Trusted Advantage: IBM and UChicago Certify Quantum Supremacy Without Classical Proof

IBM and University of Chicago researchers did not claim quantum advantage on a toy circuit. On July 30, 2026, they reported a 70-logical-qubit, depth-70 Clifford circuit doped with 468 T gates — encoded in spacetime quantum error correction — that produced roughly 2,000 samples in 16 minutes while classical simulation approaches faced prohibitive runtimes. The paper (arXiv:2607.25941) introduces something equally important: a device-dependent fidelity certificate that does not require classical reference simulations.

Trusted advantage, not headline advantage

Random circuit sampling (RCS) has long been the benchmark for separating quantum from classical compute — but cross-entropy benchmarking requires calculating ideal output probabilities classically, which becomes intractable at scale. Without a reference, researchers cannot prove the quantum output was faithful.

The IBM–UChicago team addresses both problems with structured circuits that retain RCS-style hardness while admitting encoding in a quantum code. Syndrome post-selection on 97 physical qubits suppresses effective gate error rates by 10×, yielding a 95%-confidence fidelity lower bound of 0.284 for the doped magic state.

Jay Gambetta, IBM Research director, put the milestone bluntly: "We are now firmly in the quantum advantage era."

The numbers that matter

From the University of Chicago announcement and PRNewswire release:

• 70 logical qubits shielded by error correction.
• 2,415 logical two-qubit operations and 468 logical T gates.
• Effective logical error rates 10× lower than physical error rates.
• IBM quantum hardware completed the task in ~15 minutes; leading classical approaches faced prohibitive runtimes.

The circuits and results are openly released on IBM's Quantum Advantage Tracker.

Why verification changes the field

Prior advantage demonstrations often traded verifiability for scale — or required classical simulations that themselves could not run at the demonstrated size. Here, the certificate comes from circuit structure and syndrome measurements, with substantially weaker noise assumptions than fidelity-proxy benchmarks.

IBM also announced related advantage demonstrations with ecosystem partners Qedma, Algorithmiq, RIKEN, and BlueQubit — each introducing different verification frameworks for utility-scale algorithms beyond classical reach.

For developers and businesses, the July 30 result is less about one circuit and more about a new foundation for trusting quantum outputs when classical verification is mathematically impossible.

### Sources

• arXiv — Sampling hard circuits with verifiably high fidelity (2607.25941) (July 28, 2026)
• University of Chicago News — IBM, UChicago demonstrate quantum advantage (July 30, 2026)
• PRNewswire — IBM and UChicago trusted quantum computation (July 30, 2026)