With the increasing demand of privacy protection, secure multi-party computing, as a key technology of federated learning, has received extensive attention. This paper analyzes the optimization and performance evaluation methods of secure multi-party computing protocols. Through in-depth research on computational complexity, communication overhead, execution efficiency and security enhancement technologies, a variety of optimization strategies are proposed. To verify the effectiveness and practicality of these optimization strategies, a series of experiments are designed, such as simulating computing tasks on datasets of different sizes, comparing the execution time and resource consumption of protocols before and after optimization, etc. The experimental results show that the optimized secure multi-party computing protocol not only significantly improves the computing efficiency, but also enhances the overall security of the system, providing a valuable reference for improving the execution efficiency and security of the secure multi-party computing protocol.