Borrar filtros
Borrar filtros

What is the maximum number of workers I can use with the 'Processes' or 'local' profile in the Parallel Computing Toolbox?

84 visualizaciones (últimos 30 días)
I have installed Parallel Computing Toolbox and wish to make use of all the processor cores on my computer. What is the maximum number of workers I can use with Parallel Computing Toolbox?

Respuesta aceptada

MathWorks Support Team
MathWorks Support Team el 24 de Jul. de 2024
Editada: MathWorks Support Team el 24 de Jul. de 2024
All releases newer than R2014a can utilize up to 512 workers on a local machine through Parallel Computing Toolbox. Prior to R2013b, the maximum number of workers was 12.
By default Parallel Computing Toolbox will attempt to use a number of workers equal to the number physical cores of the computer. 
Prior to R2023a if parpool('local') is called without a specific number of workers the Parallel Preferences will set a preferred number of workers in the pool. The default value for this is 12 workers or the number of workers set in the cluster profile (whichever is lower).
If you wish to use multiple computers to perform parallel and distributed job processing, MATLAB Parallel Server will be required to utilize computational resources of other machines.
See also: http://www.mathworks.com/matlabcentral/answers/2042101-how-can-i-increase-the-number-of-matlab-workers-inside-parallel-computing-toolbox-to-use-all-the-log
for information on why the number of cores detected might not be all of the cores on your machine.

Más respuestas (0)

Categorías

Más información sobre Parallel Computing en Help Center y File Exchange.

Etiquetas

Aún no se han introducido etiquetas.

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by