Recently, Professor Shi Jin and Professor Nana Liu from the Institute of Natural Sciences, School of Mathematical Sciences, and Pujiang International College at Shanghai Jiao Tong University, unveiled their groundbreaking research on a quantum algorithm designed to find the viscosity solution of the Hamilton-Jacobi equation. Their findings were published in PNAS, the esteemed journal of the National Academy of Sciences of the United States.

This study tackles the challenge of solving nonlinear partial differential equations (PDEs) within the realm of quantum computing by introducing an innovative quantum algorithm. Quantum computing, which operates on the principles of quantum mechanics, excels at solving linear PDEs, such as the Schrödinger equation. However, the realm of nonlinear equations has long posed a significant obstacle for quantum computing techniques.

Previous linear truncation methods have proven effective only for specific problems. In contrast, most nonlinear PDEs encountered in scientific and engineering computations, like the Hamilton-Jacobi equation, are characterized by strong nonlinearity and weak dissipation. These properties render them resistant to direct solution via quantum computing.

Leveraging the entropy penalty method proposed by Gomes-Valdinoci, Professors Shi Jin and Nana Liu have, for the first time, constructed a quantum algorithm tailored to solve the viscosity solution of the Hamilton-Jacobi equation. By integrating artificial viscosity methods, entropy penalty techniques, and Schrödingerization, they have developed an efficient quantum algorithm capable of solving the viscosity solution.

This algorithm not only provides error estimates that remain valid over extended periods but also showcases a polynomial-scale quantum advantage in calculating certain key physical observables. This breakthrough paves the way for broader applications of quantum computing across diverse fields, including finance, economics, materials science, earthquake prediction, image processing, and machine learning.