osu_reduce mpi benchmark with 16 nodes, 256 ranks  Scheduling frequency 6h , nodes=16:ppn=16:xe,walltime=00:05:00

Tests: MPI runtime, small section of cray network