Architectural support for parallel reductions in scalable shared-memory multiprocessors | Publicación