A hybrid quantum-classical framework for computational fluid dynamics

Abstract

Great progress has been made in quantum computing in recent years, providing opportunities to overcome computation resource poverty in many scientific computations like computational fluid dynamics (CFD). In this work, efforts are made to exploit quantum potentialities in CFD, and a hybrid classical and quantum computing CFD framework is proposed to release the power of current quantum computing. In this framework, the traditional CFD solvers are coupled with quantum linear algebra libraries in weak form to achieve collaborative computation between classical and quantum computing. The quantum linear solver provides high-precision solutions and scalable problem sizes for linear systems and is designed to be easily callable for solving linear algebra systems similar to classical linear libraries, thus enabling seamless integration into existing CFD solvers. Some typical cases are performed to validate the feasibility of the proposed framework and the correctness of quantum linear algorithms in CFD.