Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

arXiv:2305.00653

A Polynomial Time Quantum Algorithm for Exponentially Large Scale Nonlinear Differential Equations via Hamiltonian Simulation

Yu Tanaka, Keisuke Fujii

Published May 1, 2023• 15 citations
arXivHTMLPDF

Abstract

Quantum computers have the potential to efficiently solve a system of nonlinear ordinary differential equations (ODEs), which play a crucial role in various industries and scientific fields. However, it remains unclear which system of nonlinear ODEs, and under what assumptions, can achieve exponential speedup using quantum computers. In this work, we introduce a class of systems of nonlinear ODEs that can be efficiently solved on quantum computers, where the efficiency is defined as solving the system with computational complexity of $O(T {\rm log}(N) {\rm polylog}(1/ε))$, where $T$ is the evolution time, $ε$ is the allowed error, and $N$ is the number of variables in the system. Specifically, we employ the Koopman-von Neumann linearization to map the system of nonlinear ODEs to Hamiltonian dynamics and find conditions where the norm of the mapped Hamiltonian is preserved and the Hamiltonian is sparse. This allows us to use the optimal Hamiltonian simulation technique for solving the nonlinear ODEs with $O({\rm log}(N))$ overhead. Furthermore, we show that the nonlinear ODEs include a wide range of systems of nonlinear ODEs, such as the nonlinear harmonic oscillators and the short-range Kuramoto model. This is the first concrete example of solving systems of nonlinear ODEs with exponential quantum speedup by the Koopman-von Neumann linearization, although it is noted that this assumes efficient preparation of the initial state and computation of the output. These findings contribute significantly to the application of quantum computers in solving nonlinear problems.

DOI
10.48550/arxiv.2305.00653

Related papers

Efficient quantum algorithm for solving differential equations with Fourier nonlinearity via Koopman linearization

Judd Katz, Gopikrishnan Muraleedharan, Abhijeet Alase

Quantum algorithms offer an exponential advantage with respect to the number of dependent variables for solving certain nonlinear ordinary differential equations (ODEs). These algorithms typically begin by transforming the original nonlinear ODE into a higher-dimensional linear ODE using a linearization technique, most commonly Carleman linearization. Existing works restrict their analysis to ODEs where the nonlinearities are polynomial functions of the dependent variables, significantly limiting their applicability. In this work we construct an efficient quantum algorithm for solving ODEs with `Fourier' nonlinear terms expressible as $d{\bf u}/dt = G_0 + G_1 e^{i{\bf u}}$, where ${\bf u}$ denotes a vector of $n$ complex variables evolving with $t$, $G_0$ is an $n$-dimensional complex vector, $G_1$ is an $n \times n$ complex matrix and $e^{i{\bf u}}$ denotes the vector with entries $\{e^{iu_j}\}$. To tackle the Fourier nonlinear term, which is not expressible as a finite sum of polynomials of ${\bf u}$, our algorithm employs a generalization of the Carleman linearization technique known as Koopman linearization. We also make other methodological advances towards relaxing the stringent dissipativity condition required for efficient solution extraction and towards integrated readout of classical quantities from the solution state. Our results open avenues to the development of efficient quantum algorithms for a significantly wider class of high-dimensional nonlinear ODEs, thereby broadening the scope of their applications.

quant-pharXiv:2512.06488•Dec 9, 2025•4 citations

How to Map Linear Differential Equations to Schrödinger Equations via Carleman and Koopman-von Neumann Embeddings for Quantum Algorithms

Yuki Ito, Yu Tanaka, Keisuke Fujii

Solving linear and nonlinear differential equations with large degrees of freedom is an important task for scientific and industrial applications. In order to solve such differential equations on a quantum computer, it is necessary to embed classical variables into a quantum state. While the Carleman and Koopman-von Neumann embeddings have been investigated so far, the class of problems that can be mapped to the Schrödinger equation is not well understood even for linear differential equations. In this work, we investigate the conditions for linear differential equations to be mapped to the Schrödinger equation and solved on a quantum computer. Interestingly, we find that these conditions are identical for both Carleman and Koopman-von Neumann embeddings. We also compute the computational complexity associated with estimating the expected values of an observable. This is done by assuming a state preparation oracle, block encoding of the mapped Hamiltonian via either Carleman or Koopman-von Neumann embedding, and block encoding of the observable using $O(\log M)$ qubits with $M$ is the mapped system size. Furthermore, we consider a general classical quadratic Hamiltonian dynamics and find a sufficient condition to map it into the Schrödinger equation. As a special case, this includes the coupled harmonic oscillator model [Babbush et al., \cite{babbush_exponential_2023}]. We also find a concrete example that cannot be described as the coupled harmonic oscillator but can be mapped to the Schrödinger equation in our framework. These results are important in the construction of quantum algorithms for solving differential equations of large-degree-of-freedom.

arXiv:2311.15628•Nov 27, 2023•6 citations

Continuous Variables Quantum Algorithm for Solving Ordinary Differential Equations

Alice Barthe, Michele Grossi, Jordi Tura +1 more

Solving Ordinary Differential Equations (ODE) is important for a broad range of domains, such as engineering, weather forecast, and finance. Most quantum algorithms proposed to solve them revolve around transforming ODEs into a system of linear equations, in order to benefit from the exponential advantage promised by the HHL algorithm. However, there are known limitations to this subroutine such as the linear scaling with condition number. In particular, for HHL-based ODE solvers, this dependency yields a complexity that generally grows exponentially with the integration time. In this paper, we present a scheme that does not rely on the HHL subroutine. We use the Koopman Von Neumann (KvN) formalism that maps arbitrary nonlinear dynamics to a Hilbert space where the dynamics are unitary. This effectively reduces the ODE problem to a Hamiltonian time evolution, which is a well-known problem in quantum computing. However, this comes at a cost as the KvN formalism is expressed in an infinite-dimensional Hilbert space, and involves nonphysical states with infinite energies. Previous works have tackled this by truncating the Hilbert space, and by approximating all the relevant states and operations on qubit-based systems. Instead of mapping to qubit-based computations, in this paper, we investigate the direct use of continuous-variable quantum computers for this problem. We provide an algorithm to compile a sequence of Gaussian and non-Gaussian continuous variables gates to solve an arbitrary one-dimensional polynomial ODE. We analyze the algorithm and propose that it is intrinsically better suited for solving so-called initial distribution problems, rather than initial condition problems. We propose the first steps towards a comprehensive complexity and specify which steps need to be developed for a complete analysis.

IEEE International Conference on Quantum Computing and Engineering (QCE)•Sep 17, 2023•5 citations

Improved quantum algorithms for linear and nonlinear differential equations

Hari Krovi

We present substantially generalized and improved quantum algorithms over prior work for inhomogeneous linear and nonlinear ordinary differential equations (ODE). Specifically, we show how the norm of the matrix exponential characterizes the run time of quantum algorithms for linear ODEs opening the door to an application to a wider class of linear and nonlinear ODEs. In \cite{BCOW17}, a quantum algorithm for a certain class of linear ODEs is given, where the matrix involved needs to be diagonalizable. The quantum algorithm for linear ODEs presented here extends to many classes of non-diagonalizable matrices including singular matrices. The algorithm here is also exponentially faster than the bounds derived in \cite{BCOW17} for certain classes of diagonalizable matrices. Our linear ODE algorithm is then applied to nonlinear differential equations using Carleman linearization (an approach taken recently by us in \cite{Liue2026805118}). The improvement over that result is two-fold. First, we obtain an exponentially better dependence on error. This kind of logarithmic dependence on error has also been achieved by \cite{Xue_2021}, but only for homogeneous nonlinear equations. Second, the present algorithm can handle any sparse matrix (that models dissipation) if it has a negative log-norm (including non-diagonalizable matrices), whereas \cite{Liue2026805118} and \cite{Xue_2021} additionally require normality.

arXiv:2202.01054•Quantum•Feb 2, 2023•152 citations

Exponential Quantum Speedup in Simulating Coupled Classical Oscillators

Ryan Babbush, Dominic William Berry, Robin Kothari +2 more

We present a quantum algorithm for simulating the classical dynamics of 2 n coupled oscillators (e.g., 2 n masses coupled by springs).Our approach leverages a mapping between the Schrödinger equation and Newton's equation for harmonic potentials such that the amplitudes of the evolved quantum state encode the momenta and displacements of the classical oscillators.When individual masses and spring constants can be efficiently queried, and when the initial state can be efficiently prepared, the complexity of our quantum algorithm is polynomial in n, almost linear in the evolution time, and sublinear in the sparsity.As an example application, we apply our quantum algorithm to efficiently estimate the kinetic energy of an oscillator at any time.We show that any classical algorithm solving this same problem is inefficient and must make 2 ΩðnÞ queries to the oracle, and when the oracles are instantiated by efficient quantum circuits, the problem is bounded-error quantum polynomial time complete.Thus, our approach solves a potentially practical application with an exponential speedup over classical computers.Finally, we show that under similar conditions our approach can efficiently simulate more general classical harmonic systems with 2 n modes.

arXiv:2303.13012•Physical Review X•Dec 4, 2023•104 citations

Reduced basis algorithm for solving nonlinear differential equations on quantum computers

Monica Lăcătuş, Matthias Möller, Sauro Succi

As quantum computing moves toward scientific computing applications, nonlinear differential equations remain a central challenge since quantum evolution is intrinsically linear. In this work, we introduce a reduced basis algorithm (RBA) for polynomial nonlinear ordinary differential equations (ODEs) and spatially discretized partial differential equations (PDEs). After time discretization, the method composes the resulting polynomial update map over $m$ timesteps, identifies the reduced monomial basis appearing in this composed map, and constructs a linear RBA operator whose action recovers the exact $m$-timestep nonlinear dynamics. Thus, at the level of the chosen discrete update rule, the method introduces no additional approximation error beyond the time discretization error. The qubit number requirement is governed by the size of the reduced monomial basis. For an $n$-dimensional polynomial ODE system of degree $p>1$, the lifted register requires at most $q_m^{\mathrm{ODE}} = O(nm\log p)$ qubits in the full basis scenario. For PDEs discretized on $N^D$ grid points, a locality-based construction requires at most $q_m^{\mathrm{PDE}} = O(D\log N + n m^{D+1}\log p)$ qubits. Hence, the dependence on the grid size remains logarithmic, while the nonlinear overhead is controlled by local reduced basis size. The main computational burden is moved from the quantum computer to a classical preprocessing step, where the reduced monomial basis and RBA operator are constructed for the chosen timestep window. Through numerical tests on the Lorenz system and the one-dimensional Burgers equation, we verify that the RBA reproduces the corresponding discrete time nonlinear dynamics exactly, while exposing the trade-off between timestep composition, reduced basis growth, and locality.

math.NACross-listedarXiv:2606.13457•Jun 12, 2026•0 citations
Quantum computers have the potential to efficiently solve a system of nonlinear ordinary differential equations (ODEs), which play a crucial role in various industries and scientific fields. However, it remains unclear which system of nonlinear ODEs, and under what assumptions, can achieve exponential speedup using quantum computers. In this work, we introduce a class of systems of nonlinear ODEs that can be efficiently solved on quantum computers, where the efficiency is defined as solving the system with computational complexity of O(Tlog(N)polylog(1/ε))O(T {\rm log}(N) {\rm polylog}(1/ε))O(Tlog(N)polylog(1/ε)), where TTT is the evolution time, εεε is the allowed error, and NNN is the number of variables in the system. Specifically, we employ the Koopman-von Neumann linearization to map the system of nonlinear ODEs to Hamiltonian dynamics and find conditions where the norm of the mapped Hamiltonian is preserved and the Hamiltonian is sparse. This allows us to use the optimal Hamiltonian simulation technique for solving the nonlinear ODEs with O(log(N))O({\rm log}(N))O(log(N)) overhead. Furthermore, we show that the nonlinear ODEs include a wide range of systems of nonlinear ODEs, such as the nonlinear harmonic oscillators and the short-range Kuramoto model. This is the first concrete example of solving systems of nonlinear ODEs with exponential quantum speedup by the Koopman-von Neumann linearization, although it is noted that this assumes efficient preparation of the initial state and computation of the output. These findings contribute significantly to the application of quantum computers in solving nonlinear problems.
Quantum algorithms offer an exponential advantage with respect to the number of dependent variables for solving certain nonlinear ordinary differential equations (ODEs). These algorithms typically begin by transforming the original nonlinear ODE into a higher-dimensional linear ODE using a linearization technique, most commonly Carleman linearization. Existing works restrict their analysis to ODEs where the nonlinearities are polynomial functions of the dependent variables, significantly limiting their applicability. In this work we construct an efficient quantum algorithm for solving ODEs with `Fourier' nonlinear terms expressible as du/dt=G0+G1eiud{\bf u}/dt = G_0 + G_1 e^{i{\bf u}}du/dt=G0​+G1​eiu, where u{\bf u}u denotes a vector of nnn complex variables evolving with ttt, G0G_0G0​ is an nnn-dimensional complex vector, G1G_1G1​ is an n×nn \times nn×n complex matrix and eiue^{i{\bf u}}eiu denotes the vector with entries {eiuj}\{e^{iu_j}\}{eiuj​}. To tackle the Fourier nonlinear term, which is not expressible as a finite sum of polynomials of u{\bf u}u, our algorithm employs a generalization of the Carleman linearization technique known as Koopman linearization. We also make other methodological advances towards relaxing the stringent dissipativity condition required for efficient solution extraction and towards integrated readout of classical quantities from the solution state. Our results open avenues to the development of efficient quantum algorithms for a significantly wider class of high-dimensional nonlinear ODEs, thereby broadening the scope of their applications.
Solving linear and nonlinear differential equations with large degrees of freedom is an important task for scientific and industrial applications. In order to solve such differential equations on a quantum computer, it is necessary to embed classical variables into a quantum state. While the Carleman and Koopman-von Neumann embeddings have been investigated so far, the class of problems that can be mapped to the Schrödinger equation is not well understood even for linear differential equations. In this work, we investigate the conditions for linear differential equations to be mapped to the Schrödinger equation and solved on a quantum computer. Interestingly, we find that these conditions are identical for both Carleman and Koopman-von Neumann embeddings. We also compute the computational complexity associated with estimating the expected values of an observable. This is done by assuming a state preparation oracle, block encoding of the mapped Hamiltonian via either Carleman or Koopman-von Neumann embedding, and block encoding of the observable using O(log⁡M)O(\log M)O(logM) qubits with MMM is the mapped system size. Furthermore, we consider a general classical quadratic Hamiltonian dynamics and find a sufficient condition to map it into the Schrödinger equation. As a special case, this includes the coupled harmonic oscillator model [Babbush et al., \cite{babbush_exponential_2023}]. We also find a concrete example that cannot be described as the coupled harmonic oscillator but can be mapped to the Schrödinger equation in our framework. These results are important in the construction of quantum algorithms for solving differential equations of large-degree-of-freedom.
Solving Ordinary Differential Equations (ODE) is important for a broad range of domains, such as engineering, weather forecast, and finance. Most quantum algorithms proposed to solve them revolve around transforming ODEs into a system of linear equations, in order to benefit from the exponential advantage promised by the HHL algorithm. However, there are known limitations to this subroutine such as the linear scaling with condition number. In particular, for HHL-based ODE solvers, this dependency yields a complexity that generally grows exponentially with the integration time. In this paper, we present a scheme that does not rely on the HHL subroutine. We use the Koopman Von Neumann (KvN) formalism that maps arbitrary nonlinear dynamics to a Hilbert space where the dynamics are unitary. This effectively reduces the ODE problem to a Hamiltonian time evolution, which is a well-known problem in quantum computing. However, this comes at a cost as the KvN formalism is expressed in an infinite-dimensional Hilbert space, and involves nonphysical states with infinite energies. Previous works have tackled this by truncating the Hilbert space, and by approximating all the relevant states and operations on qubit-based systems. Instead of mapping to qubit-based computations, in this paper, we investigate the direct use of continuous-variable quantum computers for this problem. We provide an algorithm to compile a sequence of Gaussian and non-Gaussian continuous variables gates to solve an arbitrary one-dimensional polynomial ODE. We analyze the algorithm and propose that it is intrinsically better suited for solving so-called initial distribution problems, rather than initial condition problems. We propose the first steps towards a comprehensive complexity and specify which steps need to be developed for a complete analysis.
We present substantially generalized and improved quantum algorithms over prior work for inhomogeneous linear and nonlinear ordinary differential equations (ODE). Specifically, we show how the norm of the matrix exponential characterizes the run time of quantum algorithms for linear ODEs opening the door to an application to a wider class of linear and nonlinear ODEs. In \cite{BCOW17}, a quantum algorithm for a certain class of linear ODEs is given, where the matrix involved needs to be diagonalizable. The quantum algorithm for linear ODEs presented here extends to many classes of non-diagonalizable matrices including singular matrices. The algorithm here is also exponentially faster than the bounds derived in \cite{BCOW17} for certain classes of diagonalizable matrices. Our linear ODE algorithm is then applied to nonlinear differential equations using Carleman linearization (an approach taken recently by us in \cite{Liue2026805118}). The improvement over that result is two-fold. First, we obtain an exponentially better dependence on error. This kind of logarithmic dependence on error has also been achieved by \cite{Xue_2021}, but only for homogeneous nonlinear equations. Second, the present algorithm can handle any sparse matrix (that models dissipation) if it has a negative log-norm (including non-diagonalizable matrices), whereas \cite{Liue2026805118} and \cite{Xue_2021} additionally require normality.
We present a quantum algorithm for simulating the classical dynamics of 2 n coupled oscillators (e.g., 2 n masses coupled by springs).Our approach leverages a mapping between the Schrödinger equation and Newton's equation for harmonic potentials such that the amplitudes of the evolved quantum state encode the momenta and displacements of the classical oscillators.When individual masses and spring constants can be efficiently queried, and when the initial state can be efficiently prepared, the complexity of our quantum algorithm is polynomial in n, almost linear in the evolution time, and sublinear in the sparsity.As an example application, we apply our quantum algorithm to efficiently estimate the kinetic energy of an oscillator at any time.We show that any classical algorithm solving this same problem is inefficient and must make 2 ΩðnÞ queries to the oracle, and when the oracles are instantiated by efficient quantum circuits, the problem is bounded-error quantum polynomial time complete.Thus, our approach solves a potentially practical application with an exponential speedup over classical computers.Finally, we show that under similar conditions our approach can efficiently simulate more general classical harmonic systems with 2 n modes.
As quantum computing moves toward scientific computing applications, nonlinear differential equations remain a central challenge since quantum evolution is intrinsically linear. In this work, we introduce a reduced basis algorithm (RBA) for polynomial nonlinear ordinary differential equations (ODEs) and spatially discretized partial differential equations (PDEs). After time discretization, the method composes the resulting polynomial update map over mmm timesteps, identifies the reduced monomial basis appearing in this composed map, and constructs a linear RBA operator whose action recovers the exact mmm-timestep nonlinear dynamics. Thus, at the level of the chosen discrete update rule, the method introduces no additional approximation error beyond the time discretization error. The qubit number requirement is governed by the size of the reduced monomial basis. For an nnn-dimensional polynomial ODE system of degree p>1p>1p>1, the lifted register requires at most qmODE=O(nmlog⁡p)q_m^{\mathrm{ODE}} = O(nm\log p)qmODE​=O(nmlogp) qubits in the full basis scenario. For PDEs discretized on NDN^DND grid points, a locality-based construction requires at most qmPDE=O(Dlog⁡N+nmD+1log⁡p)q_m^{\mathrm{PDE}} = O(D\log N + n m^{D+1}\log p)qmPDE​=O(DlogN+nmD+1logp) qubits. Hence, the dependence on the grid size remains logarithmic, while the nonlinear overhead is controlled by local reduced basis size. The main computational burden is moved from the quantum computer to a classical preprocessing step, where the reduced monomial basis and RBA operator are constructed for the chosen timestep window. Through numerical tests on the Lorenz system and the one-dimensional Burgers equation, we verify that the RBA reproduces the corresponding discrete time nonlinear dynamics exactly, while exposing the trade-off between timestep composition, reduced basis growth, and locality.