王雪莹 北京邮电大学

Xueying Wang, Beijing University of Posts and Telecommunications, Posdoctor Researcher

Brief Biography

Wang Xueying focuses on parallel computing and compiler optimization for machine learning applications, including graph compilation optimization (TACO, CCF-A, 2025), quantized Winograd convolution optimization(TACO, CCF-A, 2024), hardware-aware NAS tuning (JSA, CCF-B, 2023), memory scheduling system(TACO, CCF-A, 2023), and operator fusion(Euro-par, CCF-B, 2020). Wang Xueying is also the reviewer for Neurocomputing, Journal of Systems Architecture, IEEE Cluster, IEEE TCAD, and other journals. Wang Xueying is passionate about exploring how to improve parallel computing performance to solve complex computational problems, and how compiler optimizations can further enhance the performance of supercomputing systems.

Work and Education Experience

• Postdoctor at School of Computer science, Beijing University of Posts and Telecommunications, 2023 - Now (Co-advisor: Li Shigang )
• Ph.D in Computer System Architecture, Institute of Computing Technology, Chinese Academy of Sciences, 2017 - 2023 (Advisor: Feng Xiaobing)
• B.S. in Software Engineering, Northeast Normal University (Projecti 211), 2013 - 2017

Student Competition News

• Congratulations! We have qualified for the ASC Student Challenge 2025 Final Round. It’s our first time participating in the ASC Student Challenge competition, and also the first time for BUPT appearing on the finalists list.
• Congratulations again for our outstanding performance at ASC Student Challenge Final Round, we got the third place and achieved the First Prize!
• Our group ‘OneLastCompiler’ consist of 3 undergraduates achieved Third Price at National College Students Computer System Ability Competition - Compilation system design Competition(2024全国大学生计算机系统能力大赛编译系统设计赛全国总决赛).

Research Projects

• China Postdoctoral Science Foundation, 75th China Postdoctoral Science Foundation(75批博后面上), 2024M750258, 2024-07 to date, 80,000 yuan, in research, Host. (Acceptance Rate: 15%)

Representative Publications

• [TACO’25, CCF-A] OptiFX: Automatic Optimization for Convolutional Neural Networks with Aggressive Operator Fusion on GPUs. Xueying Wang, Shigang Li, Hao Qian, Fan Luo, Zhaoyang Hao, Tong Wu, Ruiyuan Xu, Huimin Cui, Xiaobing Feng, Guangli Li, Jingling Xue. ACM Transactions on Architecture and Code Optimization, 2025.

• [TACO’24, CCF-A] Fast convolution meets low precision: Exploring efficient quantized Winograd convolution on modern CPUs[J]. Xueying Wang, Guangli Li, Zhen Jia, Xiaobing Feng, Yida Wang. ACM Transactions on Architecture and Code Optimization, 2024.

• [TACO’22, CCF-A]An application-oblivious memory scheduling system for DNN accelerators[J]. Jiansong Li*, **Xueying Wang* **(Equal Contribution), Xiaobing Chen, Guangli Li, Xiao Dong, Peng Zhao, Xianzhi Yu, Yongxin Yang, Wei Cao, Lei Liu, Xiaobing Feng. ACM Transactions on Architecture and Code Optimization, 2022.

• [JSA’23, CCF-B] Facilitating hardware-aware neural architecture search with learning-based predictive models[J]. Xueying Wang, Guangli Li, Xiu Ma, Xiaobing Feng. Journal of Systems Architecture, 2023.

• [JSA’23, CCF-B] CoAxNN: Optimizing on-device deep learning with conditional approximate neural networks[J]. Guangli Li, Xiu Ma, Qiuchu Yu, Lei Liu, Huaxiao Liu, Xueying Wang (Corresponding author). Journal of Systems Architecture, 2023.

• [EuroPar’20, CCF-B] Accelerating deep learning inference with cross-layer data reuse on GPUs[C]. Xueying Wang, Guangli Li, Xiao Dong, Jiansong Li, Lei Liu, Xiaobing Feng. European Conference on Parallel Processing, 2020.