IBM and University of Chicago Demonstrate Verified Quantum Advantage With Error-Corrected Logical Qubits

IBM and researchers at the University of Chicago announced on July 30 what they describe as a demonstration of genuine quantum advantage: a computation that ran beyond the practical reach of the best classical simulation methods while also providing verifiable confidence that the result was correct. The achievement, detailed in a paper titled 'Sampling hard circuits with verifiably high fidelity,' addresses a problem that has dogged earlier quantum advantage claims, namely proving a result is accurate when no classical computer can check it by simply redoing the calculation.

The team used a new error-correction method to encode 70 logical qubits, protecting them from the noise that normally corrupts quantum calculations, and used them to execute 2,415 logical two-qubit operations and 468 logical 'T gates,' both benchmarks for how complex a quantum circuit is. The full computation finished in roughly 15 minutes, a runtime the researchers say would require an infeasible amount of time using leading classical computing methods. Notably, the logical error rate achieved was about ten times lower than the underlying physical error rate, showing that the error-correction scheme itself was working as intended even at a large scale.

The result is being described as one of the largest error-correction demonstrations conducted to date and a step toward quantum computers that can be trusted for problems beyond the reach of ordinary supercomputers, rather than merely faster at niche, hard-to-verify tasks. Researchers say the verification method developed alongside the computation itself, not just the raw scale of qubits used, is what distinguishes this result from earlier quantum advantage claims that faced skepticism over whether their outputs could be independently confirmed.