USC, Quantum Elements show surface code scaling on IBM Heron processors

A Nature Communications paper from USC and Quantum Elements shows the surface code can deliver below-threshold error protection on IBM Heron’s heavy-hex quantum processors, even though the hardware geometry does not match the code. The work points to a more flexible path to fault-tolerant quantum computing by combining error correction with dynamical decoupling.

Why it matters: - The result suggests superconducting quantum processors do not need to match the exact geometry of the surface code to gain scaling benefits. - That could broaden the hardware options available for fault-tolerant quantum computing. - The work also shows hybrid error-mitigation approaches can outperform either technique alone.

What happened: - Quantum Elements and the University of Southern California announced the publication of a peer-reviewed Nature Communications paper on September 15, 2026. - The paper, Surface code scaling on heavy-hex superconducting quantum processors, tested the surface code on two IBM Heron-generation processors. - The authors demonstrated logical qubits and below-threshold performance on IBM Heron’s heavy-hex architecture. - The paper was co-authored by Daniel Lidar, Quantum Elements chief scientific officer and director of the USC Center for Quantum Information Science & Technology, and Quantum Elements quantum research scientist Arian Vezvaee, among others.

The details: - The surface code is built for a square grid of connected qubits. - IBM Heron processors use a heavier-hex, sparsely connected layout that is not optimized for surface codes. - Mapping the code onto that hardware requires extra routing and creates idle periods where qubits accumulate errors as the code grows. - The researchers combined a depth-efficient code with dynamical decoupling to suppress idle-time noise. - That approach later became the basis for Quantum Elements’ Qiskit Function, Orbit. - The method enabled directional subthreshold scaling, meaning the code improved protection against the corresponding logical error as the code grew in one direction. - Lidar said the team used the dynamical decoupling techniques deployed in Orbit to show the expected improvements as the surface code distance increased. - Lidar also said hybrid decoupling-error correction approaches are likely to outperform either approach alone in future quantum processors.

Between the lines: - The paper reframes a common hardware-design tradeoff in quantum computing. - If error correction only worked when chip layout matched the code, hardware teams would have to optimize around one code rather than broader engineering goals. - This result reduces that constraint and points to a more adaptable route to fault tolerance. - Quantum Elements CEO Izhar Medalsy said the work shows the power of hybrid approaches and called it the first step toward entangled logical qubits.

What's next: - Quantum Elements says it will share more details soon about entangled logical qubits. - The company says those qubits would mark another advance toward fault-tolerant quantum computing. - The broader research path now appears focused on extending hybrid decoupling-error correction methods across more superconducting architectures.

The bottom line: - USC and Quantum Elements showed that surface code scaling can work on IBM’s heavy-hex hardware, strengthening the case for flexible, hybrid approaches to quantum error correction.