arXiv:2610.11658
Error-Corrected Memory and Logic on a Heavy-Hex Superconducting-Qubit Processor
Campbell K. McLauchlan, Georgia M. Nixon, Julien M. Drouet, Xanda C. Kolesnikow, Seok-Hyung Lee, James Raftery, Karthik Siva, Stephen D. Bartlett, Benjamin J. Brown, Robin Harper
Abstract
Large quantum computations require high-fidelity error-corrected memory and fault-tolerant logical operations. To achieve these with superconducting circuits, quantum processors will need many components operating with low error rates. In practice, fabrication of monolithic superconducting devices produces unavoidable heterogeneity in component performance, which the quantum computer must accommodate. We implement strategies to operate a large dynamic compass code on an array of 156 superconducting qubits arranged on a heavy-hexagonal lattice, making use of the entire monolithic quantum processor and redesigning our measurement schedule to exclude underperforming components. We demonstrate improvements of up to $35\%$ in logical error probabilities for a memory experiment by using a noise-informed decoder combined with strategies to exclude the worst-performing qubits and couplers, along with post-selecting on shots with a low probability of leakage errors. In addition to these memory results, we also observe performance gains of up to $56\%$ when these strategies are applied to measurement-based logic, demonstrated by stability experiments. We furthermore demonstrate high-fidelity fault-tolerant logic through lattice surgery. Through pre-selection filtering techniques, we can obtain Bell fidelities above $77\%$, and further logical gap post-selection using exclusive decoding increases these fidelities to above $97\%$. Our experiments demonstrate that high performance can be maintained in key elements of fault-tolerant quantum memory and logic when working on monolithic superconducting quantum processors with underperforming components, by using a suite of noise-informed quantum error correction strategies.
AI Summary
The paper studies how heterogeneous component quality and leakage affect error-corrected memory and measurement-based logic on a 156-qubit heavy-hex superconducting processor. It implements dynamic compass codes with measurement schedules redesigned to exclude defective qubits and couplers, combining these with ACES circuit-specific noise characterization, BeliefMatching decoding, measurement soft information, and leakage filtering. Performance is assessed through memory decay, stability experiments, and lattice-surgery Bell-state preparation. For a full-device $7\times7$ memory patch, the combined strategies yield logical error probabilities per measurement round of $4.11\%$ in $Z$ and $2.56\%$ in $X$, reductions of $32.3\%$ and $35.2\%$ relative to the reported-noise baseline. The best unpostselected results are $4.4\%$ and $2.9\%$. Defect exclusion reduces both code distances from seven to five. The resulting patch approximately matches average $5\times5$ performance, but average $3\times3$ patches still perform better. In $5\times5$ stability experiments, defect exclusion extends observed error suppression to 30 measurement rounds; without it, performance plateaus around 12 rounds. Combined improvements reduce failure probability by up to $56\%$. Lattice surgery gives unfiltered Bell-fidelity bounds of $58.6\%$–$73.5\%$ for two $3\times3$ patches. Pre-selection raises the lower bound to $77.3\%$ while retaining $13.3\%$ of attempts; additional logical-gap filtering raises it above $97.6\%$ at $27\%$ relative yield, approximately $3.6\%$ overall. The processor lacks mid-circuit reset, leakage is filtered rather than actively removed, and fidelity estimates use two correlators rather than full tomography. The applicability of observed soft-information benefits beyond these experiments remains unresolved.
Central claim may not hold: The reported post-selected Bell-fidelity bounds combine correlators accepted using different readout-dependent decoding filters without establishing a common selected logical-state ensemble, so the claimed >97% Bell fidelity is not rigorously supported by those conditional correlations alone.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.
- Comments
- 28 pages, 20 figures, 5 tables