CV
Education and research and engineering experience of Guangyu Xiang.
Contact Information
| Name | Guangyu Xiang |
| Professional Title | Ph.D. Student in Data Science and Analytics |
| gxiang190@connect.hkust-gz.edu.cn |
Professional Summary
Ph.D. student at HKUST(GZ) working on machine learning systems, efficient and elastic LLM training and inference, GPU computing, distributed systems, and communication-computation co-optimization.
Experience
-
2023 - 2023 Beijing, China
Research Intern, Systems Research Group
Microsoft Research Asia
Studied Multi-Head Attention and FlashAttention optimization, developed CUDA kernels, and explored compiler cost models.
-
2022 - 2023 Beijing, China
R&D Engineer Intern, AI Framework Team
OneFlow
Reworked MLIR operator fusion and rewrite patterns with PDLL, and supported model compatibility across frameworks and accelerators.
-
2022 - 2022 Beijing, China
AI Compiler Engineer Intern
AMD
Built TVM Relay passes for ASR model compilation, including quantization-flow optimization and AIE deployment evaluation.
-
2020 - 2021 Shenzhen, China
AI Development Engineer Intern
Peng Cheng Laboratory
Contributed to resource management and job scheduling for high-performance computing clusters.
Education
-
2025 - present Guangzhou, China
Ph.D.
The Hong Kong University of Science and Technology (Guangzhou), HKUST(GZ)
Data Science and Analytics, Information Hub
- Advisor: Prof. Xiaowen Chu
- Machine Learning Systems; Distributed LLM Training and Serving; GPU Computing
-
2021 - 2024 Beijing, China
Master's Degree
Peking University (PKU)
Software Engineering
- School of Software and Microelectronics
- High-Performance Computing, AI Compiler Optimization, and CUDA Programming
-
2017 - 2021 Chengdu, China
Bachelor's Degree
University of Electronic Science and Technology of China (UESTC)
Software Engineering
- School of Information and Software Engineering
- GPA 3.86/4.00; ranked in the top 5%
Publications
-
2026