Select appropriate tasks sizes (big vs. small) that may have positive performance impacts on the workflow, and to run several (smaller) tasks by the help of job chaining.

Use mapping of processes to nodes, pinning of processes/threads to CPUs or cores, and setting memory affinities to NUMA nodes in order to speed up a parallel program.

Facilitate the speed up of program execution by setting appropriate runtime options (e.g. for MPI and OpenMP).