Cloud computing resources and energy are gradually showing a trend of geographical distribution, and an innovative computing task placement strategy needs to be designed for a more cost-effective and low-carbon environmental protection cloud deployment. In this paper, from the perspective of optimizing the use of new energy to the maximum extent, a computing power task placement model considering the power difference factors of geographically distributed new energy generation is developed. An effective power prediction model for new energy generation is designed to deal with the volatility and intermittency characteristics of new energy generation. Finally, an intelligent computing task placement algorithm is studied to ensure the maximum use of new energy and sustainable development of the computing network system. The experimental results show that the proposed algorithm can effectively predict the data fluctuation of new energy generation, and the new energy consumption can be increased by 14.9%–36.8% compared with the baseline algorithm.