Exploiting task-parallelism in message-passing sparse linear system solvers using OmpSs

Abstract

We introduce a parallel implementation of the preconditioned iterative solver for sparse linear systems underlying ILUPACK that explores the interoperability between the message-passing MPI programming interface and the OmpSs task-parallel programming model. Our approach commences from the task dependency tree derived from a multilevel graph partitioning of the problem, and statically maps the tasks in the top levels of this tree to the cluster nodes, fixing the internode communication pattern. This mapping induces a conformal partitioning of the tasks in the remaining levels of the tree among the nodes, which are then processed concurrently via the OmpSs runtime system. The experimental analysis on a cluster with high-end Intel Xeon processors explores several configurations of MPI ranks and OmpSs threads per process showing that, in general, the best option matches the internal architecture of the nodes. The results also report significant performance gains for the MPI+OmpSs version over the initial MPI code.