liamY

Hi, I’m Ziwei (Liam) Yuan. I am an undergraduate in Software Engineering at the University of Electronic Science and Technology of China (2022–2026) and an incoming Ph.D. student in Computer Science at Tsinghua University (from 2026). I am fortunate to be advised by Prof. Mingxing Zhang in Madsys team. My work lies at the intersection of machine learning system and high‑performance computing, with a focus on heterogeneous inference for cutting edge models, compiler/operator‑level optimization, and distributed systems.

Currently, I contribute to KTransformers, an open‑source library for efficient Transformer inference on CPU/GPU. My work centers on heterogeneous operator optimization, KV‑cache management, and load balancing to improve throughput while maintaining low latency. Repository

Publications

• KTransformers: Unleashing the Full Potential of CPU/GPU Hybrid Inference for MoE Models (SOSP ’25). DOI

Awards

• National Scholarship (2023, 2024)
• Tencent Scholarship (2024)

Skills

• Quantization; intrinsics optimization; kernel fusion; sparse operator optimization; embedded OS development
• Programming: C/C++, Rust, Arm assemble, Python, Go

Links

• GitHub