Parallelize heterogenous refinement

I’m often doing heterogenous refinement with several million particles and 6+ references and it would be extremely useful to have the ability to run this job type using multiple GPUs.

5 Likes