EVeREST: An Effective and Versatile Runtime Energy Saving Tool for GPUs

Abstract

Amid conflicting demands for ever-improving performance and maximizing energy savings, it is important to have a tool that automatically identifies opportunities to save power/energy at runtime without compromising performance. GPUs in particular present challenges due to (1) reduced savings available from memory bound applications, and (2) limited availability of low overhead performance counters. Thus, a successful tool must address these issues while still tackling the challenges of dynamic application characterization, versatility across processors from different vendors, and effectiveness at making the right power-performance tradeoffs for desired energy savings.

We propose Everest, a tool that automatically finds energy saving opportunities across GPUs at runtime. Specifically, Everest finds two unique avenues for saving energy using DVFS in GPUs in addition to the traditional method of lowering core clock for memory bound phases. Everest has very low overhead and works across different GPUs given its reliance on the minimum possible performance events for the needed characterization. Everest works at a finer granularity of individual application phases and utilizes built-in performance estimation to provide desired performance guarantees for an effective solution that outperforms existing solutions on the latest NVIDIA and AMD GPUs.