Parallel Computing Toolbox 4.1

Microsoft Compute Cluster Server

Parallel Computing Toolbox and MATLAB Distributed Computing Server directly support Microsoft Windows Compute Cluster Server 2003. Using this scheduler, you can execute both distributed jobs, which do not involve any intertask communication, and parallel jobs, which involve intertask communication.

CCS offers several features, including:

• Ability to manage multiple applications that include MathWorks distributed computing tools  
• Simplified cluster deployment and management with monitoring tools, setup wizards for networking, installation services, and cluster security
• Integration with existing Windows infrastructure, allowing you to use existing tools for system, workload, user, and security management
• Integrated MPI layer based on MPI-2 standard
• Visual Studio 2005, including support for parallel applications with debugging capabilities

With Microsoft Windows CCS, the toolbox and the engine use Microsoft's implementation of MPI (MS-MPI) and not the default MPI distribution that is shipped with the toolbox and the engine.

Callbacks and dynamic addition of tasks to running jobs are not supported for CCS.

With the MathWorks job manager, the MATLAB Distributed Computing Server workers on the cluster run as a service and remain running at all times. With CCS, the workers run as an application and are started for the evaluation of tasks, and stopped when their tasks are complete. As a result, you incur an extra overhead of worker start-up and stop time. Create your tasks taking this overhead into account; for small tasks, this may represent a significant percentage of the overall execution time.  

The Programming Distributed Jobs and Programming Parallel Jobs sections in the Parallel Computing Toolbox documentation illustrates how you can use Windows CCS to submit jobs for execution on your Windows cluster.

For detailed information about this scheduler visit the Microsoft Windows Compute Cluster Server product web site.