Skip to navigation Skip to content
Careers | Phone Book | A - Z Index

Hengjie Wang

Hengjie Wang
Postdoctoral Researcher

I am a postdoctoral scholar in the Center for Computational Sciences and Engineering (CCSE) in the Applied Mathematics and Computational Research (AMCR) Division at Lawrence Berkeley National Laboratory.

My research aims at designing High Performance Computing (HPC) algorithms to speed up numerical solvers. More specifically, I am interested in designing efficient and robust algorithms for load balancing, optimizing stencil computations, and minimizing communication costs on modern heterogeneous supercomputers. I am also interested in using Machine Learning (ML) techniques to build fast surrogates for conventional CFD solvers.

I achieved my PhD degree in Mechanical Engineering from the University of California-Irvine in March 2021. My thesis work focused on systematical algorithm designs to optimize conventional CFD solvers on modern hardware, including grid partitioning, cache tiling, computation-communication overlapping, as well as efficient ML surrogates. 

Journal Articles

Hengjie Wang, Robert Planas, Aparna Chandramowlishwaran, Ramin Bostanabad, "Mosaic flows: A transferable deep learning framework for solving PDEs on unseen domains", Computer Methods in Applied Mechanics and Engineering, 2022, 389:114424,

Conference Papers

Hengjie Wang, Aparna Chandramowlishwaran, "Pencil: a pipelined algorithm for distributed stencils", SC20: International Conference for High Performance Computing, Networking, Storage and Analysis, 2020, 1--16,

Hengjie Wang, Aparna Chandramowlishwaran, "Multi-criteria partitioning of multi-block structured grids", Proceedings of the ACM International Conference on Supercomputing, 2019, 261--271,