If have not compiled in MPI mode, Rosetta will use one processor per instance of Rosetta you start.
If you compiled with MPI, Rosetta will attempt to use however many processors you told MPI to use; I believe the standard mpirun application flag is -np, but I’m not sure.
Rosetta is not internally multithreaded (well, ab initio isn’t) and will never use more than one processor at a time; the MPI communication layer manages separate independent threads rather than using multiple processors for one job.