If your computing task is too big for your local computer, you may want to offload your calculation to networked nodes on a computer cluster. This may be a local cluster or a cluster in the cloud. Changes required to run your MATLAB® code on a cluster are often minimal. Use the MATLAB Job Scheduler (MJS) Profile to exploit computer clusters that have MATLAB Distributed Computing Server™ installed.
|Create cluster object|
|Create parallel pool on cluster|
|Get current parallel pool|
|Shut down cloud cluster|
|Start cloud cluster|
|Wait for cloud cluster to change state|
|Examine or set default cluster profile|
|Export one or more profiles to file|
|Import cluster profiles from file|
|Save modified cluster properties to its current profile|
|Save cluster properties to specified profile|
|Configure settings for Parallel Computing Toolbox client session|
Find out how to work with cluster profiles and discover cloud clusters running on Amazon EC2.
Run your application on workers in MATLAB Parallel Cloud™
parfor-loops on your
desktop, and scale up to a cluster without changing your code.
Specify your preferences, and automatically create a parallel pool.
Programming an independent job using a generic scheduler interface to distribute the tasks
How to program a communicating job using the generic scheduler interface