Sharing the instruction cache among lean cores on an asymmetric CMP for HPC applications | Publicación