Parallelism in homogeneous refinement

In v2, there appears to be no option to select multiple GPUs for homogeneous refinement. Is the parallelization automatic, or is it not parallelized?

Hi @abreuer,

Currently, homogeneous refinement can only run on a single GPU.