An efficient cache design for scalable glueless shared-memory multiprocessors