[datavirtue]

This algorithm is still spinning a lot of random. I would guess that this is much less overhead than hashing but still seems like it could be significant.