Louis Feng
Software Engineer at Facebook
Palo Alto, California, United States
Summary
🤩
Rockstar🎓
Top SchoolLouis Feng is a software engineer and technical architect based in Palo Alto with 13 years of experience focused on CPU/GPU performance and AI infrastructure. He combines deep performance engineering expertise—demonstrated by contributions to PyTorch profiler and execution tracing—with low-level graphics and runtime work such as improving geometry correctness in the ospray renderer and CPU optimizations in nGraph. Comfortable across ML frameworks, compilers, and rendering engines, he excels at diagnosing subtle numerical and layout bugs that impact throughput and correctness. Louis’s background suggests a strong ability to bridge research-grade systems and production tooling, often surfacing non-obvious execution details like operator sequencing and device-aware tracing to drive measurable performance gains.
13 years of coding experience
4.0/4.0, 4.0/4.0 at University of California, Davis
English, Chinese
Stackoverflow
1reputation
0reached
0answers
0questions
Github Skills (32)
pytorch10
visualization10
c-language10
python10
rendering10
raytracing10
machine-learning10
render10
pov-ray10
deep-learning10
tensorflow10
performance-optimization10
backpropagation10
profiling10
convolution10
Programming languages (6)
Github contributions (5)
NervanaSystems/ngraph
Feb 2018 - Jul 2019
nGraph - open source C++ library, compiler and runtime for Deep Learning
Role in this project:
Back-end Developer
Back-end DeveloperContributions:74 commits, 21 PRs, 199 pushes in 1 year 4 months
Contributions summary:Louis primarily contributed to the back-end development of the nGraph library, focusing on adding and improving CPU runtime functionality. Their work included implementing and optimizing convolution backpropagation for the CPU, and adding support for fusing convolution with bias operations. The user also introduced dynamic reshape and slice operations, indicating a focus on enhancing the library's capabilities in dynamic shape processing for deep learning workloads. Furthermore, the user addressed bugs and layout issues in the CPU emitter, ensuring the correctness and performance of the nGraph runtime.
inference-enginecppc-librarydeep-learningtvm
pytorch/pytorch
Sep 2020 - Oct 2022
Tensors and Dynamic neural networks in Python with strong GPU acceleration
Role in this project:Back-end Developer & ML Engineer
Back-end Developer & ML EngineerContributions:51 reviews, 29 commits, 21 PRs in 2 years 1 month
Contributions summary:Louis contributed significantly to the PyTorch framework by fixing issues related to the `record_function` feature and integrating the Execution Graph Observer into the PyTorch Profiler. They addressed checks within the record function and integrated the Execution Graph Observer, which captures the execution flow for performance analysis. Additionally, they improved the Execution Graph by adding sequence numbers to map forward and backward operators, recording tensor devices, and refactoring the GlobalStateManager. Furthermore, they enabled the capturing of parameters for communications and refactored the Execution Graph to Execution Trace.
pythongpu-accelerationdeep-learninggpunumpy
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.