HP-MPI Version 2.2.5 for Linux Release Note

HP-MPI V2.2.5 for Linux Release Note
What’s in This Version
31
n00 rank1
n01 rank2
n02 rank3
n03 rank4
% $MPI_ROOT/bin/mpirun [-prun|-srun] -n4 ./a.out
will run a.out with 4 ranks, 2 ranks per node, ranks are block allocated. Two nodes used.
n00 rank1
n00 rank2
n01 rank3
n01 rank4
% $MPI_ROOT/bin/mpirun [-prun|-srun] -n6 -O -N2 -m block ./a.out
will run a.out with 6 ranks (oversubscribed), 3 ranks per node, ranks are block allocated.
Two nodes used.
n00 rank1
n00 rank2
n00 rank3
n01 rank4
n01 rank5
n01 rank6
% $MPI_ROOT/bin/mpirun [-prun|-srun] -n6 -O -N2 -m cyclic ./a.out
will run a.out with 6 ranks (oversubscribed), 3 ranks per node, ranks are cyclic allocated.
Two nodes used.
n00 rank1
n01 rank2
n00 rank3
n01 rank4
n00 rank5
n01 rank6
Other forms of usage include allocating the nodes you wish to use, which creates a
subshell. Then jobsteps can be launched within that subshell until the subshell is exited.
% $MPI_ROOT/bin/mpirun [-prun|-srun] -A -N6
This allocates 6 nodes and creates a subshell.
% $MPI_ROOT/bin/mpirun [-prun|-srun] -n4 -m block ./a.out
This uses 4 ranks on 4 nodes cyclically from the existing allocation. Note that we asked for
block.
n00 rank1
n01 rank2
n02 rank3
n03 rank4