Quantum computers are supposed to have capabilities far beyond conventional ones, but verifying that they actually have them is surprisingly difficult, because verifying their results can require calculations that become prohibitively difficult for classical machines. This is the problem of quantum verification, and a new experiment has found an ingenious solution: a game that tests each type of system. The trick? There is a mathematically proven limit to the performance of any classical computer. And when a team led by computer scientists Marcello Benedetti and Harry Buhrman of Quantinuum in the UK ran it on a trapped ion quantum system, it easily surpassed the limit. And as the test became more difficult, the gulf between the quantum system’s performance and the best possible classical performance became increasingly wider, according to their paper published in Nature Communications. frameborder=”0″ enable=”accelerometer; autoplay; writing on clipboard; encrypted media; gyroscope; picture in picture; web-share” referrerpolicy=”strict-origin-when-cross-origin” enablefullscreen> Quantum computers gain their unusual abilities from the strange physics that governs particles at the smallest scales. While bits in a classical computer represent information as one of two states – a 1 or a 0 – quantum bits, or qubits, can exist in a superposition of the two until they are measured. Measurement collapses that superposition, producing only one defined result. The mathematical nature of that superposition can be incredibly powerful from a computational point of view, allowing a quantum computer to quickly solve certain problems that are enormously difficult for a conventional computer. It was the computational power of superposition that the researchers set out to test. So, they devised a game based on a computational task called complement sampling. a classical computer, that is not enough to continue. It knows that the answer it was given belongs to A, so it knows not to return it. But it does not know which of all the other possible answers belongs to A and which belongs to BA. In fact, the researchers were able to demonstrate mathematically how well the best possible classical strategy could work. However, a quantum computer plays by very different rules. the entire set A in superposition, rather than reducing the response to a single sample. And, more importantly, it can manipulate that superposition before measuring it. Using what researchers call a “switcher” circuit, the quantum computer transforms the state representing set A into one representing its complement, set B. Only then does it measure the state, producing a response from set B. In an ideal, error-free quantum system, this strategy always wins. measure the incoming state to get a randomly selected answer from group A and then try to return a response from group B. This creates a huge gap between what the two types of systems are capable of. While an ideal quantum system wins every round, the advantage available to even the best possible classical strategy decreases exponentially as the number of bits and therefore the number of possible answers increases. And this is not just because researchers have not found a sufficiently intelligent classical algorithm. rely on assumptions about the difficulty of the calculation. This gave the team something unusually valuable: a test whose answers are easy to verify, but whose classical performance has a hard ceiling. So they took it to a real quantum computer, with the gap between quantum and classical performance growing exponentially (Benedetti et al., Nat. Commun., 2026). The researchers ran the complement sampling game on Quantinuum’s H2 trapped-ion quantum computers. different circuits and expanding their experiments up to 55 qubits. The real machines, of course, were not as perfect as the theory. But the quantum system still consistently exceeded the classical limit. In each experiment, the quantum computer scored so well that its results were statistically inconsistent with what any classical strategy could have achieved, closely following. The “arbiter” who chooses the initial answer and the “player” who analyzes it and gives the complement were implemented on the same quantum computer, with quantum teleportation used to simulate the channel. communication between them. Related: Quantum teleportation was achieved over the Internet for the first time A more rigorous future test would put them on separate quantum computers connected by a genuine quantum communication channel that is an obstacle that can be overcome in the next round of experiments. For now, the result provides a proof of concept, a new way to test quantum hardware that is efficient to verify, scalable and, most importantly, does not rely on unproven assumptions about what classical computers can and cannot. do. “Our test,” the researchers write, “demonstrates the power of quantum superposition in a way that ignores entanglement and non-locality.” Communications This article was fact-checked by Fiona MacDonald. While we take pride in our process, we’re human.
