Abstract: The trade-off between performance and energy consumption in ternary optical computer (TOC) systems has long been a critical concern. To unlock TOC's potential in high-performance computing while meeting energy efficiency demands, this study adopts system response time as the key performance metric. By integrating classical M/M/1 and M/M/c queuing models with a synchronized vacation mechanism, a dual-objective optimization approach for improving both system performance and energy efficiency is proposed. Focusing on how to achieve a balance between the two, a normalized objective function is established and solved using an intelligent optimization algorithm. Through dynamically searching for the optimal number of partitioned optical processors and vacation parameters, the method achieves dual optimization. Experimental results demonstrate that adjusting the processor partitioning and vacation parameters can further optimize system load distribution, effectively reduce energy consumption, and improve processing performance, thereby realizing a dynamic balance between performance and energy usage. This research provides a novel approach for TOC system optimization and offers valuable insights for advancing the practical deployment of TOC technology.