Local Temperature Control in Data Center Cooling: Part II, Statistical

[摘要] The data center plays a more and more important role with the increasing demands of internet and commodity computing. As a result of the evolution of the microprocessor and increasing demands of customers, the power density in data centers becomes much greater. Mechanical engineers face challenges in cooling system design. Recent research focuses on local control algorithms which can help to optimize local temperature distribution, and thus increase the energy efficiency of the system. This paper analyzes the relationship between rack inlet temperature and vent group configuration from the view of statistics. It also examines whether rack power can affect inlet temperature. Combined with part I in this series, the conclusion can be used as the basis of local control algorithms. 23 Pages

[发布机构] HP Development Company

[学科分类] 计算机科学（综合）

[关键词] data center cooling;smart cooling;local temperature control;vent tile;time series analysis;correlation matrix