Xiaobing Zhang
Machine Learning Engineer at Intel Corporation
Minhang District, Shanghai, China
Summary
🤩
Rockstar🎓
Top SchoolXiaobing Zhang is a Machine Learning Engineer with a decade of experience specializing in optimizing PyTorch deep learning workloads for Intel hardware platforms. Based in Minhang District, Shanghai, he contributes to high-profile open-source efforts such as Intel’s AI Reference Models, where he has improved ResNet50 inference and training performance across precisions like FP32 and BF16 and adapted code for evolving IPEX APIs. His background combines engineering and management degrees with advanced study in information theory and communications, reflecting a rare mix of systems-level optimization and theoretical grounding. Known for squeezing throughput out of real-world deployments, he focuses on making models production-ready on Intel Xeon and Data Center GPUs.
10 years of coding experience
Bachelor of Engineering, Bachelor of Management, Electronic and Information Engineering, Business Administration, Bachelor of Engineering, Bachelor of Management, Electronic and Information Engineering, Business Administration at University of Petroleum (East China)Visiting student, Information Theory, Visiting student, Information Theory at UC Berkeley Electrical Engineering & Computer Sciences (EECS)Master of Engineering - MEng, Communication and Information System, Master of Engineering - MEng, Communication and Information System at Shanghai Institute of Microsystems and Information Technology, Chinese Academy of Sciences
Github Skills (8)
pytorch10
inference10
deep-learning10
ai9
performance-monitor9
performance-analysis9
cpu8
tensorflow3
Programming languages (4)
Github contributions (5)
intel/ai-reference-models
Nov 2020 - Nov 2022
Intel® AI Reference Models: contains Intel optimizations for running deep learning workloads on Intel® Xeon® Scalable processors and Intel® Data Center GPUs
Role in this project:
ML Engineer
ML EngineerContributions:39 commits in 2 years
Contributions summary:Xiaobing primarily focused on integrating and optimizing PyTorch models within the Intel AI Reference Models repository. Their contributions involved adding and updating ResNet50 models for inference, including support for different precisions like FP32 and BF16. Further work includes updating model code for newer ipex API changes, and optimizing throughput for Resnet50 model, along with enabling and updating models for training, indicating a focus on model deployment and performance.
optimizationsprocessorstensorflowzoomodel-zoo
juliusshufan/pytorch-1
Apr 2019 - Aug 2019
Tensors and Dynamic neural networks in Python with strong GPU acceleration
Contributions:42 commits in 4 months
pythongpu-accelerationdeep-learninggpuacceleration
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.