The Flying Fox Optimization Algorithm is a metaheuristic method inspired by the foraging and movement behavior of flying foxes. These creatures, a type of bat, provide a suitable model for designing an efficient algorithm through their flight between trees and search for food resources. In this study, the main objective is to develop and evaluate the performance of this algorithm in the MATLAB environment while leveraging CUDA architecture to enable parallel processing and GPU computational power. For this purpose, a proposed version of the algorithm was implemented, and its execution speed and efficiency were compared with the serial version. Simulation results indicate that parallel execution of the algorithm on the GPU significantly reduces execution time, performing over 314 times faster than the serial implementation. This remarkable improvement is primarily due to the simultaneous execution of blocks and optimal utilization of GPU computational resources. The findings suggest that the parallelized version of this algorithm demonstrates higher efficiency compared to traditional methods and can be highly effective in solving complex scientific and engineering problems, including numerical simulations, big data processing, and real-time modeling. Therefore, using CUDA architecture can play a crucial role in enhancing both the speed and quality of metaheuristic computations.