Fast and Accurate Evaluation of Cooling in Data Centers

Abstract

Cooling is a major component in the enormous energy consumption in data centers. Accurate evaluation of cooling inside a data center forms the backbone of all the attempts for improving cooling efficiency. Models based on computational fluid dynamics (CFD) are typically used for accurate evaluation, but have a drawback of high computation time. This paper presents a novel thermal predictor to evaluate data center cooling in seconds. The key idea is to extract information from a single instance of CFD simulation using metrics called as influence indices to build the fast thermal predictor. Then, this predictor can evaluate the cooling for altered operation of data center with comparable accuracy in seconds without the need for repetitive CFD simulations. This paper demonstrates the accuracy of the thermal predictor by comparing with CFD simulations for a sample, but realistic data center. The fast thermal predictor then successfully passed more challenging tests in a real production data center and proved its practical utility. The results of the thermal predictor compared with measurements carried out in the production data center are also presented. This fast thermal predictor is an important milestone in the development of a method for model-based real time control of data center cooling.

Tensor-Based Optimal Temperature Control of CRACs in Multi-Datacenters

“…To overcome the drawbacks of CFD, some dynamic models for fast estimation of the thermal environment in a data center have been developed. In [7], [8], it was found that for a fixed velocity field, temperature fields generated by multiple different boundary conditions can be superposed linearly. With this feature, the temperature field under real circumstances can be directly calculated by the superposition of a set of precalculated elementary processes.…”

Minimizing Energy Costs for Geographically Distributed Heterogeneous Data Centers

“…The airflow within the data center causes heat generated from nodes to propagate to other nearby nodes, thereby increasing the inflow temperature of those nodes. Using the notion of thermal influence indices [38] that were derived using computational fluid dynamics simulations, we can calculate the steadystate temperatures at compute nodes and CRAC units in each data center. Because we assume the same physical layout for each of the data centers ( Figure 3.1), we use thermal influence indices derived for one data center layout based on an average workload that would be executed by the data center.…”

Energy cost optimization for geographically distributed heterogeneous data centers

“…5) Thermal Model: Using the notion of thermal influence indices [13] that were derived using computational fluid dynamics simulations, we can calculate the steady-state temperatures at compute nodes and CRAC units in each data center. Because we assume the same physical layout for each of the data centers ( Fig.…”