With the rapid development of artificial intelligence and cloud computing, data centers are shifting towards high power density and energy efficiency. Traditional air-cooling methods are reaching their limits in cooling high-power-density IT cabinets. Liquid cooling, with its superior efficiency, is poised to become the mainstream solution for reducing energy consumption and addressing these challenges. This study focuses on liquid-cooled data centers with high single-cabinet power density and auxiliary air cooling. A liquid-cooled data center absorption chiller is de-signed. The absorption chiller uses the high-temperature water from the chip side of the liquid-cooled data center as the driving heat source for the generator and the heat from other components as the low-temperature heat source for the evaporator. The heat in the absorber and condenser is removed by circulating water and discharged via an outdoor cooling tower, enabling heat distribution and waste heat utilization in the liquid cooling system. A computational program is developed to simulate the absorption ma-chine’s operation. The numerical model simulates steady-state changes in chip temperature, component temperature, cooling water temperature, and dynamic conditions such as sudden external parameter changes. The absorption chiller performs better under high supply liquid temperature, high component temperature, and low cooling temperature conditions and maintains stability across multiple operational conditions. To address server load fluctuations in data centers, a device operation strategy based on solution storage is proposed. By altering the flow direction of the solution be-tween the concentrated and diluted solution tanks and the distribution of condensate water in the refrigerant water tank, the strategy provides solutions for high/low IT load and low natural temperature conditions. The study also explores energy quality storage under high heat supply and low-temperature cooling circulating water conditions.