In this case specifically, SIMD is available, via compiler supported auto-vectorization, and intrisics, no need to go down into Assembly.