Towards Energy-Efficiency: CNTD MERIC Approach for Energy-Aware MPI Applications

Abstract

Energy efficiency is a major challenge in High-Performance Computing (HPC) systems, impairing their scale, performance, and sustainability. However, despite technological and research progress, there is still a lack of software methods to measure and assess the energy efficiency of computing codes at scale. This is also exacerbated by the emergence of newer ISAs in the HPC computing spectrum with non-unified interfaces for power and energy monitoring. In this work, we present CNTD MERIC, which integrates two state-of-the-art energy monitoring and optimization libraries for HPC systems, COUNTDOWN and MERIC. COUNTDOWN is an energy-aware runtime system for MPI applications. MERIC is a platform-agnostic runtime system and energy measurement library that optimizes energy efficiency by adjusting hardware configurations. CNTD MERIC combines the benefits of the two approaches with low overhead, resulting in a portable power management runtime system for MPI applications. We tested the newly developed integrated library in both ARM and x86 compute nodes in the production environment of the IT4Innovations supercomputing center (IT4I). Our results show that CNTD MERIC achieves similar performance to the original COUNTDOWN and MERIC in terms of energy optimization and power/energy measurement, with negligible overheads within -5% to +3% with respect to the original COUNTDOWN configurations. We also implemented CNTD MERIC for multi-architecture (x86 and ARM) comparison between Intel Sapphire Rapids and A64FX processors, and the results show that A64FX achieves significantly lower execution time, reduced energy-to-solution and lower average power consumption (110–132 W vs. 400–590 W), which confirms its efficiency for energy-efficient HPC systems.