Architectural support for scalable speculative parallelization in shared-memory multiprocessors | Publicación