Parallel computing environments continue to grow in complexity, with a myriad of hardware and software components influencing the performance of an application. Open MPI exposes a large number of features of modern architectures through its run-time parameters. These parameters can be passed to the MPI library and influence the behavior of the MPI library without having to recompile the communication library or the application. Popular examples for MCA parameters include:
The Open Tool for Parameter Optimization (OTPO) is a new framework designed to aid in the optimization of the MCA paremeters. OTPO systematically tests a large numbers of combinations of Open MPI's run-time tunable parameters based on a user input file to determine the best set for a given platform.
Key features of OTPO include: