Server Energy Management in Data Centers: A Mode-Switching Model for Cost-Effective Operations

The increase in the demand for computing power has led to the deployment of a large number of computing infrastructures such as data centers. However, the operation of the computational core servers in them consumes a large amount of electricity, which brings about a very high cost of electricity consumption. Therefore, how to optimize the scheduling of servers to reduce the energy cost and improve the economic efficiency has become the focus of research. This paper explores energy modeling and optimization by distinguishing server operation modes: low, normal, and high frequency. The switching of the three operating modes in each time length is considered to make full use of the flexibility of the server’s work by cooperating with the scheduling of task loads in time. So the server’s energy consumption can be modelled in a refined way. In order to reduce the energy cost, this paper sets the lowest electricity cost as the goal, bases on the server energy consumption model, and combines it with the time-of-use price solving to get the economically optimal result, which provides energy-saving and cost-reducing scheduling references for the places deploying servers.