> Unlike crypto there is no mathematical lower bound for computation

this is what I feel, but is there any scientific proof on that?