- Sept 2026: CRISP, our work on input-adaptive sparse attention for long-context LLM prefilling, was accepted to the main conference at EMNLP 2026.
- Summer 2026: I received a research internship offer from Amazon for Summer 2026.
- Jun 2025 - Sep 2025: I completed a Research Internship at Microsoft Research, working on GPU kernel optimization for LLM inference.
- Sept 2024: I started my Ph.D. in Computer Science at the University of Oregon, advised by Assoc. Prof. Thien Huu Nguyen.
- Mar 2024: I implemented a faster RoPE embedding for Unsloth (PR #238).
- Jan 2024: We introduced Vistral, a state-of-the-art conversational LLM for Vietnamese.
- Oct 2023: Two papers were accepted at EMNLP 2023.
- July 2022: I joined VinAI Research as a Research Resident.
Hello, I'm Huy Nguyen
I am a Ph.D. researcher at the University of Oregon, co-advised by Assoc. Prof. Thien Huu Nguyen and Asst. Prof. Jieyang Chen. I specialize in the systems-level architecture and distributed acceleration of Large Language Models (LLMs). My work focuses on pushing the physical limits of hardware utilization, from bare-metal CUDA/Cutlass kernel engineering, to designing topology-aware 5D parallel training engines. I build the underlying computational and cross-GPU interconnect infrastructure required to eliminate network bubbles and compute bottlenecks at massive scale.
Prior to this, I completed my B.S. in Computer Science at Hanoi University of Science and Technology with an Excellent Degree (GPA 3.71/4.0), advised by Dr. Linh Ngo Van.