Accelerate BigInteger Montgomery Multiplication on x86 architecture. Profiling points to Montgomery Multiplication as the primary bottleneck, so I will focus on acceleration of MM using IFMA instructions, changing the implementation from native code to generated intrinsic.