Quantum Advantage: What Has Actually Been Shown So Far?

Quantum advantage (once called “quantum supremacy”) means a quantum computer doing a task that no ordinary computer can do in a reasonable time. It says nothing about whether the task is useful.

Useful quantum advantage means beating ordinary computers on a problem people actually care about. The first has arguably been achieved; the second has not yet been clearly demonstrated.

In 2019, Google reported in the journal Nature that its 53-qubit Sycamore processor completed a random sampling task in about 200 seconds, and estimated that a top supercomputer would need 10,000 years. IBM quickly argued that a better classical method could do it in about 2.5 days, and later classical techniques narrowed the gap further.

• ◆2020, China: the light-based Jiuzhang experiment performed a sampling task that its researchers estimated would take classical supercomputers vastly longer.
• ◆2023, IBM: a 127-qubit processor produced results for a physics simulation using error-mitigation techniques. Classical researchers soon reproduced many of the results, showing how quickly classical methods keep improving.
• ◆2024, Google Willow: a random sampling task finished in under five minutes that Google estimated would take a leading supercomputer 10 septillion (10²⁵) years. The same chip demonstrated error correction that improved as it scaled.

Each claim sparks new classical algorithms that shrink the gap. This back-and-forth is healthy and pushes both fields forward. But it means “beats a supercomputer” headlines should be read carefully.

The milestone that matters most for businesses is a quantum computer solving a useful problem, for example in chemistry or materials science, better than any classical method, with a result that can be checked. Most experts expect this to need error-corrected machines, so progress in error correction is the best signal to watch.