Cooperatively Improving Data Center Energy Efficiency Based on Multi-Agent Deep Reinforcement Learning

Ce Chi,Avinab Marahatta,Dehui Qiu,Penglei Song,Zhiyong Liu,Kaixuan Ji,Shikui Zhang,Fa Zhang

doi:10.3390/en14082071

Abstract

The problem of high power consumption in data centers is becoming more and more prominent. In order to improve the energy efficiency of data centers, cooperatively optimizing the energy of IT systems and cooling systems has become an effective way. In this paper, a model-free deep reinforcement learning (DRL)-based joint optimization method MAD3C is developed to overcome the high-dimensional state and action space problems of the data center energy optimization. A hybrid AC-DDPG cooperative multi-agent framework is devised for the improvement of the cooperation between the IT and cooling systems for further energy efficiency improvement. In the framework, a scheduling baseline comparison method is presented to enhance the stability of the framework. Meanwhile, an adaptive score is designed for the architecture in consideration of multi-dimensional resources and resource utilization improvement. Experiments show that our proposed approach can effectively reduce energy for data centers through the cooperative optimization while guaranteeing training stability and improving resource utilization.

Highlights

With the increasing development and application of cloud computing, data centers have become essential supporters and are located all over the world
By analysis [3], the IT system accounts for approximately 56% of the overall energy consumption in data centers, and the cooling system usually consumes about 30% of the total energy
The sum reward keeps stable and relatively high, which indicates that MAD3C has converged and stabilized

Summary

Introduction

With the increasing development and application of cloud computing, data centers have become essential supporters and are located all over the world. The rapidly growing number and scale of data centers have created a heavy burden on the energy supply and the environment. As reported in [1], 70 billion kWh of electricity in the U.S. was consumed by data centers in 2014, which accounted for about 1.8% of all electricity used in the U.S In addition, according to the statistics [2], data centers generate 0.3% of global carbon emissions every year. Energy consumption and environmental problems in data centers are becoming increasingly serious. By analysis [3], the IT system accounts for approximately 56% of the overall energy consumption in data centers, and the cooling system usually consumes about 30% of the total energy. For the energy consumption optimization of data centers, the focus reasonably tends to be on optimizing the IT system and the cooling system

Methods

Results

Discussion

Conclusion