94. Toward Understanding HPC- Big Data Convergence Using Cloud Platforms
Authors: Shweta Salaria (Tokyo Institute of Technology)Kevin Brown (Tokyo Institute of Technology)Hideyuki Jitsumoto (Tokyo Institute of Technology)Satoshi Matsuoka (Tokyo Institute of Technology)
Abstract: The path to HPC- Big Data convergence has resulted in numerous research that demonstrates the performance-cost tradeoff between running applications on supercomputers and cloud platforms. Previous studies typically focus on evaluating scientific HPC benchmarks on old cloud configurations, failing to consider the new opportunities offered by cloud platforms. We present a comparative study of the performance of representative big data benchmarks, or "Big Data Ogres", and HPC benchmarks running on supercomputer and cloud. Our work distinguishes itself from previous studies in a way that we explore the new generation of compute-optimized instances oriented for HPC applications by Amazon EC2. Our results indicate that the new cloud environment appears to be a promising HPC system alternative for a set of high performance computing applications.
Two-page extended abstract: pdf