Fast Implementation And Optimization Of GPU-based SM2 Digital Signature And Verification Algorithm

Elliptic Curve Cryptography(ECC)algorithm has great significances in the field of digital signature.SM2 is a elliptic curve public key cryptography algorithm issued by our National Password Authority.Compared to others public key encryption technology,it has the shorter length of key under the conditions of same security.With the extensive application of digital signature technology,the demand for digital signature performance is also increasing,and the rapid realization of ECC has been widely concerned by academia and industry.High performance general purpose calculation of Graphic Processing Units(GPU)has been widely used in various fields.In this paper,the SM2 digital signature and verification algorithm are implemented on the GPU platform,and the underlying algorithm(such as elliptic curve group operation,large integer domain operation)is optimized,and we optimized and analyzed SM2 algorithm.The main work is as follows:The running model of CPU and GPU heterogeneous platform is expounded in detail,and the running model of SM2 digital signature and verification algorithm on Open CL platform is proposed.By using CPU and GPU heterogeneous programming methods,we focus on the large integer inverse operation and the point multiplication operation of the elliptic curve group in the finite field.The experimental results show that the SM2 signature algorithm running on the AMD Radeon R9 290 GPU is 1.44 to 22.70 faster than the serial algorithm on the Intel Xeon E5620 CPU,and the acceleration ratio of verification is 1.39 to 2.95.The SM2 GPU parallel optimization algorithm proposed in this paper can significantly improve the speed of SM2 digital signature and verification.