave vs Scilab, Fotios Kasolis, 2010/10/06
- Re: Octave vs Scilab, Judd Storrs, 2010/10/06
- Re: Octave vs Scilab,
Tatsuro MATSUOKA <=
- Equalis Octave group, Dave Barnes, 2010/10/07
- Re: Equalis Octave group, Jordi Gutiérrez Hermoso, 2010/10/07
- Re: Equalis Octave group, Søren Hauberg, 2010/10/07
- Re: Equalis Octave group, Dave Barnes, 2010/10/08
- Re: Equalis Octave group, Judd Storrs, 2010/10/08
- Re: Equalis Octave group, Jordi Gutiérrez Hermoso, 2010/10/08
- Re: Equalis Octave group, Dave Barnes, 2010/10/10
- Re: Equalis Octave group, Judd Storrs, 2010/10/11
- Re: Equalis Octave group, Dave Barnes, 2010/10/11
- Re: Equalis Octave group, Jordi Gutiérrez Hermoso, 2010/10/11