Fast Multiple Montgomery Multiplications Using Intel AVX-512IFMA Instructions

摘要

In this paper, we propose a fast implementation of multiple Montgomery multiplications using Intel AVX-512IFMA (Integer Fused Multiply-Add) instructions. The proposed implementation is based on a modified Montgomery multiplication. For Montgomery multiplication operands with 52 bits or fewer, the proposed implementation using Intel AVX-512IFMA instructions is up to approximately 12.22 and 4.30 times faster than the implementations using Intel 64 and Intel AVX-512F (Foundation) instructions on an Intel Core i3-8121U processor, respectively.