Zheng,Hong-Sheng | Mediatek inc.

About Me

I'm Hong-Sheng Zheng, currently a quantum compiler engineer at Xanadu, focusing on quantum compilation technology. Prior to moving to Canada, I lived in Taiwan and worked at MediaTek as a compiler engineer specializing in AI and computer vision compiler design and optimization. I received my M.Sc. from the Department of Computer Science at National Chiao Tung University (NCTU), where I was advised by Prof. Wuu Yang.

Research Interests

TinyML: focus on peak memory reduciton, compiler optimization
Polyhedral Compiler: program anylsis, tiling and loop optimization
Algo/Schedule Decoupled Compiler: Halide/TVM based schedule language and compiler design

News

[1/2025] Our paper about multi-hierachy StreamNet inference in TinyML is accepted to TECS 2025.
[10/2023] Our paper about memory-efficient optimiztion in TinyML is accepted to HPCA 2024.
[9/2023] Our paper about hybrid(1D/2D) streaming approach applied on patch-based inference in TinyML is accepted to NeurIPS 2023.

Publications

StreamNet++: Memory-Efficient Streaming TinyML Model Compilation on Microcontrollers

Chen-Fong Hsu, Hong-Sheng Zheng, Yu-Yuan Liu, *Tsung Tai Yeh

ACM Transactions on Embedded Computing Systems (TECS), 2025

TinyTS:Memory-Efficient TinyML Model Compiler Framework on Microcontrollers

Yu-Yuan Liu, Hong-Sheng Zheng, Yu-Fang Hu, Chen-Fong Hsu, *Tsung Tai Yeh

High-Performance Computer Architecture (HPCA), 2024

StreamNet:Memory-Efficient Streaming Tiny Deep Learning Inference on the Microcontroller

Hong-Sheng Zheng, Yu-Yuan Liu, Chen-Fong Hsu, *Tsung Tai Yeh

Neural Information Processing Systems (NIPS), 2023

PDF

Awards

[2020] E.SUN bank Anti-Money Laundering Competition - Our team won third place.
[2020] Irwin Mango Defects Classification Competition - Gain honorable mention.