By Stephen Jenks

This blog is intended to share my experiences with programming parallel applications on the Cell Broadband Engine in a Sony Playstation 3 running Linux and on NVIDIA’s CUDA and any other asymmetric architectures I can get my hands on. I am an Assistant Professor at UC Irvine, so I do parallel architecture and systems for a living, but much of my experience has been with clusters and distributed systems. Therefore, programming the Cell, which has both shared memory parallelism and distributed memory on the SPEs, seemed like a fun challenge. The eventual goal is to see how some of our parallel execution models for chip multiprocessors map to the Cell, CUDA, and more. For details, please visit the blog.

Others are also working on scientific programming for the Cell, often with actual support and good results:

Professor David Bader runs the Sony-Toshiba-IBM Cell Center of Competence at Georgia Tech.

The Barcelona Supercomputing Center is instrumental in Cell SDK development.

The CUDA home page has resources for CUDA on Windows, Linux, and Mac.