On August 26th, the 2023 CCF National Annual Conference on High Performance Computing (CCF HPC China 2023) successful concluded. The research team led by Assist Prof. Cui Jiahuan of ZJUI won the "2023 China Supercomputing Best Application" award for their research results in the "Ultra Large Scale and High Precision Aerodynamic Thermodynamic Simulation of Aircraft Engines Based on Sunway Supercomputing" project. As one of the three most influential supercomputing conferences in the field of high-performance computing globally, this conference aims to promote the application ecology, research, and industrial digital upgrading and transformation of supercomputing technology through the exchange of cutting-edge academic achievements, innovative application technology display, and showcase China's new charm in the field of supercomputing.

"China Supercomputing Best Application" is selected annually to recognize outstanding achievements in high-performance computing in China. It is to track the development of parallel computing, stimulate innovation in the application of high-performance computing in fields such as science, engineering, and large-scale data analysis applications, and award the best performance or special achievements in the scalability and timeliness of solving important scientific and engineering problems in China.

The only one to receive this honor this year is the team led by Assist Prof. Cui Jiahuan of ZJUI. The team developed, parallelized, and tested a large eddy simulation program for complex flow in aircraft engines using the flux reconstruction (FR) method in the " Ultra Large Scale and High Precision Aerodynamic Thermodynamic Simulation of Aircraft Engines Based on Sunway Supercomputing" project. In response to the architectural characteristics of the Sunway series supercomputers, the team designed a multi-level hybrid parallel solution that deeply optimized the data exchange between the main core processing element (MPE) and the many-core processing element (CPE), as well as the execution order and mutual integration of the computing kernel. Due to the good locality of the FR method, the high-order FR format has a high computational intensity for matrices, with matrix operations accounting for up to 60% of the computational time. For a high-pressure turbine example with 1.69 billion grid cells and 865 trillion degrees of freedom (DoF), the solver achieved sustained computational performance of 115.8 DP-PFLOPs. With the development of large-scale heterogeneous computing and the advancement of numerical technology, they have demonstrated the feasibility of using large eddy simulation to solve complex flow problems in aviation engines. This work has taken a solid step towards achieving NASA's proposed numerical simulation of aircraft engines.

Dr. Cui Jiahuan is currently an Assistant Professor at ZJUI. He graduated with a PhD degree from the University of Cambridge in the UK and served as a postdoctoral researcher there. Dr. Cui mainly engaged in the development of computational fluid dynamics programs, large-scale parallel computing, computational simulation, and flow mechanism research of aerodynamic components of aviation engines such as turbines, compressors, and nacelles. He led and participated in 6 provincial level projects and published over 40 papers and led the team to be nominated for the "Gordon Bell Award", the highest award in the field of high-performance computing in 2023.