MATLAB: Using all cores in parallel computing toolbox

matlabpool

I want to use all the core present in my pc. I have used matlabpool('open',core number) command at the starting of the program and matlabpool('close') at the end. But it took longer time than single core job. I'm using R2009a version. please help.

Best Answer

You can open the Task Manager and view how many threads are crunching under the Performance tab. If all available threads show activity, then it could simply be that running in parallel is sub-optimal compared to running serially. This can happen, for example, if the size of the task is small compared to the overhead of setting up parallel workers and communicating data to them.

Related Solutions

MATLAB: Matlab Multi-Core

See the command maxnumcompthreads.

[EDITED, 25-Sep-2011 15:42 UTC]:

The warning is not really helpful, because it concerns a future release. TMW explained, that it is always possible, that a thread is starting other threads and that the limitation of core cannot be assured. But removing maxNumCompthreads for this reason is even worse. Running a program on a single core does not mean, that it is the same core all the time. The operating system can distribute the work between several cores. Under Windows you can bind a process to a core and there can be good reasons to do so (there should be equivalent methods under other OS also). But MATLAB is a very heterogenous program and I do not assume that this is really an advantage.

MATLAB: Use both multi core CPUs

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.

Best Answer

Related Solutions

MATLAB: Matlab Multi-Core

MATLAB: Use both multi core CPUs

Related Question