2017/08/31 by Kohji Nishimura, Hidetoshi Nishimori
Computer Science · Engineering · Mathematics · Physics and Astronomy · #Algorithm #Amplitude #Condensed matter physics #Engineering #Ground state #Hamiltonian (control theory) #Hamming distance #Mathematical optimization #Mathematics #Physics #Quantum #Quantum Computing Algorithms and Architecture #Quantum and electron transport phenomena #Quantum annealing #Quantum computer #Quantum electrodynamics #Quantum fluctuation #Quantum many-body systems #Quantum mechanics #Quantum noise #Statistical physics #Transverse field #Transverse plane #quant-ph
paper · pdf · doi:10.1103/physreva.96.042310
published as Phys. Rev. A 96, 042310 (2017) · 8 pages, 5 figures
openalex created_date 2017/08/08 · arxiv created 2017/10/10 · openalex publication_date 2017/10/10 · arxiv updated 2017/10/12 · openalex updated_date 2026/08/05
We study the problem to infer the original ground state of a spin-glass Hamiltonian out of the information from the Hamiltonian with interactions deviated from the original ones. Our motivation comes from quantum annealing on a real device in which the values of interactions are degraded by noise. We show numerically for quasi-one-dimensional systems that the Hamming distance between the original ground state and the inferred spin state is minimized when we stop the process of quantum annealing before the amplitude of the transverse field reaches zero in contrast to the conventional prescription. This result means that finite quantum fluctuations compensate for the effects of noise, at least, to some extent. Analytical calculations using the infinite-range mean-field model support our conclusion qualitatively.