Thermalization, freeze-out and noise: deciphering experimental quantum annealers

Abstract

By contrasting the performance of two quantum annealers operating at different temperatures, we address recent questions related to the role of temperature in these devices and their function as `Boltzmann samplers'. Using a method to reliably calculate the degeneracies of the energy levels of large-scale spin-glass instances, we are able to estimate the instance-dependent effective temperature from the output of annealing runs. Our results show that the output distributions of the annealers do not in general correspond to classical Boltzmann distributions. For the small fraction of the instances for which classical thermalization takes place, we find that the effective temperatures are significantly higher than the physical temperatures. Our results in this regime provide further evidence for the `freeze-out' picture in which the output is sampled from equilibrium distributions determined at a point in time earlier in the quantum annealing process. We also find that the effective temperatures at different programming cycles fluctuate greatly, with the effect worsening with problem size. We discuss the implications of our results for the design of future quantum annealers to act as efficient Boltzmann samplers and for the programming of such annealers.

12 pages, 16 figures. v2: updated to published version

References in corpus (6)

• Minor-embedding in adiabatic quantum computation: II. Minor-universal graph design
• Hiding Quiet Solutions in Random Constraint Satisfaction Problems
• Temperature and Disorder Chaos in Three-Dimensional Ising Spin Glasses
• Temperature scaling law for quantum annealing optimizers
• Disorder chaos in spin glasses
• Advantages of Unfair Quantum Ground-State Sampling

Cited by in corpus (27)

• Quantum Annealing for Industry Applications: Introduction and Review
• A Hybrid Solution Method for the Capacitated Vehicle Routing Problem Using a Quantum Annealer
• Probing the Universality of Topological Defect Formation in a Quantum Annealer: Kibble-Zurek Mechanism and Beyond
• Power of Pausing: Advancing Understanding of Thermalization in Experimental Quantum Annealers
• Scaling and diabatic effects in quantum annealing with a D-Wave device
• A deceptive step towards quantum speedup detection
• Improving quantum annealing of the ferromagnetic $p$-spin model through pausing
• Programmable Quantum Annealers as Noisy Gibbs Samplers
• Quantum annealing for systems of polynomial equations
• High-quality Thermal Gibbs Sampling with Quantum Annealing Hardware
• Optimal quantum annealing: A variational shortcut to adiabaticity approach
• Improved Boltzmann machines with error corrected quantum annealing
• Quantum-assisted associative adversarial network: Applying quantum annealing in deep learning
• Comparing relaxation mechanisms in quantum and classical transverse-field annealing
• Scalable effective temperature reduction for quantum annealers via nested quantum annealing correction
• Equation Planting: A Tool for Benchmarking Ising Machines
• Search range in experimental quantum annealing
• Signatures of Open and Noisy Quantum Systems in Single-Qubit Quantum Annealing
• Perils of Embedding for Sampling Problems
• Estimating the Density of States of Frustrated Spin Systems
• Classical simulation and theory of quantum annealing in a thermal environment
• Noise amplification at spin-glass bottlenecks of quantum annealing: a solvable model
• Advantage of pausing: parameter setting for quantum annealers
• Boltzmann sampling with quantum annealers via fast Stein correction
• Neural-network-assisted Monte Carlo sampling trained by Quantum Approximate Optimization Algorithm
• Magnetic Hysteresis Experiments Performed on Quantum Annealers
• Boltzmann Sampling of Frustrated J1 - J2 Ising Models with Programmable Quantum Annealers