Exploiting Dynamic Reuse Probability to Manage Shared Last-level Caches in CPU-GPU Heterogeneous Processors | Publicación