Jianfei Wang
AI Infra工程师 at 小红书
Pudong, Shanghai, China
Summary
🤩
Rockstar🎓
Top SchoolJianfei Wang is an AI infrastructure engineer with 8 years of experience specializing in GPU-accelerated HPC and deep learning runtimes, currently building AI infra at Xiaohongshu in Pudong, Shanghai. He holds a PhD in Computer Science from Shanghai Jiao Tong University and began his engineering career after a strong automation foundation from Huazhong University of Science and Technology. Previously at SenseTime he focused on CUDA-driven performance and correctness, shipping bug fixes and new operator support in production neural-net primitives. Jianfei is an active open-source contributor to ppl.nn, where he improved CUDA implementations, added ONNX unary operators and optimized format/type conversions for better throughput. His work blends low-level GPU engineering with practical ML operator compatibility—an engineer who turns algorithmic detail into measurable system speedups.
8 years of coding experience
4 years of employment as a software developer
学士, 自动化, A, 学士, 自动化, A at 华中科技大学博士, Computer Science, 博士, Computer Science at 上海交通大学
Github Skills (16)
neural-network10
cuda10
gpu-programming10
c-language10
deep-learning10
cprogramming-language10
onnx10
optimization9
operator9
tensorflow9
optmization9
operation9
optimisation9
performance-optimization9
tensorrt9
Programming languages (3)
Github contributions (5)
OpenPPL/ppl.nn
Apr 2022 - Oct 2022
A primitive library for neural network
Role in this project:
Back-end Developer
Back-end DeveloperContributions:1 review, 19 commits, 83 PRs in 5 months
Contributions summary:Jianfei primarily contributed to improving the CUDA implementation for the ppl.nn library, focusing on bug fixes and adding new functionalities. Their work included resolving issues in arithmetic operations, softmax calculations, and pooling algorithms. The user also added support for new ONNX operators, specifically focusing on unary operations like `abs` and `round`. Furthermore, they optimized the library by implementing format and type conversions for performance improvements.
neural-networkdeep-learningonnx
jianfei-wangg/ppl.kernel.cuda
Dec 2022 - Dec 2023
Contributions:49 pushes, 49 branches in 11 months
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.