Description
Parallel graph partitioning can be used from a library built against MPICH. HPC developers install it for distributed graph and sparse matrix workloads.
Numerical results depend on input data and MPI configuration. Validate outputs before using them in simulations.