Variational simulation of the Lipkin-Meshkov-Glick model on a neutral atom quantum computer
R. Chinnarasu, C. Poole, L. Phuttitarn, A. Noori, T. M. Graham, S. N. Coppersmith, A. B. Balantekin, M. Saffman · 2025
Read Variational simulation of the Lipkin-Meshkov-Glick model on a neutral atom quantum computer by R. Chinnarasu, C. Poole, L. Phuttitarn, A. Noori, T. M. Graham, S. N. Coppersmith, A. B. Balantekin, M. Saffman (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2501-06097.txt.
Introduction
Chinnarasu leads a team of eight on this paper from two thousand twenty-five. It runs a classic nuclear model, the Lipkin-Meshkov-Glick model, on a neutral atom quantum computer with a variational solver.
Ground state energies are found for spin systems of up to fifteen spins. Two encodings are tried. In one, each spin takes its own qubit. In the other, a Gray code packs the spins into far fewer qubits, a count that grows with the log of the spin count. The lean code, plus a noise stretch trick that cleans the read, lands much closer to the exact energies. The work shows a small quantum machine can host a real physics model with care.
Catalog record
| CREATOR | R. Chinnarasu, C. Poole, L. Phuttitarn, A. Noori, T. M. Graham, S. N. Coppersmith, A. B. Balantekin, M. Saffman |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 68 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2501-06097
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Variational simulation of the Lipkin-Meshkov-Glick model on a neutral atom quantum computer R. Chinnarasu, C. Poole, L. Phuttitarn, A. Noori, T. M. Graham, S. N. Coppersmith, A. B. Balantekin, M. Saffman Abstract We simulate the Lipkin-Meshkov-Glick (LMG) model using the Variational-Quantum-Eigensolver (VQE) algorithm on a neutral atom quantum computer. We test the ground-state energy of spin systems with up to 15 spins. Two different encoding schemes are used: an individual spin encoding where each spin is represented by one qubit, and an efficient Gray code encoding scheme which only requires a number of qubits that scales with the logarithm of the number of spins. This more efficient encoding, together with zero noise extrapolation techniques, is shown to improve the fidelity of the simulated energies with respect to exact solutions. I Introduction Scaling up of digital quantum processors for running large circuits requires establishing connectivity between distant qubits. Universal paradigms for enabling arbitrary connectivity rely on either moving quantum states across the processor with a sequence of swap gates [ 1 ] or using teleportation [ 2 ] . Both of these approaches incur additional resource overhead as long swap chains increase circuit depth, and efficient teleportation depends on the availability of high-fidelity resource states. Physical motion of qubits is a third alternative that was originally introduced for the trapped ion platform [ 3 , 4 , 5 ] , and has been successfully implemented with neutral atom qubits for logical processing [ 6 , 7 , 8 ] , as well as in atomic tweezer clocks [ 9 ] . In this work, we demonstrate several innovations for quantum circuit operation with neutral atom qubits. Using both atom transport for dynamical modification of the connectivity map and rapid optical beam scanning for site-specific optical control we simulate the Lipkin-Meshkov-Glick(LMG) model using the Variational-Quantum-Eigensolver(VQE) algorithm. We show that application of efficient encoding strategies together with noise mitigation techniques [ 10 ] enables finding ground-state energies with an accuracy beyond that supported by the intrinsic physical gate fidelities. The LMG model [ 11 ] was introduced in the 1960s as an exactly solvable test bed to benchmark classical computer simulations of many-body physics. There have been several recent quantum simulations of the LMG model. A quantum algorithm to find the ground-state energies of a two-particle system was implemented on the IBM Quantum Experience in [ 12 ] . The excited states of the LMG model were studied in [ 13 ] , where the energy spectrum of up to four particles was obtained. Several versions of hybrid quantum-classical approaches to the LMG model are available [ 14 , 15 , 16 , 17 , 18 ] as well as extensions to the LMG model by inclusion of a pairing term in the Hamiltonian [ 19 ] . Due to the exactly solvable nature of the LMG model Hamiltonian, it can be used to verify eigenenergies predicted on NISQ era quantum processors. In this work, we demonstrate simulation of up to 15 spin versions of the LMG Hamiltonian on a neutral atom quantum computer. Ground-state energies found with the VQE algorithm are compared with exact solutions of the LMG Hamiltonian. The remainder of the paper is organized as follows. In Sec. II we present the individual spin and Gray code encoding schemes used for VQE simulation of ground-state energies. In Sec. III the experimental approach is described. Results from both encoding schemes are presented in Sec. IV . The results are summarized in Sec. V and additional technical details are provided in Appendices. II Model Hamiltonian and spin encoding The LMG model is a nuclear shell model consisting of N fermions distributed among two N -fold degenerate levels [ 11 ] . In terms of collective quasi-spin operators J_{0},J_{+},J_{-} , the model Hamiltonian is given by [ 11 , 12 ] {\tilde{H}}=\epsilon J_{0}+\frac{1}{2}{\tilde{V}}\left(J_{+}^{2}+J_{-}^{2}\right), (1) where \epsilon is the energy level separation and \tilde{V} is the strength of a tunable interaction which flips pairs of particles between the two energy levels. In the rest of this paper, we work with the dimensionless Hamiltonian H\equiv{\tilde{H}}/\epsilon with V\equiv{\tilde{V}}/\epsilon . The LMG model is nontrivial but also exactly solvable, making it attractive for use in testing approximation methods in many-body physics. The VQE algorithm [ 20 ] leverages the variational principle to estimate the ground-state energy of a Hamiltonian. Compared to Quantum Phase Estimation which in general requires deep circuits, the VQE algorithm is potentially advantageous for implementation on NISQ processors without full error correction. Figure 1: A pictorial schematic of the three-qubit VQE circuit implementation using optical beam scanning and atom transport. The trap configurations for the different circuit operations are shown in green, yellow, blue, and orange frames. a) Shows a circuit diagram for three qubit VQE state preparation using Gray code encoding. b) An atom spacing of 16~\mu\rm m is used for initialization and single qubit operations. c) The atoms are moved to a horizontal spacing of 2.5~\mu\rm m , and Rydberg light is focused on the atom pair in the top row to implement a \sf CZ gate between q_{1},q_{2} . The atoms are then moved apart for local {\sf R}_{z} correction pulses. d) The atoms are moved to a vertical spacing of 2.5~\mu\rm m , and Rydberg light is focused on the atom pair in the left column to implement a \sf CZ gate between q_{0},q_{1} . The atoms are then moved apart for local {\sf R}_{z} correction pulses. e) The atoms are placed on a grid with 6\times 9~\mu\rm m spacing for readout. This rectangular spacing was chosen to minimize heating from beating between AOD frequencies due to trap crosstalk from scattering and reflections in the trap beam path. To implement the VQE algorithm, an ansatz wave function \ket{\psi} is selected as a function of a number of variational parameters. The expectation value of the Hamiltonian \bra{\psi}H\ket{\psi} is evaluated and the variational parameters are adjusted to minimize the observed energy. To implement the algorithm on a quantum computer, the Hamiltonian and trial wave function must be mapped onto the computational basis of the quantum register. To evaluate the expectation value of the Hamiltonian, we decompose it as H=\sum_{i}a_{i}P_{i} , where P_{i} are the products of Pauli operators and a_{i}={\rm Tr}(HP_{i}^{\dagger})/{\rm dim}(H) . By linearity of expectation values, \bra{\psi}H\ket{\psi}=\sum_{i}a_{i}\bra{\psi}P_{i}\ket{\psi} . For the results presented in this paper we have set V=1 when evaluating the weights a_{i} . We can measure a Pauli operator by measuring each qubit in the basis indicated by the operator. For example, to measure \sf ZXZ , we measure the middle qubit in the \sf X basis and the others in the \sf Z basis. Only Pauli operators with non-zero weight need to be measured, and operators can be grouped into sets where any pair of operators in the same set commute and therefore can be measured simultaneously with the same final basis transformation. One unique circuit must be run for each commuting set for each value of the variational parameters. Each unique circuit was allocated 400 bitstring measurements to compute the expectation value of the associated observables. We used two different encoding schemes for VQE simulation of the LMG model. A naive encoding scheme represents each spin