MATLAB: Use both multi core CPUs

cpuImage Processing ToolboxMATLABmulticoreparallel computingParallel Computing Toolbox

Hi,

I have a two CPU multi core machine (16 cores each). I want to use a parfor loop on all 32 cores. But maxNumCompThreads gives me 16. In preferences I changed the maximum number of workers to 32, but when I open a parallel pool it still has only 16 workers. What do I need to configure to make use of all cores? If it matters my parfor loop is a perfect parallelisable task, i.e. no communication between iterations is needed.

Thanks for your help!

Best Answer

You need to edit the 'local' parallel profile to set the maximum number of workers to 32. See the doc. Once you've done that, you should be able to issue the following command:

parpool('local', 32)

and launch 32 workers.

Note that if MATLAB by default does not consider hyper-threaded cores, as these tend not to perform well for numerically intensive computations. So, you might well find that running 32 workers is actually slower than running 16, depending on the precise nature of your computation.

Related Solutions

MATLAB: What is the maximum numbers of workers I can use with MATLAB Parallel Computing Toolbox

All releases newer than R2013b do not have a limit on the number of workers (CPU cores) a local machine can utilize through Parallel Computing Toolbox. Prior to R2014a, the maximum number of workers was 12.

Please note that the default number of workers Parallel Computing Toolbox will attempt to use is 12. This number can be set higher in the "local" cluster profile. Additionally, if you edit the profile and leave the number of workers as unset, Parallel Computing Toolbox will attempt to use a number of workers equal to the physical cores of the computer.

If you wish to use multiple computers to perform parallel and distributed job processing, MATLAB Distributed Computing Server will be required to utilize computational resources of other machines.

MATLAB: -maxNumCompThreads, hyperthreading, and parpool

I was wrong about maxNumCompThreads not working on parpool workers. I think the problem was that the code I was using:

parfor j = 1:2
  tic
      maxNumCompThreads(2);
      workersCompThreads(j) = maxNumCompThreads;
      i = 1;
      while toc < 200
          a = randn(10^i)*randn(10^i);
          i = i + 1;
      end
end

used so much memory by the time I checked CPU utilization that the bottleneck was I/O and the extra threads were already shut down. When I did the following:

parfor j = 1:2
  tic
      maxNumCompThreads(2);
      workersCompThreads(j) = maxNumCompThreads;
      i = 4;
      while toc < 200
          a = randn(10^i)*randn(10^i);
      end
end

The extra threads started and stayed running.

As for the second issue, I got a confirmation from the Mathworks that the parent Matlab process won't start more threads than the number of physical cores, even if you explicitly raise the limit beyond that. So in the documentation, the sentence:

"Currently, the maximum number of computational threads is equal to the number of computational cores on your machine."

should say:

"Currently, the maximum number of computational threads is equal to the number of physical cores on your machine."

Best Answer

Related Solutions

MATLAB: What is the maximum numbers of workers I can use with MATLAB Parallel Computing Toolbox

MATLAB: -maxNumCompThreads, hyperthreading, and parpool

Related Question