The Parallel Computing Toolbox is a toolbox within MATLAB. It lets you solve computationally-intensive and data-intensive problems using MATLAB and Simulink on your local multicore computer or the Shared Computing Cluster (SCC). GPU operations are also supported provided that Nvidia GPU graphics cards are installed. Many of the SCC nodes are equipped with 1 or 3 Nvidia GPUs. Parallel processing operations such as parallel for-loops, parallel numerical algorithms, and message-passing functions let you implement task- and data-parallel algorithms in MATLAB. Converting serial MATLAB applications to parallel MATLAB applications generally requires few code modifications and no programming in a low-level language is necessary. You can run your applications interactively or in batch.

On MATLAB R2011a or newer, the PCT supports up to 12 workers (i.e., cores) on a local computer, such as a PC. Running a MATLAB PCT job on a single SCC node is, for practical purposes, similar to running it on a PC. However, to use more than 12 workers, you must run it on the SCC via the Mathworks’ Distributed Computing Server (DCS). This requires that you set up, a priori, a batch configuration (R2011a, R2012b) for jobs in excess of 12 cores.

If your MATLAB application is to run multiple independent tasks, such as parametric studies of an analysis; image processing of large number of image files; or Monte Carlo simulations, please consult the Running Multiple MATLAB Tasks page. This alternative method does not require the PCT and hence you won’t need to learn PCT.

Please visit the following Boston University-developed tutorial on the PCT. This tutorial also includes details on how to run MATLAB multi-processor batch jobs on the SCV Katana Cluster.

A PCT User’s Guide is available online at the Mathworks site. (HTML,   PDF).