Jared Casper
Senior Deep Learning Research Scientist at NVIDIA
Palo Alto, California, United States
Summary
🤩
Rockstar🎓
Top SchoolJared Casper is a Senior Deep Learning Research Scientist with 17 years of experience blending foundational research and production-grade engineering at organizations including NVIDIA, Baidu, and Oracle Labs. He holds a Ph.D. from Stanford and a BS from MIT and focuses on scaling transformer training and dependable ML infrastructure—contributions to NVIDIA’s Megatron-LM underline his work on PyTorch RNG state handling and large-model sampling. Jared also has a strong systems and compiler background, evidenced by substantive back-end contributions to the Icarus Verilog project that improved compatibility and added SystemVerilog 2009 features. Based in Palo Alto, he operates at the intersection of deep learning research and low-level software reliability, bringing academic rigor to pragmatic, shipping systems.
17 years of coding experience
11 years of employment as a software developer
B.S., Electrical Engineering and Computer Science, B.S., Electrical Engineering and Computer Science at Massachusetts Institute of TechnologyPh.D., Computer Science, Ph.D., Computer Science at Stanford University
Stackoverflow
170reputation
11kreached
7answers
1question
Github Skills (20)
pytorch10
verilog10
c-language10
systemverilog10
machine-learning10
compiler-design10
cprogramming-language10
memory-management9
computer-engineering9
gcc8
api-management7
socket6
emulation6
android6
tcp6
Programming languages (6)
Github contributions (5)
NVIDIA/Megatron-LM
Sep 2019 - Jan 2023
Ongoing research training transformer models at scale
Role in this project:
ML Engineer & Data Scientist
ML Engineer & Data ScientistContributions:6 releases, 2 reviews, 351 commits in 3 years 4 months
Contributions summary:Jared's contributions primarily focused on supporting and improving the PyTorch RNG state API and updating the code base with latest changes, implying a role in maintaining or optimizing the codebase. The user also addressed a specific bug and added code to generate samples. These contributions suggest involvement in both model functionality and related software engineering aspects of the project.
pytorchnlplanguage-modeltransformer-modelsbert
steveicarus/iverilog
Sep 2008 - Jul 2013
Icarus Verilog
Role in this project:Back-end Developer
Back-end DeveloperContributions:22 commits in 4 years 10 months
Contributions summary:Jared primarily contributed to improving the Icarus Verilog compiler. They fixed compiler compatibility issues with different GCC versions by including necessary header files and addressing memory leaks in specific functions. Furthermore, the user implemented the addition of new features, such as the ability to generate SystemVerilog 2009 constructs, along with corresponding tokens and language elements. They also addressed the correction of the system tasks and some of its arguments.
icarusicarus-verilogverilog
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.