Exploiting the Cell/BE Architecture with the StarPU Unified Runtime System

Abstract

Core specialization is currently one of the most promising ways for designing power-efficient multicore chips. However, approaching the theoretical peak performance of such heterogeneous multicore architectures with specialized accelerators, is a complex issue. While substantial effort has been devoted to efficiently offloading parts of the computation, designing an execution model that unifies all computing units is the main challenge.

We therefore designed the StarPU  runtime system for providing portable support for heterogeneous multicore processors to high performance applications and compiler environments. StarPU  provides a high-level, unified execution model which is tightly coupled to an expressive data management library. In addition to our previous results on using multicore processors alongside with graphic processors, we show that StarPU  is flexible enough to efficiently exploit the heterogeneous resources in the Cell  processor. We present a scalable design supporting multiple different accelerators while minimizing the overhead on the overall system. Using experiments with classical linear algebra algorithms, we show that StarPU  improves programmability and provides performance portability.