作者: Raymond Namyst , Usman Dastgeer , Uwe Dolinsky , Siegfried Benkner , Sabri Pllana
关键词:
摘要: We discuss three complementary approaches that can provide both portability and an increased level of abstraction for the programming heterogeneous multicore systems. Together, these also support performance portability, as currently investigated in EU FP7 project PEPPHER. In particular, we consider (1) a library-based approach, here represented by integration SkePU C++ skeleton library with StarPU runtime system dynamic scheduling selection suitable execution units parallel tasks; (2) language-based Offload-C++ high-level language extensions Offload compiler to generate platform-specific code; (3) component-based specifically PEPPHER component annotating user-level application components metadata, thereby preparing them performance-aware composition. strengths weaknesses show how they could complement each other integrational framework