Lightweight threading and vectorization with OpenMP in ACME