What does OpenMP (the shared memory parallel runtime) have to do with optimizing away less-used branches? Am I missing something here or did you mean something

What does OpenMP (the shared memory parallel runtime) have to do with optimizing away less-used branches? Am I missing something here or did you mean something else instead of OpenMP?