Google's "Quantum Echoes": the first verifiable quantum advantage?
Google says its Willow chip ran a physics calculation about 13,000 times faster than the best classical method, and that the answer can be checked. The result is real and peer-reviewed. The "verifiable" part hasn't been verified by anyone else yet.
- Announced
- Published
- Peer-reviewed
- Independently repeated
- Useful in practice
Verdict: Real and still standing. "Verifiable" isn't verified yet.
- The claim
- Google says its Quantum Echoes algorithm, run on the Willow chip, is the first verifiable quantum advantage: about 13,000 times faster than the best classical approach on a leading supercomputer, with an answer other quantum computers can check.
- What was actually shown
- A peer-reviewed experiment (Nature, October 2025) on Google's superconducting Willow processor, using 103 qubits in total. It measured a quantity physicists use to study how information spreads through a quantum system, at a scale the authors say known classical methods can't match.
- What it doesn't show
- A repeat on another quantum computer, which is what "verifiable" promises. A useful result that classical computers can't get. Google's chemistry demonstration (a "molecular ruler" for two molecules) is a proof of principle, not an advantage.
- Why it matters
- Earlier "quantum advantage" claims used tasks with answers nobody could check, and some were later matched by better classical methods. A result that other machines can check would be a real step towards trustworthy quantum computing.
- What would settle it
- Another quantum computer of similar quality reproducing the result, and classical methods continuing to fall short as they improve.
What Google claimed
In October 2025, Google announced that an algorithm it calls Quantum Echoes, run on its Willow chip, ran about 13,000 times faster than the best classical approach on one of the world's fastest supercomputers. Google called it the first verifiable quantum advantage: unlike earlier demonstrations, the answer is a measurable quantity that another quantum computer of similar quality should be able to reproduce.
What the evidence shows
The experiment is published in Nature, so it has passed peer review. It ran on Willow, Google's superconducting processor, with 103 qubits taking part. The team sent a signal into the quantum system, nudged a single qubit, then ran the system "backwards" to pick up an amplified echo. Physicists use this quantity, called an out-of-time-order correlator, to study how information spreads and scrambles.
Two points matter for anyone reading the headline:
- The 13,000 figure is Google's comparison. It's measured against the best classical method the team knew of. Speed-up claims like this can shrink when someone finds a smarter classical method; that has happened to earlier quantum advantage claims.
- So far, the result is holding up. In April 2026, an outside group posted a preprint concluding that one of the strongest families of classical techniques (tensor networks with belief propagation) can't feasibly reproduce the experiment. That supports Google's claim, though the preprint isn't peer-reviewed yet.
What it doesn't show yet
"Verifiable" is the headline word, and it's the part still waiting. As of 27 October 2026, we found no published repeat of the result on another quantum computer. Until that happens, the advantage is checkable in principle, not checked in practice.
It also isn't a useful application yet. Google and UC Berkeley showed a "molecular ruler": using the technique on measurement data from two organic molecules. That's a proof of principle for chemistry, not a result classical computers can't reach.
Why it matters
Most earlier "quantum advantage" results used random tasks whose answers couldn't be checked, which made them hard to trust and easy to challenge. A result built on a checkable physical quantity is a better kind of claim. If another machine reproduces it, this becomes one of the strongest pieces of evidence yet that quantum computers can do something classical ones can't.
Our verdict: real and still standing; "verifiable" isn't verified yet. We'll update this page if the result is repeated on another machine, or if a classical method catches up.
Sources
- Observation of constructive interference at the edge of quantum ergodicity, Nature (Google Quantum AI and collaborators), DOI 10.1038/s41586-025-09526-6.
- A verifiable quantum advantage, Google Research blog, 22 October 2025.
- Our Quantum Echoes algorithm is a big step toward real-world applications for quantum computing, Google blog, 22 October 2025.
- Constructive interference at the edge of quantum ergodic dynamics, arXiv preprint 2506.10191.
- Tensor Networks with Belief Propagation Cannot Feasibly Simulate Google's Quantum Echoes Experiment, arXiv preprint 2604.15427.
Machine records on QubitBenchmark: google-willow.