Benchmarking variational quantum eigensolvers for the square-octagon-lattice Kitaev model

• DOI: 10.48550/arxiv.2108.13375
• Publication Date: 2021-01-01
• Authors: Andy Y. Li; M. Sohaib Alam; Thomas Iadecola; Ammar Jahin; Joshua Job; Dog̃a Murat Kürkç...; Richard Li; Peter P. Orth; A. Barış Özgüler; Gabriel Perdue; Norm M. Tubman
• Keywords: Ansatz; Hamiltonian (control theory)

Quantum spin systems may offer the first opportunities for beyond-classical quantum computations of scientific interest. While general simulation algorithms likely require error-corrected qubits, there be applications interest prior to practical implementation error correction. The variational eigensolver (VQE) is a promising approach finding energy eigenvalues on noisy computers. Lattice models are broad use near-term hardware due sparsity number Hamiltonian terms and possibility matching lattice geometry geometry. Here, we consider Kitaev model hardware-native square-octagon qubit connectivity map, examine efficiently probing its rich phase diagram with VQE approaches. By benchmarking different choices Ansatz states classical optimizers, illustrate advantage mixed optimization using (HVA) potential system's VQE. We further demonstrate HVA circuits Rigetti's Aspen-9 chip mitigation.