Towards a framework for abstracting accelerators in parallel applications: experience with cell

While accelerators have become more prevalent in recent years, they are still considered hard to program. In this work, we extend a framework for parallel programming so that programmers can easily take advantage of the Cell pro-cessor’s Synergistic Processing Elements (SPEs) as seam-lessly as possible. Using this framework, the same appli-cation code can be compiled and executed on multiple plat-forms, including x86-based and Cell-based clusters. Further-more, our model allows independently developed libraries to efficiently time-share one or more SPEs by interleaving work from multiple libraries. To demonstrate the framework, we present performance data for an example molecular dynam-ics (MD) application. When compared to a single Xeon core utilizing streaming SIMD extensions (SSE), the MD program achieves a speedup of 5.74 on a single Cell chip (with 8 SPEs). In comparison, a similar speedup of 5.89 is achieved using six Xeon (x86) cores. 1