Rong Ou
Principal Systems Software Engineer at NVIDIA
Palo Alto, California, United States
Summary
🤩
Rockstar🎓
Top SchoolRong Ou is a Principal Systems Software Engineer with 13+ years building high-performance AI/ML infrastructure and distributed systems, currently leading GPU-accelerated ML work at NVIDIA from Palo Alto. He combines deep systems roots from Google and Sabre with hands-on contributions to prominent open-source projects like Spark RAPIDS, XGBoost, cuDF, RMM and Thrust, driving features such as arena memory allocators, GPUDirect Storage, out-of-core GPU training and federated XGBoost. Rong’s work spans low-level CUDA optimizations, orchestration (Kubernetes MPI Operator), and production-scale ML pipelines, enabling much larger and faster training on GPUs while reducing fragmentation and I/O overhead. He has a track record of inventing practical algorithms (bit-mask based federated learning) and shipping them into production and industry adoption, including financial and medical-imaging use cases. A skilled technical leader, he mentors teams across continents and authored design proposals that seeded new projects and patents. His background in physics (Peking University) and an MS in CS from UT Austin inform a methodical, performance-first approach to complex engineering trade-offs.
13 years of coding experience
21 years of employment as a software developer
Master of Science (M.S.) Computer Science, Master of Science (M.S.) Computer Science at The University of Texas at AustinBachelor of Arts (B.A.) Physics, Bachelor of Arts (B.A.) Physics at Peking University
English, Spanish, Chinese
Stackoverflow
81reputation
23kreached
1answer
0questions
Github Skills (51)
thrust10
apache-spark10
kubernetes10
memory-allocation10
c-language10
strings10
gpgpu10
jvm10
back-end-development10
python10
data-engineering10
memory-management10
mpi10
machine-learning10
java10
Programming languages (19)
Github contributions (5)
kubeflow/mpi-operator
May 2018 - Jan 2021
Kubernetes Operator for MPI-based applications (distributed training, HPC, etc.)
Role in this project:
Back-end & DevOps Engineer
Back-end & DevOps EngineerContributions:1 release, 33 commits, 37 PRs in 2 years 8 months
Contributions summary:Rong primarily contributed to the core functionality of the MPI operator by implementing features, addressing review feedback, and fixing minor issues within the controller logic. They made changes related to GPU resource allocation, including the ability to specify GPUs explicitly. The user also updated dependencies and configurations, demonstrating an understanding of the project's infrastructure and deployment. These contributions show a focus on refining the operator's capabilities and maintainability.
pytorchmpihorovoddistributed-applicationsoperator
rapidsai/rmm
Apr 2020 - Mar 2022
RAPIDS Memory Manager
Role in this project:Backend Developer
Backend DeveloperContributions:231 reviews, 88 commits, 49 PRs in 1 year 11 months
Contributions summary:Rong primarily contributed to the `rmm` repository, focusing on memory management within the RAPIDS ecosystem. Their commits involved adding support for stdout/stderr logging, enhancing the logging resource adaptor, and updating build scripts related to Anaconda uploads. Further contributions include implementing an option to automatically flush logs and incorporating a warning for performance considerations.
cudamemory-managementmemorycpppython
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.