SAN JOSE, Calif. The MathWorks is rolling out a new release of its MatLab environment that includes new parallel programming capabilities. Release 2008b available Monday (Oct. 20) lets users distribute parallel MatLab applications as standalone executables or software components that can more easily take advantage of computing clusters and multicore processors.
The enhancement takes the form of a new language construct in MatLab's Parallel Computing Toolbox. The new feature helps mix serial programs with parallel ones using the familiar single program, multiple data construct (also known as spmd) used by parallel programmers in technical applications for computer clusters.
"Initially this will probably be used by people familiar with parallel programming techniques, but we believe it is applicable to a broad set of users," said Loren Dean, director of engineering for MatLab products at The MathWorks (Natick, Mass.).
"We continue to look for ways to create high level abstractions to ease the job of tapping into parallel hardware, and we'll continue to explore high-level language constructs that help in this trend," he added.
Earlier this year, The MathWorks integrated parallel computing capabilities with many of the optimization routines called solvers used in key MatLab tool sets. "We have been moving from an explicit to an implicit parallel-programming model in which users don't need to program anything," said Dean.
The upgrade also includes the release of SimElectronics. The product handles modeling and simulating electronic and electromechanical systems in the Simulink environment.
The new release is available immediately and is being shipped to customers with current subscriptions to the MathWorks upgrade service.