Improving Accuracy of Temperature Distribution and Energy-Saving Technology of Air Conditioners in Data Centers

Reducing the energy consumption of a data center has recently become more important because the data center market is rapidly expanding. We developed an “IT-facility linkage system” to deal with this energy saving requirement. This system reduces the amount of energy consumed in a data center by linking two systems. One is an optimized server load allocation system, and the other is an air conditioning optimization control system. One of the key technologies of the “IT-facility linkage system” is the precise prediction of the server inlet temperature. If we can comprehend future temperature distributions, we can reduce the amount of energy consumed in a data center by consolidating the IT load to more effectively cooled servers. Then, we carried out computational thermal fluid dynamics to predict the server inlet temperatures and evaluated the prediction precision by comparing the measured temperature data. As a result, we found that when we use a detailed rack model and reproduce the characteristic airflow of a data center such as the recirculation, we can predict the server inlet temperature to within less than 1.5°. Next, we carried out an experiment to verify the optimal server consolidation and confirmed that the effect of optimal consolidation can be predicted by simulation.