Highly anticipated: IBM and researchers at the University of Chicago say they have reached a key milestone in quantum computing: running a computation that appears to be beyond the reach of classical machines while also providing evidence that the result is correct. The work addresses a problem that has lingered in the field for years. It is not enough for a quantum computer to produce an answer that is difficult for classical systems to replicate; researchers also need a way to verify that the answer is accurate.
Until now, most experiments have relied on random circuit sampling, a benchmark designed to produce outputs that are difficult for classical computers to simulate. The issue is that as those computations become more complex, verifying the results becomes nearly impossible without making assumptions about the quantum system itself.
The IBM-UChicago team took a different approach. They used a more structured type of quantum circuit that maintains the same level of computational difficulty as random circuit sampling while allowing errors to be detected during the computation. This makes it possible to evaluate how reliable the result actually is.
"Verification remains one of the biggest challenges in firmly establishing experimental quantum advantage," said Bill Fefferman, associate professor of computer science at UChicago and co-author of the paper. "This experiment develops techniques to better characterize the fidelity of hard quantum states under noise, increasing confidence that the quantum computer is solving a computationally hard problem."
The system relies on encoded quantum circuits, which are tied to error-correction methods designed to stabilize quantum operations. In this case, the team demonstrated a relatively large-scale setup involving 70 logical qubits. These qubits include built-in error protection, rather than being raw physical qubits.
Within that system, the researchers executed 2,415 logical two-qubit operations and 468 logical T gates. Both are standard measures of quantum-computing complexity, particularly in systems designed to handle errors.
The emphasis on error correction is notable. Rather than treating noise solely as a limitation, the experiment incorporates it into the design, allowing researchers to track how closely the computation follows its intended path.
Soumik Ghosh, a graduate student and co-author on the research, said the implications go beyond validation. "Beyond strengthening experimental validation, advances in verification have the potential to unlock practical applications for the next generation of quantum computers."
In terms of performance, the quantum system completed the computation in about 15 minutes. The researchers say comparable classical simulations would require impractical amounts of time, reinforcing the claim that the task falls outside the range of feasible classical computation.
IBM is framing the result as a turning point in how progress in quantum computing is evaluated. Jay Gambetta, director of IBM Research, said the company believes the field has entered what it considers the quantum advantage era, pointing to the successful execution of a computation that would be impractical for classical systems to replicate. He added that the work provides a measurable level of confidence in the accuracy of the result, which could help build trust in quantum systems as they scale to tackle more complex problems beyond the reach of classical computing.
The broader significance is tied to trust. Earlier demonstrations of quantum advantage focused mainly on whether a quantum system could outperform a classical one on a narrow task. This work places more emphasis on whether those results can be independently checked and understood.
That shift may be necessary if quantum computing is to move beyond controlled experiments and into practical use. For now, the results remain experimental, but they suggest a more realistic path toward systems that are not only powerful but also verifiable.