OpenMP does not seem to work when compiled with nvcc.

In particular:
`paracabs::multi_threading::set_n_threads()` does not seem to have any effect anymore.

Currently this is not much of an issue since we only want to use MPI in combination with CUDA.