PoLiMEr

Abstract

Power and energy consumption are now key design concerns in HPC. To develop software that meets power and energy constraints, scientific application developers must have a reliable way to measure these values and relate them to applicationspecific events. Scientists face two challenges when measuring and controlling power: (1) diversity-power and energy measurement interfaces differ between vendors-and (2) distribution-power measurements of MPI simulations should be unaffected by the mapping of MPI processes to physical hardware nodes. While some prior work defines standardized software interfaces for power management, these efforts do not support distributed environments. The result is that the current state-of-the-art requires scientists interested in power optimization to write tedious, error-prone applicationand system-specific code. To make power measurement and management easier for scientists, we propose PoLiMEr, a user-space library that supports fine-grained application-level power monitoring and capping. We evaluate PoLiMEr by deploying it on Argonne National Laboratory's Theta system and using it to measure and cap power, scaling the performance and power of several applications on up to 1024 nodes. We find that PoLiMEr requires only a few additional lines of code, but easily allows users to detect energy anomalies, apply power caps, and evaluate Theta's unique architectural features.

Performance and power modeling and prediction using MuMMI and 10 machine learning methods

“…10 Several vendor-specific power management tools exist, such as Cray's CapMC and out-of-band and in-band power monitoring capabilities, 46 IBM EMON API on BG/Q, 45 Intel RAPL, 47 and NVIDIA's power management library. 48 In this work, we use simplified PoLiMEr 49 to measure power consumption for the node, CPU, and memory at the node level on Theta. PoLiMEr uses Cray's CapMC to obtain power and energy measurements of the node, CPU, and memory on Theta.…”