Message passing-aware power management on many-core systems

Dynamic frequency and voltage scaling (DVFS) techniques have been widely used for meeting energy constraints. Single-chip many-core systems bring new challenges owing to the large number of operating points and the shift to message passing from shared memory communication. DVFS, however, has been mostly studied on single-chip systems with one or few cores, without considering the impact of the communication among cores. This paper evaluates the impact of voltage and frequency scaling on the performance and power of many-core systems with message passing (MP) based communication, and proposes a power management policy that leverages the communication pattern information to efficiently traverse the search space for finding the optimal voltage and frequency operating point. We conduct experiments on a 48-core Intel Single-Chip Cloud Computer (SCC), as our target many-core platform. The paper first introduces the runtime monitoring infrastructure and the application suite we have designed for an in-depth evaluation of the SCC. We then quantify the effects of frequency perturbations on performance and energy efficiency. Experimental results show that runtime communication patterns lead to significant differences in power/performance tradeoffs in many-core systems with MP-based communication. We show that the proposed power management policy achieves up to the 70% energy-delayproduct (EDP) improvements compared to existing DVFS policies, while meeting the performance constraints.