High-Performance Generic-Point Parallel Scalar Multiplication

A high-performance parallel scalar multiplication method is presented in this paper. The proposed method takes a set of generic points and performs their scalar multiplications in parallel using parallel elliptic curve cryptoprocessors. The method requires neither precomputations nor postcomputations to speed up scalar multiplication and outperforms existing methods. Furthermore, it is scalable and its performance improves as the number of consecutive requests increases. An implementation of the proposed method on an FPGA is presented. The area-time performance metric of the implementation outperforms other recently proposed fast implementations. Accordingly, our method is very attractive for high-speed applications.