Xinxin Wang
Postdoctoral Scholar at Stanford University
Palo Alto, California, United States
Summary
🤩
Rockstar🎓
Top SchoolXinxin Wang is a postdoctoral scholar in Palo Alto specializing in in-memory computing architectures for AI acceleration, with seven years of experience bridging device-level non-idealities and system-level DNN performance. She designed a tiled architecture for DNN inference, quantified the effects of bitcell imperfections and ADC quantization on model accuracy, and proposed practical bitcell requirements and multi-range ADC schemes to reduce computational error. Her work balances layer-level throughput via weight-array duplication and ADC multiplexing, and includes end-to-end performance analyses (latency, power, capacity, TOPS/W) for ResNet50, MobileNet, Inception-V4, and Transformer models. At Stanford following a PhD from the University of Michigan and internship experience validating RRAM-based accelerators at Applied Materials, she blends hands-on lab validation, architecture-aware training, and AI simulation—bringing both experimental rigor and system-level insight to hardware-aware ML.
7 years of coding experience
Doctor of Philosophy - PhD, Electrical Engineering and Computer Science, Doctor of Philosophy - PhD, Electrical Engineering and Computer Science at University of MichiganBachelor of Science - BS, Microelectronics Science and Engineering, Bachelor of Science - BS, Microelectronics Science and Engineering at Peking University
Github Skills (14)
neuroscience10
spike10
spiking-neural-networks10
python9
tensorflow9
machine-learning9
deep-learning9
incremental-learning9
pytorch9
neural-network6
benchmarking3
hardware2
autograd2
benchmark2
Programming languages (2)
Github contributions (4)
xxwang1/DNN-accelerator-based-on-tiled-architecture
Jan 2019 - Apr 2020
Contributions:8 commits, 5 pushes, 5 branches in 1 year 3 months
jeshraghian/snntorch
Oct 2020 - Apr 2021
Deep and online learning with spiking neural networks in Python
Contributions:28 commits, 19 pushes, 1 branch in 6 months
pytorchspiking-neural-networksneurosciencepythondeep-learning
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.