[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
speed issues (was Re: Dot product (nx1).*(nxm))
From: |
Mike Miller |
Subject: |
speed issues (was Re: Dot product (nx1).*(nxm)) |
Date: |
Mon, 6 Dec 1999 11:43:26 -0600 (CST) |
This was my result:
octave:1> p=rand(100,1);
octave:2> q=ones(1,50);
octave:3> p2=p*q;
octave:4> t0=cputime; for i=1:1000, p2=p*q; end , cputime-t0
ans = 2.4600
But Michael Hanke <address@hidden> had this result:
octave:1> p=rand(100,1);
octave:2> q=ones(1,50);
octave:3> p2=p*q;
octave:4> t0=cputime; for i=1:1000, p2=p*q; end , cputime-t0
ans = 0.34773
Mike Miller Michael Hanke
=========== =============
version 2.0.14 2.0.13
OS Solaris 2.6 Solaris 2.5.1
machine Ultra 2 2200 Ultra 1
CPU speed 200 MHz 170 MHz
Michael Hanke wrote:
"Octave: The Fortran stuff is completely replaced by the Sun Performance
Library which is well tuned to the hardware. So the break-even appears at
rather high dimensions. Caching effects?"
Of course I'm very concerned that I have done a poor job of compiling
Octave on my Sun. I was happy that I could make it work, but now I think
I should find a way to make it run fast. I compiled using gcc. Was that
my mistake? Do I have to buy software from Sun that will allow me to
better optimize Octave? I'm pretty naive about compiler issues, so keep
it simple if you can! :-) (For example, I don't know what the Sun
Performance Library is and I don't know which Fortran stuff is replaced by
that library.)
Thanks in advance for any suggestions,
Mike
--
Michael B. Miller
University of Missouri--Columbia
http://taxa.psyc.missouri.edu/~mbmiller/
-----------------------------------------------------------------------
Octave is freely available under the terms of the GNU GPL.
Octave's home on the web: http://www.che.wisc.edu/octave/octave.html
How to fund new projects: http://www.che.wisc.edu/octave/funding.html
Subscription information: http://www.che.wisc.edu/octave/archive.html
-----------------------------------------------------------------------
- Dot product (nx1).*(nxm), Andre Bonfrer, 1999/12/04
- RE: Dot product (nx1).*(nxm), Heber Farnsworth, 1999/12/04
- RE: Dot product (nx1).*(nxm), Mike Miller, 1999/12/04
- Re: Dot product (nx1).*(nxm), Joao Cardoso, 1999/12/05
- Re: Dot product (nx1).*(nxm), Joao Cardoso, 1999/12/05
- Re: Dot product (nx1).*(nxm), Mike Miller, 1999/12/06
- Re: Dot product (nx1).*(nxm), John W. Eaton, 1999/12/06
- Re: Dot product (nx1).*(nxm), Joao Cardoso, 1999/12/06
- Re: Dot product (nx1).*(nxm), Mike Miller, 1999/12/06
- speed issues (was Re: Dot product (nx1).*(nxm)),
Mike Miller <=
Re: Dot product (nx1).*(nxm), Dr Andre Bonfrer, 1999/12/05