Enlisting GPU Power for Constraint Solving

Abstract

Applying parallelism to constraint solving seems a promising approach and it has been done with varying degrees of success. Early attempts to parallelize constraint propagation, which constitutes the core of traditional interleaved propagation and search constraint solving, were hindered by its essentially sequential nature. Recently, parallelization efforts have focussed mainly on the search part of constraint solving, as well as on local-search based solving. The most obvious source of parallelism are multicore processors and shared-memory multiprocessors, which impose the least burden on the developer. Lately, another source of parallelism has become pervasive, in the guise of GPUs, able to run thousands of parallel threads, and they have naturally drawn the attention of researchers in parallel constraint solving. And even if it turned out that the computational model of a GPU may be ill-suited to backtracking search, and that its parallelism potential may not be fully exploitable, they can still be a valuable resource. In this paper, we present ongoing work on a parallel solver, able to harness some of the GPUs' computational power to assist in the constraint solving process, and results showing why they should still be taken into account in parallel constraint solving.