Some Experiments and Issues to Exploit Multicore Parallelism in a Distributed-Memory Parallel Sparse Direct Solver

MUMPS is a parallel sparse direct solver, using message passing (MPI) for parallelism. In this report we experiment how thread parallelism can help taking advantage of recent multicore architectures. The work done consists in testing multithreaded BLAS libraries and inserting OpenMP directives in the routines revealed to be costly by profiling, with the objective to avoid any deep restructuring or rewriting of the code. We report on various aspects of this work, present some of the benefits and difficulties, and show that 4 threads per MPI process is generally a good compromise. We then discuss various issues that appear to be critical in a mixed MPI-OpenMP environment