Comparison of variational quantum eigensolvers in light nuclei

Miquel Carrasco-Codina, Emanuele Costa, Antonio Márquez Romero, Javier Menéndez, Arnau Rios · 2025

Read Comparison of variational quantum eigensolvers in light nuclei by Miquel Carrasco-Codina, Emanuele Costa, Antonio Márquez Romero, Javier Menéndez, Arnau Rios (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-2507-13819.txt.

Introduction

Miquel Carrasco-Codina and co-authors compare two quantum algorithms for computing the ground states of light atomic nuclei. Their paper appeared in 2025. The algorithms, called variational quantum eigensolvers, are leading candidates for near-term quantum machines. The team simulated them on classical computers for nuclei from helium-6 to boron-10, using the nuclear shell model. They also propose a new score for the quantum resources each run consumes.

Both methods worked best when starting from simple single-configuration states. One method, ADAPT, used fewer resources for nuclei near closed shells. The other, UCC, was leaner toward mid-shell. The authors frame the study as groundwork for fair benchmarks of quantum algorithms in nuclear physics, before real hardware takes over.

Catalog record

CREATORMiquel Carrasco-Codina, Emanuele Costa, Antonio Márquez Romero, Javier Menéndez, Arnau Rios
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE44 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2507-13819

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Comparison of variational quantum eigensolvers in light nuclei Miquel Carrasco-Codina, Emanuele Costa, Antonio Márquez Romero, Javier Menéndez, Arnau Rios Abstract Quantum computing is one of the most promising technologies of the near future, and the simulation of quantum many-body systems is a natural application. In this work, we present classical simulations of the ground states of light atomic nuclei within the p p shell, from 6 He to 10 B, calculated within the nuclear shell model. We compare the performance of two leading variational quantum eigensolver algorithms: the Unitary Coupled Cluster (UCC) and the Adaptive Derivative-Assembled Pseudo-Trotter (ADAPT) methods, introducing a new metric to quantify the use of quantum resources in each simulation. We find that Slater determinants are the most useful reference states for both approaches. Our analysis suggests that ADAPT is more efficient for nuclei close to magic numbers, while UCC tends to require fewer resources toward the mid shell. This work lays the groundwork for robust benchmarking of quantum algorithms in nuclear structure studies. I Introduction The simulation of quantum many-body systems in classical devices is challenged by the exponential growth of the dimension of the Hilbert space and the entanglement correlations between particles Feynman 1982 . For this reason, the study of nuclear structure is limited by a lack of precise simulations, especially in heavy nuclei. The development of new techniques to address these systems can enhance the understanding of strong Mukha et al. 2006 ; Taniuchi et al. 2019 ; Butler et al. 2019 ; Tsunoda et al. 2020 ; Hu et al. 2022 ; Kondo et al. 2023 ; Gade et al. 2025 and weak Hinke et al. 2012 ; Gysbers et al. 2019 interactions in nuclei, as well as phenomena such as entanglement correlations among nucleons Robin et al. 2021 ; Johnson and Gorton 2023 ; Tichai et al. 2023 ; Pérez-Obiol et al. 2023a ; Brökemeier et al. 2025 . Moreover, accurate simulations of nuclear structure are essential to tackle fundamental questions in astrophysics Cowan et al. 2021 ; Aalbers et al. 2022 ; Huth et al. 2022 and particle physics Agostini et al. 2023 ; Engel et al. 2013 . Quantum computing offers an alternative to classical simulations for solving many-body problems, exploiting the quantum-mechanical behavior of its fundamental elements, qubits Nielsen and Chuang 2010 ; DiVincenzo 2000 . The concept of a computer that simulates complex many-body systems by encoding information in quantum particles has been established for decades Feynman 1982 , but interest in this approach has surged recently due to rapid advancements in quantum-computing devices Preskill 2018 . However, the inherently sensitive nature of qubits still causes several limitations Preskill 2018 ; Bravyi et al. 2024 ; Hornibrook et al. 2015 . These challenges have driven the development of hybrid algorithms, which optimize computations by distributing tasks between quantum and classical devices, thus mitigating the high error rates associated with quantum circuits in the current noisy intermediate-scale quantum (NISQ) era. Various quantum approaches have been proposed to study nuclear structure Ayral et al. 2023 . Some algorithms rely on projection techniques Stetcu et al. 2023 ; Li et al. 2024 ; Rule et al. 2024 ; Yoshida et al. 2024 while others propose annealing protocols Du et al. 2021 ; Costa et al. 2024 . Many works focus on hybrid algorithms for many-body simulations called Variational Quantum Eigensolvers (VQEs) Peruzzo et al. 2014 ; Cerezo et al. 2021 ; Tilly et al. 2022 , initially using the Unitary Coupled Cluster (UCC) approach Dumitrescu et al. 2018 ; Kiss et al. 2022 ; Stetcu et al. 2022 ; Sarma et al. 2023 . More recently, also the Adaptive Derivative-Assembled Pseudo-Trotter (ADAPT) Romero et al. 2022 ; Pérez-Obiol et al. 2023b ; Zhang et al. 2024 ; Pérez-Obiol et al. 2024 ; Zhang and Lacroix 2025 ; Singh et al. 2025 and other problem-inspired, hardware-efficient algorithms have been employed Bhoy and Stevenson 2024 ; Singh et al. 2025 . Promising developmemts related to magic and non-stabilizerness have been recently explored in nuclear systems too Brökemeier et al. 2025 ; Robin 2025 . VQEs are especially appealing since they have been applied to a variety of many-body systems, including quantum chemistry Grimsley et al. 2019 ; Sapova and Fedorov 2022 ; McArdle et al. 2020 , Ising Cervera-Lierta 2018 and Lipkin-Meshkov-Glick models Cervia et al. 2021 ; Harsha et al. 2018 ; Faba et al. 2022 ; Robin and Savage 2023 ; Baid et al. 2024 , and superfluids Lacroix 2020 ; Ruiz Guzman and Lacroix 2022 . In the context of nuclear structure, while various strategies within the same VQE framework have been compared in detail for UCC Kiss et al. 2022 , an equal-footing evaluation of the performance of various VQEs applied to the same nuclei is still missing. In this work, we employ classical simulations to analyze the performance in solving nuclear ground states of two state-of-the-art VQEs: UCC Anand et al. 2022 and ADAPT Grimsley et al. 2019 . We apply these algorithms to the same nuclei within a common many-body framework: the nuclear shell model (NSM). Our aim is to evaluate the quantum resources for each method, thereby identifying the best-suited algorithm for each NSM simulations. To achieve this, we introduce a new metric based on the number of quantum operations required to simulate a given system, comparing the performance of both VQEs across isotopes in the p shell. Our comparisons also employ different reference states and randomized initial parameters, in an effort to obtain the most general possible method-independent conclusions. This approach provides us with unique insights into the suitability of quantum algorithms for studying nuclear structure. Figure 1: Left panel: Single-particle states in the lowest nuclear shells. Each state is labeled by its quantum numbers nl_{j} ( y axis) and m ( x axis). The index corresponds to a qubit in the VQE circuit, according to a Jordan-Wigner mapping Jordan and Wigner 1928 . Right panel: Nuclear chart for the lightest nuclei, based on the number of neutrons, N , ( x axis) and the number of protons, Z , ( y axis). The p shell is displayed in red and nuclei simulated in this work are highlighted with grey circles. This paper is structured as follows: Sec. II offers a brief introduction to the NSM and presents the nuclei studied in our work. We then discuss the details of the UCC and ADAPT methods in Sec. III . The results of our comparison are provided in Sec. IV . Finally, we draw our conclusions and outline future research in Sec. V . II Nuclear Shell Model The NSM describes nuclear structure based on a valence space comprised of orbitals with similar energies that form a shell . These shells contain quantum single-particle states for protons and neutrons, the constituents of the nucleus. When fully occupied, shells lead to the so-called magic numbers Mayer 1949 ; Haxel et al. 1949 , which are associated to large energy gaps between orbitals. Within the NSM Caurier et al. 2005 , the complexity of full nuclear systems is simplified by assuming that nucleons in the valence space can describe all the nuclear structure of each isotope in

Read the complete text