Managing Parallelism

For the foreseeable future, parallelism will be one of the main (if not the main) source of continued performance improvements, from scaling out larger clusters, increasing the number of processors on a chip, to wider data parallelism within cores. However the parallelism available in many emerging technologies is often restricted in
some way that doesn't fit into the standard models of parallel computation (for example the PRAM model). Managing parallelism in these new technologies requires rethinking the design of algorithms and schedulers that run these algorithms. The mornings will primarily be devoted to talks by domain experts that explain various emerging technologies that involve parallelism in some way. The afternoons would be devoted to discussions of these technologies, and interesting algorithmic issues that arise in the management of these technologies.