About me
Education
Sun Yat-sen University B.E. · MA.Eng in Computer Science , 2019/09 - 2026/06 (expected)
- Supervised by X. Zhang, interested in compiler and Machine Learning.
Research
VecTrans: LLM Transformation Framework for Better Auto-vectorization on High-performance CPU
Z. Zheng, L. Cheng, L. Li, R. Rocha, T. Liu, W. Wei, X. Zhang, Y. Gao. LCTES’25 submitted
- VecTrans leverages LLMs to enhance compiler-based code vectorization.
- VecTrans first employs compiler analysis to identify potentially vectorizable code regions. It then utilizes an LLM to refactor these regions into patterns that are more amenable to the compiler’s auto-vectorization.
mLOOP: Optimize Loop Unrolling in Compilation with a ML-based Approach
Z. Zheng, Y. Wu, X. Zhang. GitHub. NAS’24
- mLOOP employs the XGBoost model to predict loop unroll factors.
- mLOOP is implemented as an LLVM optimization pass, enabling seamless deployment within existing compiler pipelines.
GoPTX: Fine-grained GPU Kernel Fusion by PTX-level Instruction Weaving
K. Wu, Z. Lin, X. Meng, Z. Zheng, W. Pan, X. Zhang, Y. Lu. GitHub. DAC’25
REFIT: Improve Code Efficiency via Binary Level Loop Optimization
T. Zhang, G. Chen, W. Pan, Z. Zheng, G. Sun and X. Zhang. ICCCS’25
Research Interest
Compiler Optimizations & ML-Augmented Code Generation
- Focus on synergistic techniques between compiler optimizations (e.g., loop unrolling, vectorization) and machine learning, enhancing performance in HPC and edge computing.
- Future work: Repository-level code optimization (cross-procedural/inter-file analysis) and scalable parallel code generation for heterogeneous architectures (CPU/GPU/TPU).
Experience
2024/08 - now, Huawei 2012 Lab, Compiler Engineer (Intern)
2024/02 - 2024/07, Sun Yat-sen University, School of Computer Science ,Teach Assistant for DCS292 - Compiler Construction.
Developed and guided course materials for compiler design, with key contributions to:
- Experiment 3 (IR Generation) –Translating AST to LLVM IR.
- Experiment 4 (Optimization) –Implementing and analyzing IR-level optimizations.
- Integrating LLMs for code generation and performance optimization.
Honor
2024/10, First-Prize Scholarship, Sun Yat-sen University
2024/07, First Prize, Best Teaching Case, China Computer Education Conference(CCEC’24)
2023/11, Second-Prinze, Greater Bay Area Cup AI for Science Technology Competition
2023/08, First-Prize, Huawei Bisheng Cup Compiler System Design Competition (CSC)
2022/09, CCF Elite Collegiate Award (102 in China), China Computer Federation (CCF)
2022/06, Top 5%, CCF Certified Software Professional
2022/04, Gold Award,iGEM Competition
2021/10, Second-Prize Scholarship, Sun Yat-sen University