The opportunity and grand challenges in computational fluid dynamics by exascale computing

Exascale computer system will come into service in around 2020, and exascale computing will lead to transformational changes in computing scientific and scientific research. In Computational Fluid Dynamics (CFD), one of the main application fields of high performance computing, computing at the exascale promises dramatic advances in our capabilities to model and simulate complex phenomena of flows, at levels of fidelity that have the potential to dramatically change both our understanding and our ability to comprehend. Thus, there are almost certain to be great benefits to going to the exascale. In this paper, the opportunity and grand challenges in CFD were discussed with approaching of exascale computing. Firstly, the current progress of high performance computing in CFD was reviewed briefly. Secondly, the grand challenges and development tendency in CFD itself and applications of exascale computing in CFD were discussed. Finally, the suggestion for the development of future CFD was presented in the authors' opinion. In order to achieve the goal of exascale computing in CFD, the co-design of CFD, applied the applied math, the computer science, and the hardware community (including the hardware vendors) should be closely coupled in the exascale program.