Performance and energy efficiency analysis of HPC physics simulation applications in a cluster of ARM processors

摘要

We analyze the feasibility and energy efficiency of using an unconventional cluster of low-power AdvancedrnRISC Machines processors to execute two scientific parallel applications. For this purpose, we have selectedrntwo applications that present high computational and communication cost: the Ondes3D that simulates geophysicalrnevents, and the all-pairs N-Body that simulates astrophysical events. We compare and discussrnthe impact of different compilation directives and processor frequency and how they interfere in Time-to-rnSolution and Energy-to-Solution. Our results demonstrate that by correctly tuning the application at compilerntime, for the Advanced RISC Machines architecture, we can considerably reduce the execution time andrnthe energy spent by computing simulations. Furthermore, we observe reductions of up to 54.14% in Timeto-rnSolution and gains of up to 53.65% in Energy-to-Solution with two cores. Additionally, we consider thernimpact of two processor frequency governors on these metrics. Results indicate that the powersave governorrnpresents a smaller instantaneous power consumption. However, it spends more time executing tasks,rnincreasing the energy needed to achieve the solution. Finally, we correlate the energy consumption with thernexecution time in the experimental results using Pareto. These findings suggest that it is possible to explorernlow-powered clusters for high-performance computing applications by tuning application and hardwarernconfiguration to achieve energy efficiency.