QCE20 Workshop on
Tuning Strategies for Quantum Annealing


Date and Time

  • Wed, Oct 14, 2020
  • 10:45─16:45 Mountain Time (MDT) — UTC-6


  • Erica Grant, Oak Ridge National Laboratory (ORNL) & University of Tennessee – Organizing Committee Chair
  • Catherine McGeoch, D-Wave Systems – Organizing Committee Chair


Quantum annealing is a metaheuristic implemented in quantum hardware, which utilizes quantum properties like superposition to find the global minimum solution of problems that can be formulated as Hamiltonians. As the capabilities of quantum annealers improve, it is important to measure the performance of these devices and develop strategies to refine the quality of solutions so that we can inform hardware and algorithm development, better understand how these parameters influence device physics, and discover which applications may demonstrate quantum advantage over classical heuristics. There are three categories of tuning strategies: (1) Pre-processing involves the preparation of the problem to be solved, including the formulation of the Hamiltonian, weights applied to the Hamiltonian (e.g., Lagrange multipliers), and the embedding of the problem onto the hardware; (2) Modifying annealing parameters involves strategies for optimizing and tuning control options such as qubit initialization, reverse anneal, pause times, and anneal offsets; and (3) Post-processing considers manipulation of the solutions returned by the quantum annealer. Hybrid quantum-classical solution approaches may combine tuning strategies. For example, one may implement a pre­processing method to find a best-guess initial state before annealing, combined with a post-processing method using local search to improve solutions. Hybrid approaches can also be applied to solve larger-than-chip inputs.



  • Time: 11:15 – 11:45
  • Catherine McGeoch, D-Wave Systems & Erica Grant, Oak Ridge National Laboratory

Overcoming analog errors in quantum annealing

  • Time: 11:15 – 11:45
  • Daniel Lidar, University of Southern California
  • Abstract: Quantum annealing has the potential to provide a speedup over classical algorithms in solving optimization problems. Just as for any other quantum device, suppressing Hamiltonian control errors will be necessary before quantum annealers can achieve speedups. Such analog control errors are known to lead to J-chaos, wherein the probability of obtaining the optimal solution, encoded as the ground state of the intended Hamiltonian, varies widely depending on the control error. Here, we show that J-chaos causes a catastrophic failure of quantum annealing, in that the scaling of the time-to-solution metric becomes worse than that of a deterministic (exhaustive) classical solver. We demonstrate this empirically using random Ising spin glass problems run on the D-Wave 2X and D-Wave 200Q devices. We then proceed to show that this doomsday scenario can be mitigated using a simple error suppression and correction scheme known as quantum annealing correction (QAC). By using QAC, the time-to-solution scaling of the same D-Wave devices is improved to below that of the classical upper bound, thus restoring hope in the speedup prospects of quantum annealing.

Simulating low-dimensional systems effectively using a quantum annealer

  • Time: 11:45 – 12:15
  • Andrew King, D-Wave Systems
  • Abstract: The simulation of condensed matter systems is a promising application of quantum annealing processors.  Systems of interest usually involve highly degenerate state spaces, and maintaining this degeneracy is a challenge in quantum annealing due to limitations in device homogeneity and environmental control.  Here we discuss several methods for fine-tuning quantum annealers, and give case studies in how these examples improve simulation results.


  • Time: 12:15—13:00pm

Improving the performance of logical qubits by topology compensation

  • Time: 13:00 – 13:30
  • Jack Raymond, D-Wave Systems
  • Abstract: Optimization or sampling of arbitrary pairwise Ising models, in a quantum annealing protocol of constrained interaction topology, can be enabled by a minor-embedding procedure. The logical problem of interest is transformed to a physical (device programmable) problem, where one binary variable is represented by a logical qubit consisting of multiple physical qubits. Inhomogeneities in coupling strength between logical qubits arising from minor embedding are shown to be mitigated by efficient strategies accounting for logical qubit topology.

Parameter optimization for improving quantum annealing accuracy

  • Time: 13:30 – 14:00
  • Hristo Djidjev, Los Alamos National Laboratory
  • Abstract: Current D-Wave quantum annealers offer new features that allow users to set values for a number of parameters for controlling the anneal process. Such parameters include anneal time, qubit offsets, and complex anneal schedules using dozens of points including reverse annealing, pausing, and quenching. Previous research has suggested that using such features can significantly improve the accuracy of the solutions found by D-Wave, but choosing the right combination of parameter values is challenging given the large dimension of the search space. We are reporting on our experience in parameter optimization using techniques such as genetic algorithms and Bayesian optimization.

Pre- and post-processing in quantum-computational hydrologic inverse analysis

  • Time: 14:00 – 14:30
  • John Golden, Los Alamos National Laboratory
  • Abstract: Certain hydrological inverse problems, such as determining the composition of an aquifer from pressure readings can be solved on a quantum annealer. However, quantum annealer performance suffers when solving problems where the aquifer is composed of materials with vastly different permeability, which is often encountered in practice. In this talk, we show why this regime is difficult and use several pre- and post-processing tools to address these issues. These processing techniques include the roof dual, D-Wave’s ‘optimization’ post-processing, and the Multi-Qubit Correction algorithm recently introduced by Dorband.


  • 14:30 – 15:15

Panel discussion with all speakers:

  • Time: 15:15 – 16:45
  • Moderator – Erica Grant, Oak Ridge National Laboratory

Contact Us

For more information about the Tuning Strategies for Quantum Annealing workshop, contact Erica Grant egrant8@vols.utk.edu, Oak Ridge National Laboratory (ORNL) & the University of Tennessee or Catherine McGeoch cmcgeoch@dwavesys.com, D-Wave Systems.