IBM and researchers at the University of Chicago announced a milestone on July 30: a quantum computation they say achieves the fundamental criteria for 'quantum advantage' — performing computations beyond the reach of leading classical simulation methods while providing trust that the results are accurate.

In their paper 'Sampling hard circuits with verifiably high fidelity,' the team executed one of the world's largest-known error-correction demonstrations: 70 logical qubits shielded from errors, running 2,415 logical two-qubit operations and 468 logical T gates. Because the circuits were encoded, effective logical error rates came out 10 times lower than physical error rates. The computation took about 15 minutes on IBM hardware — a task many leading classical simulation approaches could not complete in feasible time. The circuits and results are openly released on the Quantum Advantage Tracker.

'We are now firmly in the quantum advantage era,' said Jay Gambetta, Director of IBM Research. The work also tackles quantum computing's trust problem: as problems get harder, classical verification becomes infeasible, so the team used a structured alternative to random circuit sampling that retains the same hardness while detecting errors during computation.

The same day, Algorithmiq and IBM announced a second demonstration: a simulation of heterogeneous quantum matter run on an IBM Heron processor that has withstood open challenge for eight months on the Quantum Advantage Tracker, with no classical method reliably reproducing results across the full problem regime. Algorithmiq is also open-sourcing monoprop, a benchmark package that lets any group stress-test quantum advantage claims. As CEO Sabrina Maniscalco put it: 'For an exponential technology like quantum computing, a verified, openly contested instance of advantage is the inflection point.'

Sources