quantum annealing

quantum annealing
Name	Quantum annealing
Type	Computational optimization method
Introduced	1990s
Developer	Perimeter Institute for Theoretical Physics; D-Wave Systems
Field	Quantum computing
Related	Adiabatic quantum computation, Simulated annealing, Quantum tunnelling

quantum annealing

Introduction

Quantum annealing is a heuristic optimization technique designed to find low-energy solutions of combinatorial problems by exploiting quantum mechanical effects. It draws on ideas from Richard Feynman, Alan Turing, Paul Dirac, Peter Shor, and early work at institutions such as the Perimeter Institute for Theoretical Physics and Los Alamos National Laboratory. Practical implementations have been pursued by companies and organizations including D-Wave Systems, NASA, Lockheed Martin, Google, and research groups at University of Southern California, MIT, and University of Maryland. Research milestones intersect with events and projects like the Quantum Information Processing conference, ARPA-E, and collaborations involving Rigetti Computing and Microsoft Research.

Principles and Theory

The method is grounded in principles from Erwin Schrödinger's formulation, Werner Heisenberg's matrix mechanics, and concepts developed in John von Neumann's work. It maps combinatorial problems to an energy landscape equivalent to an Ising model used in studies by Lars Onsager and Lev Landau. The process begins with a transverse-field Hamiltonian inspired by Paul Dirac and evolves under an adiabatic schedule related to the Adiabatic theorem as studied by Max Born and J. Robert Oppenheimer. Quantum tunnelling, a phenomenon linked historically to experiments by Leo Esaki and theoretical work of Gamow, allows the system to traverse energy barriers in ways contrasted with classical trajectories examined by Metropolis algorithm origins in Nicholas Metropolis. Connections to classical methods include comparisons to Simulated annealing formalized by S. Kirkpatrick and later theoretical analyses by Edward Farhi and collaborators at Harvard University and Massachusetts Institute of Technology.

Implementation and Hardware

Physical realizations often use superconducting flux qubits based on technology advanced by John Clarke and Frank Wilhelm, and manufacturing techniques related to IBM's superconducting efforts and NIST fabrication facilities. Major hardware platforms have been produced by D-Wave Systems and tested in collaborations with Los Alamos National Laboratory, NASA Ames Research Center, and Google Quantum AI. Alternative proposals involve trapped ions developed in laboratories such as University of Innsbruck and NIST Boulder, and proposals leveraging spin qubits from groups at University of New South Wales and University of California, Santa Barbara. Control electronics draw on microwave engineering traditions found in Bell Labs and cryogenic infrastructure from projects at CERN. Benchmarking and scaling analyses reference standards and competitions like those organized by IEEE and collaborations with National Science Foundation programs.

Applications and Use Cases

Use cases span optimization challenges in industries and research institutions including routing and logistics problems explored by UPS and Airbus, portfolio optimization investigated by Goldman Sachs and JPMorgan Chase, and machine learning tasks studied by teams at Google, Microsoft Research, and DeepMind. Scientific applications include protein folding problems linked to studies at Cold Spring Harbor Laboratory and Broad Institute, materials design efforts tied to Lawrence Livermore National Laboratory and Argonne National Laboratory, and scheduling problems tackled in projects with NASA and DARPA. Case studies involve collaborations with Volkswagen for traffic optimization and investigations with Fujitsu and Hitachi on supply-chain scenarios. Academic explorations connect to work at ETH Zurich, University of Cambridge, and Princeton University.

Performance, Limitations, and Comparisons

Empirical comparisons often reference classical algorithms and architectures from Google DeepMind, IBM Research, and teams using high-performance clusters at Oak Ridge National Laboratory and Los Alamos National Laboratory. Debates over quantum speedup cite milestone demonstrations involving Google and criticisms by groups at University of Toronto and University of Oxford. Limitations include decoherence issues studied by Yasunobu Nakamura and thermalization effects analyzed in experiments at Yale University and University of Waterloo. Embedding constraints relate to graph-minor theorems from Kurt Gödel-adjacent combinatorics and mappings that reference foundational work by Richard Karp on NP-completeness. Comparative assessments examine hybrid approaches combining classical solvers from Zuse Institute Berlin and quantum processors from D-Wave Systems.

Algorithms and Software Ecosystem

Algorithmic frameworks build on Ising and quadratic unconstrained binary optimization (QUBO) formulations used by software stacks developed at D-Wave Systems, Rigetti Computing, Xanadu Quantum Technologies, and open-source projects supported by GitHub and communities around Qiskit at IBM and Cirq at Google. Toolchains integrate with cloud services from Amazon Web Services, Microsoft Azure, and Google Cloud Platform to enable access for teams at Stanford University, Cornell University, and University of California, Berkeley. Research on error suppression, anneal-path optimization, and reverse annealing involves authors from Perimeter Institute, University of British Columbia, and University of Tokyo. Educational and outreach efforts appear in workshops at International Conference on Quantum Technologies and summer schools hosted by Perimeter Institute for Theoretical Physics and Institute for Quantum Computing.

Category:Quantum computing