Using SMB Benchmark Suite to Evaluate the Potential of UPC on the Cray XE6
Hongzhong Shan, Naidong Zhao · International Conference on Multimedia Information Networking and Security · 2013
With the exponential increase of the degree of parallelism on the High-Performance Computing (HPC) platforms, the de facto MPI programming model has started to face great challenges. Researchers start to investigate other programming models. One of the most promising alternatives is Unified Parallel C (UPC). In this work, we compare the performance of MPI and UPC on a new HPC platform, the Cray XE6, using the popular Sandia MPI Micro-benchmark Suite (SMB). The results show that both MPI and UPC exhibit similar message overhead. However, UPC can deliver several times higher message rates than MPI for relatively small messages.