About Me

---

This is Zihao Ye, a second-year Ph.D. student at the University of Washington's Paul G. Allen School of Computer Science and Engineering, advised by Luis Ceze in the SAMPL research group. I also worked with Tianqi Chen on Tensor IR of Apache TVM project.

Prior to joining UW, I worked at AWS with Minjie Wang and Zheng Zhang on the DGL project. I obtained my bachelor's degree from ACM Honors Class at Shanghai Jiao Tong University.

Research

---

My research interests mainly lie on Compiler and Programming Language for modern hardwares. More specifically, I'm working on following topics:

• Deep Learning Compilers
  • IR design and automatic scheduling of irregular workloads.
  • Compilation stack for DSA (Domain Specific Architectures).
• Hardware Description Language

Feel free to drop me an email if we have aligned interests, and I'm open to collaborations.

Current Projects

---

Sparse TIR

• Sparse TIR is a unified abstraction for representing and optimizing sparse/irregular workloads in Deep Learning on top of TVM Tensor IR. It aims to generate efficient code for various sparse formats on modern tensorized hardwares.
• We presented a lightning talk at TVMCon 2021 on design principle of Sparse TIR.

Past Projects

---

Graph Learning Systems(2018-2021)

• DGL is an open-source library for Deep Learning on Graphs. Featgraph and Graphiler are two JIT compilation solutions for efficient GNN runtime system.
• BP-Transformer is an attempt on accelerating Transformer for long-range context by applying self-attention on manually designed graph structure (sparse pattern).

Publications

---

Before UW

• Graphiler: Optimizing Graph Neural Networks with Message Passing Data Flow Graph
  Zhiqiang Xie, Minjie Wang, Zihao Ye, Zheng Zhang, Rui Fan
  Conference on Systems and Machine Learning Foundation (MLSys) 2022
  [ Code ]
• FeatGraph: A Flexible and Efficient Backend for Graph Neural Network Systems
  Yuwei Hu, Zihao Ye, Minjie Wang, Jiali Yu, Da Zheng, Mu Li, Zheng Zhang, Zhiru Zhang and Yida Wang
  The International Conference for High Performance Computing, Networking, Storage, and Analysis (SC) 2020
  [ Code ]
• BP-Transformer: Modelling Long-Range Context via Binary Partitioning
  Zihao Ye, Qipeng Guo, Quan Gan, Xipeng Qiu and Zheng Zhang
  ArXiv Preprint
  [ Code ]
• Deep Graph Library: A Graph-Centric, Highly-Performant Package for Graph Neural Networks
  Minjie Wang, Da Zheng, Zihao Ye, Quan Gan, Mufei Li, Xiang Song, Jinjing Zhou, Chao Ma, Lingfan Yu, Yu Gai, Tianjun Xiao, Tong He, George Karypis, Jinyang Li and Zheng Zhang
  ArXiv Preprint
  [ Project Page ] [ Code ]

Activity and Service

---

• Artifact Evaluation Committee: MLSys 2022