Re: OpenMP

On Thu, 2 Aug 2007, Daniel Kobras wrote:
>
> Back at university, co-workers of mine have used a few OpenMP-parallel
> simulation codes in fluid dynamics. Experiences have been mostly similar
> to yours, but they've indeed seen an increase in overall performance.
> The improvements have been relatively small, though, compared to what
> similar applications gained from MPI parallelization or High-Performace
> Fortran. Which compiler did you test with, and on which platform?

It does seem to me that OpenMP is best for loops which don't do any 
system calls, and run for a fairly long time.

I tested on a 2 CPU SPARC Solaris system using GCC 4.2.1.

I don't think that OpenMP is responsible for the -blur slowness since 
I have observed that ImageMagick has become considerably slower in the 
past 1-1/2 years.  GraphicsMagick is usually getting a bit faster and 
better tuned while ImageMagick seems to be more cumbersome and slower 
as it adds more exotic features.

Bob
======================================
Bob Friesenhahn
[email protected], http://www.simplesystems.org/users/bfriesen/
GraphicsMagick Maintainer,    http://www.GraphicsMagick.org/


-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >>  http://get.splunk.com/