Measurement of data centers' energy efficiency: A data envelopment analysis approach

With the high costs of power, energy efficiency has rapidly become a critical consideration issue when evaluating data centers. Three of the most important data center metrics, PUE, DCEP, and CUE, are examined closely. Data Envelopment Analysis (DEA) is a fractional linear programming-based technique that has gained wide acceptance in centers years due to its effectiveness in comparing energy efficiency of data centers. First, expounds feasibility of the data centers' energy efficiency assessment with DEA is introduced. Second, the application method and model properties are discussed. Finally, based on the DEA method for evaluation of energy efficiency, the results are concluded. The results indicate that the proposed method can overcome the limitations of single index parameters (PUE), without subjective analysis of the input element weights. Some proposals are given for improving the conservation planning and construction in data center energy efficiency evaluation.