RhoninYang

Maybe it would be a good idea to also provide the code that the benchmark results are compared to, such as the C-loop and AVX versions?