Improving researcher access to Parallel MATLAB technologies

Researchers need look no further than the Ohio Supercomputer Center for a convenient way to access the MATLAB Parallel Computing Toolbox and ParaM, two versions of Parallel MATLAB technologies.

Parallel MATLAB enables researchers to access remote supercomputers within MATLAB, a user-friendly computer-programming language developed by The MathWorks. The Parallel Computing Toolbox is a commercial product from the Math- Works, while ParaM, an open-source addition to MATLAB, was developed by MIT Lincoln Laboratories and OSC. These technologies allow researchers to increase their productivity through the use of parallel computing without needing to re-write their MATLAB codes in a more typical parallel computing language.

“The Center’s work with the Department of Defense High Performance Computing Modernization Program, combined with OSC’s top-notch training programs, fortifies our national reputation as experts in deploying Parallel MATLAB. Our goal is to share our knowledge,” said Ashok Krishnamurthy, Ph.D., senior director of research at OSC. The service will support a number of concurrent remote users, offering access to multiple nodes on the Center’s IBM Cluster 1350. User manuals, training courses and real-world examples of production code will round out its offerings.

“While our initial focus is serving Ohio’s user community, the project will expand to businesses and national participants,” Dr. Krishnamurthy added.

For example, scientists using the Teragrid, the world’s largest distributed cyberinfrastructure for open scientific research, soon will have access to OSC’s Parallel MATLAB services as part of Pittsburgh Supercomputer Center’s pending award (Track2C) from the National Science Foundation. This prestigious grant supports acquiring, deploying and operating a high performance computing system for the national science and engineering community.

--

Project lead: Ashok Krishnamurthy, Ph.D., OSC

Research title: Applications in parallel MATLAB

Funding source: Ohio Supercomputer Center