In multi-core computing environment, reasonable allocation of computing resources is very important. Total factor collaborative computing should not only pay attention to performance improvement, but also pay attention to energy consumption optimization. This paper analyzes the current theories and techniques of computing resource scheduling, and points out the limitations of traditional methods in dealing with large-scale tasks. A scheduling scheme based on iterative optimization algorithm is proposed, which dynamically adjusts the computing power distribution to minimize the power consumption of each compute node and ensure that the computing requirements are met. With this method, the system can significantly reduce the overall power consumption and improve the energy efficiency on the basis of guaranteeing the computing performance. Experiments show that compared with traditional methods, this scheme has obvious advantages in power consumption control and computing efficiency, especially in the multi-task parallel processing environment, it can better adapt to the changing computing load, and provide more stable and efficient computing services. This study provides a new solution for co-optimization of computing power and power consumption, which is helpful for resource allocation in data centers and other computing intensive environments.