Structural Comparison of Error Mitigation Methods for Ising Machines: Penalty-Spin Model versus Stacked Model

Abstract

Error-mitigation methods for Ising machines are reexamined not merely as noise-suppression techniques but as a structural design problem of replica-coupled Ising models. Using simulated annealing as a hardware-noise-free testbed, we systematically compare the penalty-spin (PS) model, which couples replicas through a centralized auxiliary layer, with the stacked model, which couples adjacent replicas directly. Numerical experiments on the quadratic assignment problem reveal that the ferromagnetically coupled stacked model stably maintains constraint satisfaction and improves solution quality over a broad parameter range, exhibiting favorable scalability with both the number of replicas and problem size. In contrast, the PS model suffers from cooperation collapse at large parallelism: many-replica averaging in the PS layer washes out sparse solution information, preventing effective inter-replica coordination. These findings demonstrate that the topology of inter-replica couplings decisively influences search robustness, and provide practical guidelines for model selection and parameter tuning in constrained optimization.

14 pages, 11 figures

References in corpus (27)

• Ising formulations of many NP problems
• Quantum Annealing in the Transverse Ising Model
• Parallel Tempering: Theory, Applications, and New Perspectives
• Parallel Tempering Algorithm for Conformational Studies of Biological Molecules
• Quantum Annealing and Analog Quantum Computation
• Perspectives of quantum annealing: Methods and implementations
• Mathematical Foundation of Quantum Annealing
• Error corrected quantum annealing with hundreds of qubits
• Expanding the horizon of automated metamaterials discovery via quantum annealing
• Quantum annealing correction for random Ising problems
• Quantum Annealing Correction with Minor Embedding
• Temperature and Disorder Chaos in Three-Dimensional Ising Spin Glasses
• Adiabatic quantum optimization with the wrong Hamiltonian
• Chaos and Universality in a Four-Dimensional Spin Glass
• Nested Quantum Annealing Correction
• Quantum annealing correction at finite temperature: ferromagnetic $p$-spin models
• Convergence of Quantum Annealing with Real-Time Schrodinger Dynamics
• Analog Errors in Ising Machines
• Mean Field Analysis of Quantum Annealing Correction
• Towards optimization of photonic-crystal surface-emitting lasers via quantum annealing
• Scalable effective temperature reduction for quantum annealers via nested quantum annealing correction
• Nested Quantum Annealing Correction at Finite Temperature: $p$-spin models
• Mechanism of Slow Relaxation due to Screening Effect in a Frustrated System
• Dynamical process of a bit-width reduced Ising model with simulated annealing
• Quantum optimization with linear Ising penalty functions for customer data science
• Using copies to improve precision in continuous-time quantum computing
• Collaborative filtering based on nonnegative/binary matrix factorization