[apw]

It's possible that someone has found a way to speed up scrypt. We could solve for a lower bound on what that speedup must be.