Quantum Singular Value Transformation for Solving the Time-Dependent Maxwell’s Equations

2025

Read Quantum Singular Value Transformation for Solving the Time-Dependent Maxwell’s Equations (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-2507-09686.txt.

Introduction

Gal Shaviner and co-authors build a quantum method for solving the time-dependent Maxwell equations, the rules that govern electric and magnetic fields. Their 2025 paper uses quantum singular value transformation, a broad tool for working with matrices on a quantum machine. The method encodes the problem in blocks and applies a polynomial stand-in for the inverse function.

The team ran the method in the PennyLane simulator on a one-dimensional test case with a Gaussian pulse. The circuit used nine qubits for a grid of 128 points, and the quantum answer matched the classical one with very high fidelity. The authors note that reading out a full quantum state is not feasible on real hardware yet. Quantum computing readers get both a working demo and its stated limits.

Catalog record

YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE30 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2507-09686

Browse the full library · About the Tower of Babel

Full text

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

Quantum Singular Value Transformation for Solving the Time-Dependent Maxwell’s Equations Abstract This work presents a quantum algorithm for solving linear systems of equations of the form 𝐀 ​ ∂ 𝐟 ∂ 𝐱 = 𝐁𝐟 \mathbf{A}{\frac{\mathbf{\partial f}}{\mathbf{\partial x}}}=\mathbf{B}\mathbf{f} , based on the Quantum Singular Value Transformation (QSVT). The algorithm uses block-encoding of A A and applies an 21st-degree polynomial approximation to the inverse function f ⁡ ( x ) = 1 / x f(x)=1/x , enabling relatively shallow quantum circuits implemented on 9 qubits, including two ancilla qubits, corresponding to a grid size of 128 points. Phase angles for the QSVT circuit were optimized classically using the Adagrad gradient-based method over 100 iterations to minimize the solution cost. This approach was simulated in PennyLane and applied to solve a 1D benchmark case of Maxwell’s equations in free space, with a Gaussian pulse as the initial condition, where the quantum-computed solution showed high fidelity of more than 99.9 % when compared to the normalized classical solution. Results demonstrate the potential of QSVT-based linear solvers on simulators with full quantum state access. However, practical hardware implementations face challenges because accessing the complete quantum state is infeasible. This limitation restricts applicability to cases where only O ⁡ ( p ​ o ​ l ​ y ​ ( n ) ) O({poly}(n)) observables are needed. These findings highlight both the promise and current limitations of using quantum algorithms, such as QSVT, to solve linear systems of equations, and they point to the need for the development of measurement-efficient algorithms for near-term quantum devices. Keywords Maxwell’s Equations \cdot QSVT \cdot Quantum Computing \cdot Compact Schemes \cdot PennyLane 1 Introduction Maxwell’s equations and the Lorentz force law form the foundation of classical electromagnetism, underpinning technologies such as optics, wireless communication, and electronics [ 1 , 2 , 3 ] . Due to their inherent nonlinearity and hyperbolic nature, obtaining exact analytical solutions to these equations is challenging, particularly in multidimensional scenarios. As a result, computational approaches such as the finite-difference time-domain (FDTD) technique [ 4 ] , finite-element methods [ 5 ] , and finite-volume schemes [ 6 ] have become essential tools for producing approximate numerical solutions. These numerical methods often rely on solving large systems of linear equations, which arise naturally from the discretization of Maxwell’s equations and many other partial differential equations. Linear systems of equations play an important role in many areas of science and technology. For example, in biology they arise in modeling metabolic networks, population dynamics, and biochemical reaction pathways [ 7 , 8 , 9 ] . In machine learning, training algorithms such as linear regression or support vector machines often require solving systems of equations derived from optimization objectives [ 10 ] . In computational physics and engineering, discretizing partial differential equations using finite difference or finite element methods reduces the problem to solving large, sparse linear systems [ 11 ] . In fluid dynamics, the discretization and linearization of the Navier–Stokes equations during numerical simulation also necessitates robust and efficient solvers [ 12 ] . The ubiquity of linear systems across scientific disciplines highlights the fundamental importance of developing scalable algorithms for their solution. In the past decade, significant attention has been given to the possibility of solving linear systems on quantum computers. Classically solving an N \times N linear system (N equations for N unknowns) scales polynomially in N. In contrast, Harrow-Hassidim-Lloyd (HHL) [ 13 ] introduced a quantum algorithm that scales logarithmically in N, suggesting that quantum computers may provide an exponential speedup for certain linear system problems. More recently, the Variational Quantum Linear Solver (VQLS) [ 14 ] has been proposed as a hybrid quantum-classical alternative that is more suitable for near-term quantum devices. VQLS formulates the solution of Ax=b as a variational optimization problem, where a parameterized quantum circuit prepares an approximate solution state |x\rangle by minimizing a cost function that encodes the residual error in the linear system. This method avoids deep quantum circuits and phase estimation, making it more practical under realistic noise and coherence constraints. However, VQLS relies on repeated ansatz optimization, which can be computationally expensive and less practical for solving time-dependent PDEs that require solutions at multiple time steps. An alternative, fully quantum paradigm is the Quantum Singular Value Transformation (QSVT) [ 15 ] , which provides a powerful and modular framework for implementing matrix functions on quantum states. QSVT leverages block-encoded matrices and polynomial approximations to perform transformations directly on the singular values of an operator. By constructing a polynomial approximation to the reciprocal function f(x)\approx 1/x over the spectrum of A , QSVT enables the implementation of an inverse transformation A^{-1} as a quantum circuit. This approach generalizes and improves upon earlier linear system solvers by avoiding phase estimation and offering provable bounds on approximation error and complexity, while remaining compatible with fault-tolerant quantum computation. Several recent works have addressed the use of QSVT for solving partial differential equations. Lubasch et al. [ 16 ] leveraged QSVT alongside the quantum Fourier transform (QFT) to design quantum circuits with optimal asymptotic gate complexity. Their constructions, applied to canonical PDEs such as incompressible advection, heat, isotropic acoustic wave, and Poisson’s equations, yield circuits that are conceptually simple and hardware-efficient. Novikau et al. [ 17 ] proposed a quantum algorithm using QSVT to simulate dissipative wave propagation in inhomogeneous linear media as a boundary-value problem. Their approach constructs quantum circuits modeling electromagnetic wave propagation in one-dimensional systems with outgoing boundary conditions. When it comes to Maxwell’s equations, several studies have investigated their solution using quantum computing approaches. Jin et al. [ 18 ] proposed quantum algorithms based on the Schrödingerisation technique, enabling unitary evolution for simulating Maxwell’s equations. Chen et al. [ 19 ] used Quantum Physics Informed Neural Networks (QPINN) to solve two-dimensional (2D) time-dependent Maxwell’s equations. Their approach utilized a parameterized quantum circuit (PQC) in conjunction with the classical neural network architecture while enforcing physical laws, including a global energy conservation principle, during the training of the neural network. Nguyen and Thompson [ 20 ] investigated the potential of using the variational quantum imaginary time evolution (VarQITE) algorithm on near-term quantum hardware to solve for the Maxwell’s equations. Koukoutsis et al. [ 21 ] developed a quantum Schrödinger representation of Maxwell’s equations tailored for wave propagation in cold, inhomogeneous magnetized plasmas, enabling unitary, energy-preserving evolution. They also proposed a qubit lattice algorithm (QLA) that can be implemented on quantum computers and tested classically, demonstrated with simulations of electromagnetic wave scattering. The central idea of QSVT is to encode a target matrix into a block of a unitary operator, then apply a sequence of single-qubit rotations and controlled operations to approximate a polynomial function of its singular values. This framework includes several quantum algorithms, including quantum walks, amplitude amplification, and Hamiltonian simulation, under a single algebraic technique. Unlike HHL, QSVT-based methods do not require explicit eigenvalue estimation or ancilla post-selection, which can significantly reduce circuit depth and improve practicality on near-term quantum devices. Recent theoretical work [ 22 ] shows that QSVT achieves nearly optimal query complexity for various matrix functions, including the inverse, sign, and exponential functions, depending on the spectral properties of the matrix. For the linear

Read the complete text