Math Doesn’t Lie

Ars Technica posted a bit of a delve into the math behind using a quantum computer to break AES-128 encryption and the news is good.  The rumour has always been that QRQC will halve the amount of work needed to crack algorithms like AES-128, rendering it relatively quick to crack and that we would have to use AES-256 at the very least.  However, that math does not check out.

Two mathematicians tried to express the somewhat complex math by using much smaller numbers.  Instead of the 3.4 x 10^38 possible combinations that AES-128 boasts, they reduced it to a possible 256 combinations for their examples.  If you split the work four ways, then each of you could try 32 combinations which is a reasonable amount of work.  However quantum computers do not behave the same as classic computers and when you assign several of them parallel tasks they use Grover’s algorithm to split the work. If you break apart the task between several quantum computers, the amount of tries they have to chew through ends up being the same, not half the tried as the doomsayers have been predicting.

Follow the link to read “Quantum Computers Are Not a Threat to 128-bit Symmetric Keys” if you want to read a proper explanation of why AES-128 is not going anywhere … yet.