Every few months a headline announces that a quantum computer has done something no classical machine could match. A few weeks later, researchers argue about whether it really did. Both reactions make sense once you know what quantum advantage means, what it takes to prove, and why even a confirmed win does not mean your laptop is obsolete.
What Quantum Advantage Means
IBM describes quantum advantage as the point at which a quantum computer is shown to be faster, more accurate, or more cost-effective than classical-only methods for some problem, in a way that can be rigorously validated. IBM’s blog on its quantum advantage tracker adds that advantage “will not come from a single announcement.” That framing is useful: the bar is a comparison against the best classical method, not against a straw man, and it has to hold up to outside checking.
Note what is missing. Nothing in the definition says the problem must be useful. A quantum computer can win at a contrived task and still be of no practical value, which is exactly what happened with the first big claim.
The 2019 Claim: “Quantum Supremacy”
In October 2019, Google published a paper in Nature titled “Quantum supremacy using a programmable superconducting processor.” Its Sycamore chip used 53 working qubits, and the authors reported that it took about 200 seconds to sample a quantum circuit a million times, a task they estimated would take a state-of-the-art classical supercomputer approximately 10,000 years.
The task itself was designed to be hard for classical computers and easy for the quantum chip. It had no everyday use. The classical estimate was also open to challenge, since improved classical algorithms can shrink such gaps. The field has since leaned toward the milder term “quantum advantage,” which is the wording both IBM and Google use in their recent announcements.
The 2025 Claim: A Result You Can Check
A more recent headline came in October 2025. In its Quantum Echoes announcement, the company said an algorithm running on its Willow chip, a 105-qubit array, was 13,000 times faster than the best classical algorithm on one of the world’s fastest supercomputers. The key word is “verifiable.” Google defines that as a result that can be repeated on its own quantum computer, or one of similar quality, so the answer can be checked rather than just trusted.
Google is also careful about what it does not claim. It calls the work a significant step toward real-world applications, not an application itself. The molecular-structure experiments, run on molecules of 15 and 28 atoms, are described as a proof of principle, and the post puts practical uses in the future as hardware scales. As with any speed-up claim, the real test is whether researchers can find a faster classical method, which is why the community scrutinizes such comparisons.
Why IBM Built a Tracker
Rather than relying on one-off announcements, IBM and partners launched an open community Quantum Advantage Tracker, where researchers submit candidate problems and compare them against leading classical methods. IBM’s post says the tracker had more than 30 submissions using IBM and Quantinuum hardware at the time of writing. The idea is that a claim should survive repeated attack from classical-algorithm experts before anyone calls it settled.
How to Read the Next Quantum Headline
-
What was the task? A purpose-built benchmark is very different from a problem a business cares about. For real uses, see our look at quantum computing applications already happening today.
-
What was the classical baseline? Look for comparison against the best known classical algorithm, not an old one.
-
Can others verify it? Repeatable, checkable results carry more weight than one-time outputs.
-
How noisy was the hardware? Errors limit how far these machines can scale, which is the heart of why quantum error correction is so hard.
-
Does the wording match the claim? “Advantage” on one task is not a general speed-up, a point we cover in our piece on quantum computing myths.
Different hardware designs are also competing to reach advantage first, and they have different strengths, as our comparison of the types of quantum computers explains.
FAQs
What is the difference between quantum advantage and quantum supremacy?
They describe similar milestones: a quantum computer beating classical methods on a task. “Supremacy” was the term in Google’s 2019 paper, while “advantage” is the milder wording both Google and IBM use in their recent announcements.
Has quantum advantage been achieved?
Google claims it, including a verifiable version in 2025, but claims are judged against the best classical methods and keep being tested. IBM stresses that advantage will not come from a single announcement.
Does quantum advantage mean quantum computers are useful now?
Not by itself. The 2019 task had no practical use, and Google describes its 2025 molecular work as a proof of principle, with practical applications expected as hardware scales.
Why do classical computers sometimes catch up?
Researchers can devise better classical algorithms after a quantum claim appears, narrowing the gap. That back-and-forth is why shared tools like IBM’s tracker compare candidates against leading classical methods.
Discussion about this post