Yutian Li
Core Developer at Hudson River Trading
New York, New York, United States
Summary
🤩
Rockstar🎓
Top SchoolYutian Li is a core developer with 13 years of experience building high-performance back-end and systems software, currently focused at Hudson River Trading in New York. He has deep expertise in distributed machine learning infrastructure and low-level memory/concurrency engineering, contributing to prominent open-source projects like MXNet and dmlc-core where he implemented storage managers, CUDA debugging utilities, spinlocks, and concurrent queues. His background spans research and industry — from Microsoft Research Asia and Stanford to Tesla and Face++ — reflecting a blend of production-grade engineering and academic rigor. Comfortable across C++/Python toolchains and shell tooling (notably prezto enhancements), he quietly optimizes developer experience and performance, for example by caching brew shellenv results and enabling editor keybindings.
13 years of coding experience
2 years of employment as a software developer
Exchange Student, Computer Science, Junior, Exchange Student, Computer Science, Junior at The University of Texas at AustinBachelor of Engineering (B.E.), Computer Science, Bachelor of Engineering (B.E.), Computer Science at Tsinghua UniversityMaster's Degree, Computer Science, Master's Degree, Computer Science at Stanford University
English, Chinese
Github Skills (31)
algorithm10
algorithms10
c-language10
scripting10
memory-management10
mxnet10
dm10
data-structure10
zshrc10
script10
spinlock10
cuda10
data-structures10
sh10
datastructures-algorithms10
Programming languages (22)
Github contributions (5)
dmlc/minerva
May 2014 - Jul 2015
Minerva: a fast and flexible tool for deep learning on multi-GPU. It provides ndarray programming interface, just like Numpy. Python bindings and C++ bindings are both available. The resulting code can be run on CPU or GPU. Multi-GPU support is very easy.
Role in this project:
Back-end Developer
Back-end DeveloperContributions:613 commits, 6 PRs, 171 pushes in 1 year 2 months
Contributions summary:Yutian appears to be working on a deep learning tool for multi-GPU acceleration. The commits demonstrate the implementation of a chunk-level operator, suggesting the development of core functionalities within the framework. The changes involve defining and modifying classes and data structures, specifically related to DAG nodes and operations.
pythonndarraygpu-supportpython-numpymulti-gpu
apache/mxnet
Aug 2015 - May 2016
Lightweight, Portable, Flexible Distributed/Mobile Deep Learning with Dynamic, Mutation-aware Dataflow Dep Scheduler; for Python, R, Julia, Scala, Go, Javascript and more
Role in this project:Back-end Developer & Systems Engineer
Back-end Developer & Systems EngineerContributions:91 commits, 9 PRs, 51 pushes in 9 months
Contributions summary:Yutian made several commits focused on the storage backends within the MXNet deep learning framework. Their work included implementing CUDA debugging utilities and refining memory allocation strategies across CPU and GPU devices. The user's changes involved defining and implementing storage managers, including naive and pooled storage managers, and addressing low-level memory management aspects for optimized performance. These modifications directly contributed to improving the framework's ability to utilize different hardware resources effectively.
pythonschedulerdataflowmutationdata-science
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.