Modelling classical physics has helped us to understand the origins of the universe and observe deep space phenomena that occurred light years ago. But accurately modelling highly unpredictable physical systems here on earth, from complex weather patterns to improved aerodynamics during turbulence, remains extremely challenging.
This is because the equations governing such chaotic ‘non-linear’ systems are highly complex to solve, if not unsolvable, using existing differential equations and algorithms. The more non-linear the system, the more complex the differential equations it requires. The solution is to use more powerful quantum algorithms to solve these equations. But until now, such algorithms have been limited in speed, accuracy and functionality.
New research published in Quantum Journal in February 2023 by Riverlane’s Dr Hari Krovi, in partnership with MIT and the US Department of Energy, represents significant progress in this field by developing quantum algorithms that can be applied to a substantially larger class of linear and nonlinear differential equations with greater speed and accuracy. The research builds on a previous paper published in the PNAS journal in August 2021 Krovi and several others.
The practical use cases apply to a range of non-linear as well as linear systems. These include: accurate simulation of fluid dynamics in the presence of viscosity and turbulence; weather modelling; and simulations of plasma physics applied to inertial confinement fusion, aiding clean nuclear energy development.
Krovi’s research is a stepping stone to further research and development in the field. We still need better quantum algorithms for systems with far higher levels of non-linearity, for example.
The full paper in Quantum Journal can be found here.