Parallel Matlab: The Next Generation
Jeremy Kepner, Nadya Travinin · Defense Technical Information Center (DTIC) · 2003
The true costs of high performance computing are currently dominated by software. Addressing these costs requires shifting to high productivity languages such as Matlab. The development of MatlabMPI (www.ll.mit.edu/MatlabMPI) was an important first step that has brought parallel messaging capabilities to the Matlab environment, and is now widely used in the community. The ultimate goal is to move beyond basic messaging (and its inherent programming complexity) towards higher level parallel data structures and functions. The pMatlab Parallel Toolbox provides these capabilities, and allows any Matlab user to parallelize their program by simply changing a few characters in their program. The performance has been tested on both shared and distributed memory parallel computers (e.g., Sun, SGI, HP, IBM, Linux and MacOXS) on a variety of applications.