With the rapid development of the digital economy, the energy consumption and carbon emissions of data centers have significantly increased. The concept and methodology of Data Center Integrated Energy Systems (DC-IES) have emerged as an important trend for energy efficiency and emissions reduction in the context of the carbon peaking and carbon neutrality goals. this paper employs an energy "quality" analysis method based on exergy and characterizes the change of key equipments’ exergy efficiency with load rates. It constructs a capacity configuration and operational optimization model for DC-IES driven by low-carbon energy sources, achieving multi-objective optimization of DC-IES energy system economic cost and efficiency. The computational results indicate that, The optimization scheme that takes into account the dynamic exergy efficiency outperforms the scheme treating the exergy efficiency as a fixed value, with reductions in the rate of loss of exergy, economic cost and carbon emissions of 2.6%, 1.9% and 4.8%, respectively. Compared to the economically optimal single-objective optimization scenario, the multi-objective optimization scenario achieves a notable reduction of 22.72% in DC-IES carbon emissions and a decrease in the exergy loss rate by 20.73%. Furthermore, compared to scenarios with the minimum exergy loss rate and the minimum carbon emissions separately, the multi-objective optimization scenario reduces DC-IES economic costs by 54.54% and 60.78%, respectively. In comparison to scenarios relying solely on grid electricity supply, carbon emissions witness a significant decrease by 40.97%.