Computing results of large order eigen value problem faster

See for instance: FEniCS + MPI on docker inefficient? - #8 by dokken
or
https://fenicsproject.org/qa/8459/what-is-the-simplest-way-to-use-mpi/