Run many functions (adaptively) on many cores (>10k) using mpi4py.futures, ipyparallel, or dask-mpi. :tada: