Abstract: Aiming at the problems of the increasing demand for energy consumption and the low utilization of resources in current data center,the paper proposes a task scheduling and energy management model for multi-regional energy efficiency between multiple geographic data centers. First of all,photovoltaic power generation,electricity price,task quota,UPS energy storage battery status information are obtained. Then according to the maximum resource utilization principle and the boxing approximate algorithm,a task allocation strategy and the number of servers activated in each time slot are dynamically adjusted. Finally,according to the size of the task,different power supply methods are adopted,and the cycle cost of the battery and other information are calculated to determine whether to conduct photovoltaic grid connection or battery charging and discharging operation. The effectiveness of the proposed strategy is verified with an example.