Experience

  1. Research Assitant

    Kent State University

    Program resilience analysis on High-performance computing (HPC) systems. Responsibilities include:

    • Built a soft error simulation platform on LLVM
    • Developed a resilience analysis platform with transformer/LLM
    • Improved the resilience prediction accuracy by 30%
    • Built a visualization framework to study error propagation using the control-flow graph
  2. Research Aide Technical

    Argonne National Lab

    Research topic: translation between parallelization languages:

    • Complete the metadata process in the project “EXCELLENT” which translates parallelization language to each other in the compilation,
    • Engineering in a combination of Noelle and SPLENDID
    • Setup experiment environment for OpenMP/CUDA decompilation
  3. Summer Internship

    Los Almos National Lab

    Responsibilities include:

    • Developed FI-VIS tool based on Pin and Pinfi
    • Trace and visualize error propagation in program execution at the instruction level.

Education

  1. PhD in Computer Science

    Kent State University
    Thesis on Research on resilience in high-performance Computing (HPC) applications with Large Language Models. Supervised by Prof Qiang Guan. Presented papers at 3 IEEE conferences with the contributions being published in 2 Springer journals.
    Read Thesis
  2. MEng in IC Engineering

    University of Chinese Academy of Science, Beijing, China
    Thesis on The Study of Cu2ZnSnS4 films generation by sulfur-free annealing process and device application.
  3. BSc in Electronic Science and Technology

    Xidian University, Xi’an, China
    Thesis on A novel infrared object tracking algorithm.
Skills & Hobbies
Technical Skills
Programming

C++/C, Python, Shell/Cmake

Tools and Techs

LLVM/Clang, CUDA/OpenMP, pytorch/TensorFlow, GNN, LLM/BERT

Research Aid

Git, Latex, VS studio

Hobbies
Marathon
Badminton/Basketball
Photography
Awards
Graduate Student Travel Awards
Kent State University ∙ November 2024
Languages
100%
English
100%
Chinese