We do not know that. Using thermodynamics we can establish a limit on how much computation we can do (if I recall correctly Applied Chryptography says that a su

We do not know that. Using thermodynamics we can establish a limit on how much computation we can do (if I recall correctly Applied Chryptography says that a super nova does not have sufficient energy to count to 2^256 in base 2). However, we have no theoretical bases regarding how much computation is involved in breaking AES256. All we do know is that we (the public) do not know of a way to do it.