Shuzhang Zhong

Shuzhang Zhong

Second-Year PhD student

Peking University

Biography

I’m currently a third-year Ph.D. student advised by Prof. Meng Li at the School of Integrated Circuit, Peking University. Before attending Peking University, I received a bachelor’s degree from Beihang University and served as the captain of SuperComputer Team under the supervision of Prof. Hailong Yang. My research interests focus on efficient multi-modality AI acceleration algorithms, systems and hardwares.

Interests
  • Efficient Multi-Modality Artificial Intelligence
  • Algorithm/System/Hardware Co-Design/Co-Optimization
Education
  • PhD in Integrated Circuit, 2023-

    Peking University

  • BSc in Computer Science, 2019-2023

    Beihang University

Experience

 
 
 
 
 
Research Intern
MicroSoft Research
September 2025 – Present Beijing
Efficient, Robust and Scalable agent system.
 
 
 
 
 
Research Intern
Alibaba DAMO Academy
July 2024 – July 2025 Beijing
Scheduling optimization on 3D Near-Memory Processing Hardware.

Recent Publications

Quickly discover relevant content by filtering publications.
(2025). H2EAL: Hybrid-Bonding Architecture with Hybrid Sparse Attention for Efficient Long-Context LLM Inference. In ICCAD 2025.

(2025). HD-MoE: Hybrid and Dynamic Parallelism for Mixture-of-Expert LLMs with 3D Near-Memory Processing. In ICCAD 2025.

(2025). HybriMoE: Hybrid CPU-GPU Scheduling and Cache Management for Efficient MoE Inference. In DAC 2025.

(2025). SpecASR: Accelerating LLM-based Automatic Speech Recognition via Speculative Decoding. In DAC 2025.

(2024). AdapMoE: Adaptive Sensitivity-based Expert Gating and Management for Efficient MoE Inference. In ICCAD 2024.

Contact