Zhiyi Zhang
my photo

Hi there!
I was born in 2000, raised in Anhui province, China. Currently, I am a thrid-year Master student at the University of Science and Technology of China (USTC), and I will start my Doctoral career after September 2025.
My research interests include Deep Learning, Parallel Computing, Computer Architecture. Recently, I pay more attentions on Large Language Models (LLMs). Let's chat if you are into them too. My interests might evovle, but I sincerely want to stay curious and keep learning.
Beyond research, I like sports, art, computer games, delicious foods, science, military, politics ... I am out-going and always happy to make new friends!

mail-icon   gilgamesh@mail.ustc.edu.cn
github-icon   https://github.com/GilgameshXYZ123

Education

Sep. 2025 - Future          University of Science and Technology of China

Doctoral Researcher in Computer Science

Sep. 2022 - Jun. 2025      University of Science and Technology of China

Master of Pattern Recognition and Intelligent System

Sep. 2018 - Jun. 2022      Hainan University, China

Bachelar of Software Engineering


First-Author Publications

pdf-link 2025    WinRS: Accelerate Winograd Backward-Filter Convolution with Tiny Workspace

Zhiyi Zhang, Junshi Chen, Jingwei Sun, Pengfei Zhang, Zhuopin Xu, Junshi, Qi Wang.

Accepted by International Conference on Parallel Processing (ICPP 2025).

pdf-link 2025    Common Feature Representation Learning for Multi-Crop Seed Classification via Hyperspectral Images

Zhiyi Zhang, Jiading Yuan, Pengfei Zhang, Zhuopin Xu, Qi Wang.

In Submission.

pdf-link 2024    Im2col-Winograd: An Efficient and Flexible Fused-Winograd Convolution for NHWC Format on GPUs

Zhiyi Zhang, Pengfei Zhang, Zhuopin Xu, Bingjie Yan, Qi Wang.

Accepted by International Conference on Parallel Processing (ICPP 2024)

pdf-link 2023    Reduce Computational Complexity for Convolutional Layers by Skipping Zeros

Zhiyi Zhang, Pengfei Zhang, Zhuopin Xu, Bingjie Yan, Qi Wang.

IEEE International Conference on High Performance Computing, Data, and Analytics (HiPC 2023)

pdf-link 2023    Dragon-Alpha&Cu32: Open-source Java Deep-Learning Framework and High-Performance GPU Library

Zhiyi Zhang, Pengfei Zhang, Zhuopin Xu.

Arix Preprint.