Scalable Montgomery Modular Multiplication Architecture with Low-Latency and Low-Memory Bandwidth Requirement

PROJECT TITLE :

Scalable Montgomery Modular Multiplication Architecture with Low-Latency and Low-Memory Bandwidth Requirement (2014)

ABSTRACT :

In public-key cryptosystems, Montgomery modular multiplication is used widely. In traditional word-based algorithms, this work demonstrates how to relax the data dependence to increase the probability of reusing the current variable terms. We then introduced a new scheduling scheme to minimize the amount of memory access in the scalable architecture built with the greatly relaxed data dependence. Analytical findings show that the proposed scalable architecture's memory bandwidth requirement is almost 1/(w - 1) times that of traditional scalable architectures, where w denotes word size. In two consecutive iterations of the Montgomery modular multiplication algorithm, the proposed one also maintains a delay of exactly one cycle between the operations of the same words by using sufficiently processing elements. Experimental results show that the proposed one achieves an almost 54 percent reduction in power consumption with no loss in throughput compared to the one in the related work. The reduced amount of memory access not only contributes to lower power consumption, but also makes it simpler for operands to build scalable architectures for any accuracy.

Did you like this research project?

To get this research project Guidelines, Training and Code... Click Here