Undergraduate at Northwestern Polytechnical University
Computer Systems · Algorithms & Hardware · Computer Vision · Embodied AI
I am an undergraduate student in the Computer Science Top-Notch Base Class (计算机科学拔尖基地班) at Northwestern Polytechnical University. My current academic interests lie in computer systems & architecture, algorithms & hardware co-design, and computer vision for embodied AI.
💡 I enjoy exploring both ends of the computing spectrum—from hardware microarchitecture, out-of-order execution, and RTL design, to high-level visual grounding algorithms and intelligent systems.
| Area | Topics | |
|---|---|---|
| 💻 | Computer Systems & Architecture | Out-of-order superscalar processors, LoongArch architecture, MMU & TLB, Verilog RTL |
| 🤖 | Embodied AI & 3D Vision | Functional 3D visual grounding, graph retrieval (GNN), multi-modal reasoning |
| ⚡ | Algorithms & Hardware | Hardware-software co-design, algorithm optimization, digital logic design |
| 📊 | Mathematical Modeling & Tools | Mathematical modeling (MCM/ICM), PyTorch, Vivado, Linux environment |
Northwestern Polytechnical University
Undergraduate in Computer Science and Technology (Top-Notch Base Class) · 2024.09 – 2028.06
- Academic Standing: GPA: 3.945 / 4.1 · GPA Rank: 14/336 (Top 4%) · Overall Rank: 3/336 (Top 1%)
- Selected Coursework: Computer Architecture & Organization (100), Data Structures (99), Digital Logic Design II (98), Linear Algebra (97)
LoongArch Verilog RTL Vivado Computer Architecture Hardware Debugging
- Designed and verified an out-of-order 4-issue superscalar CPU supporting 8-way fetch, 4-way decode, in-order commit, register renaming, branch prediction, and partial out-of-order Load execution.
- Responsible for MMU and TLB design, memory pipeline supporting partial out-of-order load execution, multi-port issue queue, and an 8-to-4 InstFIFO buffer.
- Achieved 80 MHz operating frequency with an IPC of 1.518× compared to openla500; awarded National Second Prize in the 2026 Loongson Cup.
3D Grounding Graph Retrieval (GNN) Embodied AI Python Robotics
- Developed a query-conditioned edge-gated graph retrieval framework for natural-language-driven functional 3D grounding, evaluated and demonstrated on the Unitree G1 humanoid robot.
- Built automated data filtering and verification pipelines in Python to rectify annotations, extract geometric spatial properties (position, scale), and enhance multi-step relational reasoning benchmarks.
- (Targeted for submission to IEEE RAL)
- 🥇 National Scholarship (国家奖学金), Ministry of Education (2025.09 & 2026.09)
- 🥈 National Second Prize, 10th Loongson Cup National Student Computer System Capability Challenge (CPU Design) (2026.08)
- 🏆 Finalist, Interdisciplinary Contest in Modeling (MCM/ICM) (2026.05)
- 🥉 Third Prize (Provincial/Northwest), China Computer Design Contest (2026.06)
- 🥈 Second Prize (Provincial/Shaanxi), China Undergraduate Mathematical Contest in Modeling (CUMCM) (2025.12)