Optimal fermion-qubit mappings via quadratic assignment

Mitchell Chiew 1 , Cameron Ibrahim 2 , Ilya Safro 2 , and Sergii Strelchuk 3 · 2025

Read Optimal fermion-qubit mappings via quadratic assignment by Mitchell Chiew 1 , Cameron Ibrahim 2 , Ilya Safro 2 , and Sergii Strelchuk 3 (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-2504-21636.txt.

Introduction

Mitchell Chiew, Cameron Ibrahim, Ilya Safro, and Sergii Strelchuk search for the best way to translate electron behavior into quantum computer language, in a 2025 paper. Simulating electrons is a flagship use for quantum machines, from chemistry to materials. The translation, called a fermion-qubit mapping, sets how many qubits and operations a simulation will cost.

Older mappings force a choice: few qubits or few gates, rarely both. The team recast the search as a quadratic assignment problem, a classic matching puzzle. Their computational methods then hunt mappings that balance both costs under a qubit budget. Better mappings stretch today's small quantum machines further, letting them tackle larger molecules sooner.

Catalog record

CREATORMitchell Chiew 1 , Cameron Ibrahim 2 , Ilya Safro 2 , and Sergii Strelchuk 3
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE52 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2504-21636

Browse the full library · About the Tower of Babel

Full text

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

Optimal fermion-qubit mappings via quadratic assignment Thanks: This work was in part supported by NSF award #2427042 and DARPA ONISQ program. M.L.C. received the support of a Cambridge Australia Allen & DAMTP Scholarship during the preparation of this work. S.S. was supported by the Royal Society University Research Fellowship and “Quantum simulation algorithms for quantum chromodynamics” grant (ST/W006251/1). Mitchell Chiew 1 , Cameron Ibrahim 2 , Ilya Safro 2 , and Sergii Strelchuk 3 Abstract Simulation of fermionic systems is one of the most promising applications of quantum computers. It spans problems in quantum chemistry, high-energy physics and condensed matter. Underpinning the core steps of any quantum simulation algorithm, fermion-qubit mappings translate the fermionic interactions to the operators and states of quantum computers. This translation is highly non-trivial: a burgeoning supply of fermion-qubit mappings has arisen over the past twenty years to address the limited resources of early quantum technology. Previous literature has presented a dichotomy between ancilla-free fermion-qubit mappings, which minimise qubit count, and local encodings, which minimise gate complexity. We present two computational approaches to the construction of general mappings while working with a limited number of qubits, striking a balance between the low-qubit and low-gate demands of present quantum technology. The first method frames the order of fermionic labels as an instance of the quadratic assignment problem to minimize the total and maximum Pauli weights in a problem Hamiltonian. We compare the order-optimized performance of several common ancilla-free mappings on systems of size up to 225 fermionic modes. The second method is a computational approach to incrementally add ancilla qubits to Jordan–Wigner transformations and further reduce the Pauli weights. By adding up to 10 ancilla qubits, we were able to reduce the total Pauli weight by as much as 67% in Jordan–Wigner transformations of fermionic systems with up to 64 modes, outperforming the previous state-of-the-art ancilla-free mappings. Reproducibility: source code and data are available at https://github.com/cameton/QCE_QubitAssignment I Introduction Fig. 1 : Demonstration of the differing localities of fermion-qubit mappings. Ancilla-free mappings such as the Jordan–Wigner transformation produce hopping terms that act nonlocally, i.e. on a number of qubits that increases with system size, while local encodings such as the Bravyi–Kitaev superfast transformation employ ancilla qubits to produce local hopping terms which act on a constant number of qubits. Each vertical line on the qubit register depicts a hopping term in the qubit Hamiltonian, acting with Pauli X , Y or Z matrices on different qubits as indicated by the circles. The study of second-quantized fermionic systems in condensed matter physics and quantum chemistry presents a significant challenge to the computational sciences. Hamiltonian simulation and spectral analysis such as ground-state energy estimation of fermionic systems remains an exponentially scaling problem for classical computers [ 1 , 2 , 3 ] . Quantum computing provides a potential advantage through alternative strategies such as Trotterization [ 4 ] , quantum signal processing [ 5 ] and variational quantum eigensolvers [ 6 ] , limited by the fidelity and scale of contemporary quantum technology. Due to the lack of straightforward efficient representations of fermionic interactions in the standard circuit model, any attempt must begin with a fermion-qubit mapping . The complexity of modelling evolution and extraction of key information about a fermionic system hinges on the complexity of such representations. A central factor is the locality of the qubit Hamiltonian, relating the qubit number or geometry of the support of each term to the task of simulation [ 7 ] or ground-state estimation [ 6 ] . The most well-known mapping, the Jordan–Wigner transformation [ 8 ] , is conceptually intuitive but leads to representations of local fermionic excitations with supports that scale linearly in system size. For example, all qubits on the quantum register may need to be engaged simultaneously to represent an electron hopping between adjacent molecular orbitals. The non-locality of the Jordan–Wigner transformation is seen as an obstacle for near-term devices and has led to the development of many alternatives in recent years. Fermion-qubit mappings in the literature generally fall into two paradigms: ancilla-free mappings which use the minimum number of qubits to describe the fermionic algebra [ 9 , 10 , 11 , 12 , 13 , 14 , 15 , 16 ] and local encodings which employ a number of ancilla qubits that scales linearly with the system size [ 9 , 17 , 18 , 19 , 20 , 21 , 22 , 23 , 24 ] . As seen in Figure 1 , local encodings are designed to produce qubit Hamiltonians that are sums of local Pauli operators for fermionic systems such as the Fermi–Hubbard model [ 25 ] . Meanwhile, ancilla-free mappings have been optimized to produce Pauli representations of fermionic hopping terms with logarithmically scaling support size through the Bravyi–Kitaev [ 9 ] and Ternary Tree transformations [ 12 , 11 ] . However, in the near- and mid-term future for quantum computing, it is not necessarily the case that local encodings are always preferable to ancilla-free mappings [ 26 , 27 ] , with qubit number itself being a scarce resource [ 28 ] competing with the Pauli weights of the qubit Hamiltonian which serve as theoretical complexity parameters. The fine line between ancilla-free mappings and local encodings presents remarkable opportunities for optimization, and there are numerous strategies to drive the costs of fermionic simulation lower avoiding the unwelcome scaling of ancillary qubits with the problem size. In [ 29 ] , authors introduced a strategy to vary the fermionic labeling order in Jordan–Wigner transformations to reduce the total Pauli weight of the square lattice Fermi–Hubbard model; an approach which has now expanded to Ternary Tree transformations [ 30 ] . We also presented an analytic strategy to add two ancilla qubits to the Jordan–Wigner transformation, creating a hybrid mapping with a constant number of ancilla qubits. The addition of ancilla qubits allows cancellation of long strings of Pauli operators in the Hamiltonian terms, allowing a more dramatic reduction in Pauli weights over the ancilla-free version. Our contribution: In this work, we present two new strategies to further reduce the Pauli weight costs of fermion-qubit mappings in the absence of scaling qubit resources. We extend the method in [ 29 ] for optimal fermionic labeling to apply to a prominent class of ancilla-free mapping known as “linear encodings” beyond Jordan–Wigner transformation, applying it in particular to the Bravyi–Kitaev, Ternary Tree and Parity Basis transformations. Through converting the task to a quadratic assignment problem, we also expand the scope of our strategy to fermionic systems with any hopping and interaction pattern, beyond just the square-lattice Fermi–Hubbard model. Our second protocol generalises the two-ancilla qubit strategy from [ 29 ] to a method for incrementally adding any number of qubits to a Jordan–Wigner transformation. By also converting this task to a quadratic assignment problem, we expand the scope of these ‘constant-ancilla mappings’ to Jordan–Wigner transformations for fermionic systems with any hopping and interaction model. In the process, we demonstrate that with sufficient ancilla qubits, we can reduce the Pauli weight associated with the Jordan–Wigner transformation by as much as 67%, improving upon even the best known ancilla-free transformations. Furthermore, we determine optimal or

Read the complete text