Your browser doesn't support javascript.
loading
General limit to thermodynamic annealing performance.
Luo, Yutong; Zhen, Yi-Zheng; Liu, Xiangjing; Ebler, Daniel; Dahlsten, Oscar.
Afiliación
  • Luo Y; Blackett Laboratory, Imperial College London, London SW7 2AZ, United Kingdom.
  • Zhen YZ; Department of Physics, Southern University of Science and Technology, Shenzhen 518055, China.
  • Liu X; Hefei National Research Center for Physical Sciences at the Microscale and School of Physical Sciences, University of Science and Technology of China, Hefei 230026, China.
  • Ebler D; Shanghai Research Center for Quantum Science and CAS Center for Excellence in Quantum Information and Quantum Physics, University of Science and Technology of China, Shanghai 201315, China.
  • Dahlsten O; Department of Physics, Southern University of Science and Technology, Shenzhen 518055, China.
Phys Rev E ; 108(5): L052105, 2023 Nov.
Article en En | MEDLINE | ID: mdl-38115520
ABSTRACT
Annealing has proven highly successful in finding minima in a cost landscape. Yet, depending on the landscape, systems often converge towards local minima rather than global ones. In this Letter, we analyze the conditions for which annealing is approximately successful in finite time. We connect annealing to stochastic thermodynamics to derive a general bound on the distance between the system state at the end of the annealing and the ground state of the landscape. This distance depends on the amount of state updates of the system and the accumulation of nonequilibrium energy, two protocol and energy landscape-dependent quantities which we show are in a trade-off relation. We describe how to bound the two quantities both analytically and physically. This offers a general approach to assess the performance of annealing from accessible parameters, both for simulated and physical implementations.
Texto completo
Añadir a Mi BVS
Imprimir
XML
PubMed Links
Buscar en Google

Texto completo: 1 Colección: 01-internacional Base de datos: MEDLINE Idioma: En Revista: Phys Rev E Año: 2023 Tipo del documento: Article País de afiliación: Reino Unido Pais de publicación: Estados Unidos
Texto completo
Añadir a Mi BVS
Imprimir
XML
PubMed Links
Buscar en Google

Texto completo: 1 Colección: 01-internacional Base de datos: MEDLINE Idioma: En Revista: Phys Rev E Año: 2023 Tipo del documento: Article País de afiliación: Reino Unido Pais de publicación: Estados Unidos