USC and Quantum Elements Demonstrate Surface Code Scaling on IBM Heavy-Hex Processors - Quantum Computing Report

USC and Quantum Elements Demonstrate Surface Code Scaling on IBM Heavy-Hex Processors Surface code on heavy-hex. Researchers from the University of Southern California (USC) and quantum software developer Quantum Elements have demonstrated subthreshold surface code scaling on non-native processor geometries, publishing their peer-reviewed findings in Nature Communications. Executed across 156-qubit IBM Heron-generation superconducting QPUs, the study confirms that physical quantum processors do not require a native square-grid lattice to achieve distance-scaling error suppression under topological quantum memory codes. IBM’s heavy-hex architecture, arranged on the sites and links of a honeycomb lattice, introduces routing delays and idle gaps that typically cause non-Markovian dephasing and coherent ZZ crosstalk to accumulate. To overcome these connectivity constraints, the research team co-designed a depth-minimizing “fold-unfold” SWAP embedding using bridge ancillas alongside robust dynamical decoupling (DD). Implemented via Quantum Elements’ Orbit Qiskit Function, the DD protocol suppressed idle-time noise, allowing directional subthreshold scaling as the surface code distance expanded from d = 3 (37 qubits) to anisotropic (dx, dz) = (3, 5) and (5, 3) configurations (65 qubits). [ IBM Heavy-Hex Surface Code Demonstration & Benchmarking Metrics ]Hardware & Embedding SpecsError Suppression & Mitigation StackDemonstrated Subthreshold Scaling• QPU: IBM Heron 156-Qubit Processors• Connectivity: Heavy-Hex Honeycomb Lattice• Embedding: SWAP-based Fold-Unfold• Software Layer: Orbit Qiskit Function• Control Technique: Gap-Aware Robust DD• Noise Suppressed: ZZ Crosstalk & Dephasing• Scale: d = 3 (37Q) → (3,5) & (5,3) (65Q)• QEC Execution: Up to 10 Cycles (Depth >140)• Operation Count: 2,200 Entangling Gates Co-authored by Daniel Lidar (USC Center for Quantum Information Science & Technology Director and Quantum Elements CSO) and Arian Vezvaee, the paper introduces an SPAM-aware Entanglement Fidelity (EF) metric that accurately isolates per-cycle logical error rates without assuming stationary noise. The study also warns against “spurious subthreshold claims,” proving that unmitigated idle noise can falsely mimic scaling behaviors if advanced DD error suppression is not integrated into hardware control sequences. This experimental milestone builds on previous collaborative work between USC and Quantum Elements on classical digital-twin simulation. Earlier in June 2026, the partners published a Quantum Monte Carlo (QMC) method in Physical Review Letters that stochastically compresses open-system density matrices to simulate 97-qubit surface code circuits without encountering the classical sign problem, establishing a simulation foundation for hardware-calibrated QEC design. Review the official news release via EIN Presswire here, read the peer-reviewed paper in Nature Communications (Vezvaee et al., 2026) here, explore integration details via IBM Quantum Functions (Orbit) here, and examine our prior report on Quantum Elements & USC’s QMC Noisy Circuit Simulation Method here. September 17, 2026 Mohamed Abdel-Kareem2026-09-17T22:36:24-07:00 Leave A Comment Cancel replyComment Type in the text displayed above Δ This site uses Akismet to reduce spam. Learn how your comment data is processed.