Utilizing the Double-Precision Floating-Point Computing Power of GPUs for RSA Acceleration

Algorithm 3

DPF-based parallel Montgomery multiplication () algorithm: Converting Phase.

Input:

: Thread ID;

: Number of processed limbs per thread;

: Number of threads per Montgomery multiplication, where ;

: Redundant-format sub-result, where

;

Output:

: Simplified-format sub-result, where

;

(1)

(2) for   to   do

(3)

(4)

(5) end for

(6) while  carry of any thread is non-zero  do

(7)

(8)for   to   do

(9)

(10)

(11)end for

(12) end while