Wushi Dong

Wushi Dong is a software engineer at Meta, focusing on compiler runtime and efficiency for the Meta Training and Inference Accelerator (MTIA) since February 2025. Prior to this role, Wushi served as an applied scientist at Amazon Web Services from April 2020 to February 2025, where projects involved accelerating machine learning on the cloud through advanced compiler and High-Performance Computing technologies. Wushi also held a postdoctoral researcher position at the University of Chicago, leading a project on distributed training for 3D image segmentation in partnership with Argonne Leadership Computing Facilities. In addition, Wushi completed a research internship at IBM, creating a deep learning model for video classification that achieved state-of-the-art accuracy, and a research aide role at Argonne National Laboratory.