Berlin 2024 – scientific programme
Parts | Days | Selection | Search | Updates | Downloads | Help
QI: Fachverband Quanteninformation
QI 24: Verification and Benchmarking of Quantum Systems
QI 24.14: Talk
Thursday, March 21, 2024, 13:15–13:30, HFT-TA 441
Proposed method to produce large multipartite nonlocality and to benchmark quantum computers — •Jan Lennart Bönsel1, Otfried Gühne1, and Adán Cabello2 — 1Universität Siegen, Germany — 2Universidad de Sevilla, Spain
Nonlocality is a characteristic of quantum mechanics that does not occur in local realistic models. The violation of a Bell inequality can thus be used to verify nonclassicality.
In this contribution, we address the problem of producing n-partite Bell nonlocality with a very large number of qubits n. The main limiting factors are: (i) A restricted connectivity of the quantum computer might not allow an easy preparation of arbitrary entangled states. (ii) Noise limits the observable violation, and (iii) the number of different combinations of local measurements grows exponentially with n. Here, we point out that, for a given two-qubit connectivity, there is a particular entangled state that can be efficiently prepared, effectively overcoming (i). For the extreme cases of connectivity, we consider the GHZ and the linear cluster state. For these states, there exist n-partite Bell inequalities for which the resistance to white noise increases exponentially with n, which limits the impact of (ii). Finally, we introduce a method to address (iii). As a result, we show how to produce n-partite Bell nonlocality with unprecedented large n using quantum computers under the assumption that there is no qubit communication after the local measurements. The n-partite nonlocality allows to quantify the nonclassicality and thus benchmark quantum computers. We test our approach on the IBM Quantum platform.
Keywords: Bell inequality; Nonlocality; Quantum computer; Benchmark; Graph states