MATLAB won't use idle Processors

  • Context: MATLAB 
  • Thread starter Thread starter blumist
  • Start date Start date
  • Tags Tags
    Matlab
Click For Summary

Discussion Overview

The discussion revolves around optimizing MATLAB performance for a 2-dimensional Dirichlet boundary condition heat transfer problem. Participants explore methods to utilize more CPU resources during computation, focusing on parallel processing options available in MATLAB.

Discussion Character

  • Technical explanation
  • Exploratory

Main Points Raised

  • One participant describes a MATLAB program that only utilizes 25% of CPU power, leading to slow processing times.
  • Another participant suggests using the distributed computing toolbox or opening multiple instances of MATLAB to distribute the workload.
  • A participant inquires about the "distributed computing toolbox" after confirming their operating system is Windows 7 64-bit.
  • There is a suggestion to check if the toolbox is included in the participant's MATLAB version, with a recommendation to consult the documentation if available.
  • A participant mentions finding a Parallel Computing Toolbox and expresses intent to seek further assistance if issues arise.
  • Another participant recommends GPU computing as an alternative, specifically mentioning Jacket as a potentially superior option to the Parallel Computing Toolbox.

Areas of Agreement / Disagreement

Participants generally agree on the need for parallel processing to improve MATLAB performance, but there are multiple approaches suggested, including the use of the distributed computing toolbox, multiple MATLAB instances, and GPU computing. The discussion remains open without a consensus on the best solution.

Contextual Notes

Participants have not confirmed the availability of the distributed computing toolbox in their MATLAB versions, and there are unresolved questions regarding the effectiveness of different parallel processing methods.

blumist
Messages
18
Reaction score
0
I wrote a program on Matlab that is supposed to solve a 2-dimensional dirichlet boundary condition heat transfer problem. In programming context, it uses a number of iterations in series to calculate the next value using data acquired from the previous iterations.
I am running MATLAB 64-bit on a core i-3 2.3 Ghz with 4 GB RAM. During the processing it tends to use only 25% of the complete computing power that my CPU can support and takes quite long computing otherwise simple problems. Any ideas on how I can increase it?

Thanks !
 
Physics news on Phys.org
The distributed computing toolbox allows use of multiple processors. Alternatively, you could open up multiple instances of MATLAB and manually distribute the work load.

The core-i usually comes on windows machines iirc. What your OS?
 
Pythagorean said:
The distributed computing toolbox allows use of multiple processors. Alternatively, you could open up multiple instances of MATLAB and manually distribute the work load.

The core-i usually comes on windows machines iirc. What your OS?
My OS is Windows 7 64-bit. Can you enlighten me a bit over the "distributed computing toolbox"? Thanks :)
 
Do you have the toolbox in your version? Go to the MATLAB menu (bottom left corner) and them to toolboxes, if it's not in that list, try "more".

If you don't have it (your license has to include it) then it's not much use learning about.

If you do have it, then read up on the documentation and ask a more specific question after trying to implement it.
 
I found a Parallel Computing Toolbox. I'll get back to you if I face a problem. Thanks!
 
For parallel MATLAB, GPU computing is a really good option. I recommend taking a look at Jacket for that (better than the Parallel Computing Toolbox), http://accelereyes.com/compare
 
Last edited by a moderator:

Similar threads

  • · Replies 1 ·
Replies
1
Views
3K
  • · Replies 6 ·
Replies
6
Views
4K
  • · Replies 3 ·
Replies
3
Views
5K
  • · Replies 4 ·
Replies
4
Views
3K
  • · Replies 6 ·
Replies
6
Views
4K
  • · Replies 10 ·
Replies
10
Views
3K
  • · Replies 3 ·
Replies
3
Views
3K
  • · Replies 4 ·
Replies
4
Views
12K
  • · Replies 2 ·
Replies
2
Views
3K
  • · Replies 1 ·
Replies
1
Views
3K