Simulating strongly interacting Hubbard chains with the variational Hamiltonian ansatz on a quantum computer - CNRS - Centre national de la recherche scientifique Accéder directement au contenu
Article Dans Une Revue Physical Review Research Année : 2022

Simulating strongly interacting Hubbard chains with the variational Hamiltonian ansatz on a quantum computer

Résumé

Hybrid quantum-classical algorithms have been proposed to circumvent noise limitations in quantum computers. Such algorithms delegate only a calculation of the expectation value to the quantum computer. Among them, the variational quantum eigensolver has been implemented to study molecules and condensed matter systems on small size quantum computers. Condensed matter systems described by the Hubbard model exhibit a rich phase diagram alongside exotic states of matter. In this paper we try to answer the question: How much of the underlying physics of a 1D Hubbard chain is described by a problem-inspired variational Hamiltonian ansatz in a broad range of parameter values? We start by probing how much the solution increases fidelity with increasing ansatz complexity. Our findings suggest that even low fidelity solutions capture energy and number of doubly occupied sites well, while spin-spin correlations are not well captured even when the solution is of high fidelity. Our powerful simulation platform allows us to incorporate a realistic noise model and shows a successful implementation of noise-mitigation strategies—postselection and the Richardson extrapolation. Finally, we compare our results with an experimental realization of the algorithm on IBM Quantum's ibmq_quito device.

Dates et versions

hal-03469437 , version 1 (07-12-2021)

Identifiants

Citer

Baptiste Anselme Martin, Pascal Simon, Marko J. Rančić. Simulating strongly interacting Hubbard chains with the variational Hamiltonian ansatz on a quantum computer. Physical Review Research, 2022, 4 (2), pp.023190. ⟨10.1103/PhysRevResearch.4.023190⟩. ⟨hal-03469437⟩
121 Consultations
0 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More