Researchers demonstrate quantum computers can solve constraint puzzles faster than classical methods — on tiny problems only.

A team used quantum computers to solve magic square puzzles by reformulating the problem as a quantum search, then compared their approach against classical computers. The quantum method showed the theoretical speedup predicted by quantum computing theory, but only worked on small instances because larger problems require too much memory.

This is proof that the theoretical advantage of quantum computers over classical ones actually shows up in implemented code, not just in math. The catch is immediate: the quantum approach works on toy problems only. Right now, this tells you quantum computers can beat classical computers at searching a space of possibilities — but the moment you need to solve something real, memory limits kick in.