Parallelizing constraint programs
Laurent D. Michel · 2010
The availability of commodity multicore and multiprocessor machines and the inherent parallelism in constraint programming search offer significant opportunities for constraint programming. Both constraint-based local search and finite-domain techniques can dramatically benefit from parallelization. Yet, currently available libraries and languages offer very limited support to exploit the inherent parallelism and the high human cost incurred to develop parallel solutions confine programmers to sequential implementation. The fundamental challenge boils down to: how to exploit parallelism transparently to speed up constraint programs. This talk shows how to parallelize constraint programs (both constraint-based local search and finite-domain programs) transparently with minimal changes to the code and illustrates all the ideas with the COMET platform.