A PROVABLE ORACLE-FREE QUANTUM ALGORITHM FOR NONLINEAR DYNAMICS ON HYBRID OSCILLATOR–QUBIT PROCESSORS

Research Question and Approach

This paper addresses the challenge of simulating nonlinear ordinary differential equations (ODEs) on quantum hardware. Because standard quantum algorithms are optimized for linear systems, the authors utilize the Fokker-Planck embedding to transform a nonlinear ODE into a linear partial differential equation governing the evolution of a probability density.

To simulate this linear system, the authors employ the Schrödingerisation technique, which maps the non-unitary evolution of the density into a parametrised family of unitary Schrödinger equations. A key innovation is the use of a single continuous-variable (CV) qumode to represent the Fourier-mode parameter of this family, allowing the entire continuum to be evolved within a single coherent quantum circuit.

Structural Contributions

The core technical contribution is a bipartite Pauli decomposition of the generator's Hermitian and skew-Hermitian parts. The authors prove that the non-zero Pauli strings can be organized into O(log N) mutually commuting families. This structure allows for the exact factorization of each family exponential into a product of monomial-controlled momentum displacements, effectively eliminating intra-family Trotter errors. Because the gates are derived directly from the polynomial coefficients of the drift, the algorithm is entirely oracle-free, meaning all gate costs are explicitly accounted for.

This work provides a rigorous, resource-efficient path for simulating nonlinear dynamics, which are ubiquitous in fields like chemical kinetics, fluid dynamics, and neural modeling. By replacing the traditional discretised mode register with a single physical qumode, the algorithm demonstrates an accuracy-per-resource advantage. Furthermore, by avoiding the need for black-box sparse-access oracles or block encodings, the construction offers a transparent and implementable framework for current and near-term hybrid quantum processors.