arXiv.itsnick.co

an independent field guide to 23 million papers from arXiv & OpenAlex — not affiliated with arXiv.org

Computer Science · Hardware and Architecture

Assessing Application Efficiency and Performance Portability in Single-Source Programming for Heterogeneous Parallel Systems

August Ernstsson, Dalvan Griebler, Christoph Keßler

· International Journal of Parallel Programming

See on the map

Abstract

Abstract We analyze the performance portability of the skeleton-based, single-source multi-backend high-level programming framework SkePU across multiple different CPU–GPU heterogeneous systems. Thereby, we provide a systematic application efficiency characterization of SkePU-generated code in comparison to equivalent hand-written code in more low-level parallel programming models such as OpenMP and CUDA. For this purpose, we contribute ports of the STREAM benchmark suite and of a part of the NAS Parallel Benchmark suite to SkePU. We show that for STREAM and the EP benchmark, SkePU regularly scores efficiency values above 80% and in particular for CPU systems, SkePU can outperform hand-written code.

More like this