CARTAD: Compiler-Assisted Reinforcement Learning for Thermal-Aware Task Scheduling and DVFS on Multicores

Abstract

As the power density of modern CPUs is gradually increasing, thermal management has become one of the primary concerns for multicore systems, where task scheduling and dynamic voltage/frequency scaling (DVFS) play a pivotal role in effectively managing the system temperature. In this article, we propose <i>CARTAD</i>, a new reinforcement learning (RL)-based task scheduling and DVFS method for temperature minimization and latency guarantee on multicore systems. The novelty of <i>CARTAD</i> framework is that we exploit the machine learning technique to analyze the applications&#x2019; intermediate representations (IRs) generated by a compiler and identify an important feature which is critical for predicting the application&#x2019;s performance. With the newly explored feature, we construct an RL-based scheduler with the more effective state representation and reward function such that the system temperature can be minimized while guaranteeing applications&#x2019; latency. We implement and evaluate <i>CARTAD</i> on real platforms in comparison with the state-of-the-art approaches. Experimental results show <i>CARTAD</i> can reduce the maximum temperature by up to 16 &#x00B0;C and the average temperature by up to 10 &#x00B0;C.