Xi Xie

prof_pic.jpg

About Me

I am a Ph.D. student in the Department of Computer Science and Engineering at the University of Connecticut since Fall 2022.

My research interests include high-performance GPU algorithm design, GPU/ML system co-design, and secure and trustworthy AI/ML systems. I am skilled in C++/CUDA programming and CUDA algorithm design, and I enjoy challenging the performance of official libraries.

News

Sep, 2025 Our paper InfantAgent-Next: A Multimodal Generalist Agent for Automated Computer Interaction has been accepted by NeurIPS 2025.
May, 2025 I’m starting a new internship at NVIDIA Linear Algebra Libraries (Manager: Joost Rommes; Mentor: Quang Anh Pham).
Apr, 2025 Our papers Graph Convolutional Network Acceleration Using Adiabatic Superconductor Josephson Devices and DR-CircuitGNN: Training Acceleration of Heterogeneous Circuit Graph Neural Network on GPUs have been accepted by ICS 2025.
Jan, 2025 Our paper RANKFLOW: A Multi-Role Collaborative Reranking Workflow Utilizing Large Language Models has been accepted by WWW-RelWeb 2025.
Jan, 2025 Our paper RTop-K: Ultra-Fast Row-Wise Top-K Selection for Neural Network Acceleration on GPUs has been accepted by ICLR 2025.

Selected Publications

  1. ICLR
    topk_kernel_0325.png
    RTop-K: Ultra-Fast Row-Wise Top-K Selection for Neural Network Acceleration on GPUs
    Xi Xie, Yuebo Luo, Hongwu Peng, and 1 more author
    In The Thirteenth International Conference on Learning Representations, 2025
    PDF Code
  2. ASPLOS
    maxk_forward_kernel.png
    MaxK-GNN: Extremely Fast GPU Kernel Design for Accelerating Graph Neural Networks Training
    Xi Xie*, Hongwu Peng*, Kaustubh Shivdikar, and 6 more authors
    In Proceedings of the 29th ACM International Conference on Architectural Support for Programming Languages and Operating Systems, Volume 2, La Jolla, CA, USA, 2024
    HTML PDF Code
  3. ICCAD
    block_level_partition0517_new.png
    Accel-GCN: High-Performance GPU Accelerator Design for Graph Convolution Networks
    Xi Xie, Hongwu Peng, Amit Hasan, and 7 more authors
    In 2023 IEEE/ACM International Conference on Computer Aided Design (ICCAD), 2023
    HTML PDF Code