Comment by kardos

So is the SIMD the magic piece here, or is it the interpolation search? If the data is evenly distributed, that is pretty optimal for the interpolation search..