Based on GPU high-performance parallel computation, a parallel acceleration of regular integration in boundary face method (BFM) has been implemented in CUDA (Compute Unified Device Architecture) programming environment.Comparative computations compared with traditional regular element integration were made on both NVIDIA GTX680 GPU and Intel(R) Core(TM) i7-3770K CPU.Numerical examples show that, a speedup of 8.3x has been achieved at the same level of accuracy.