Sean Kelley
Principal System Software Architect, Servers at NVIDIA
Portland, Oregon, United States
Summary
👤
Senior🎓
Top SchoolSean Kelley is a principal system software architect with over 15 years building platform-level software across servers, clients, real-time Linux, media/video, and embedded devices. He combines deep hardware and silicon insight—rooted in early ASIC and avionics work—with broad software leadership, having driven Linux platform and video codec strategy at Intel and performance architecture at NVIDIA. Sean has led large engineering teams to ship consumer products (notably Garmin’s nüvi line) and influenced silicon planning to make real-time features ubiquitous. His open-source contributions include performance and DMA-focused work on Stanford Legion and low-level fixes to the Terra language toolchain, reflecting a knack for systems-level performance tuning and parallel I/O. Based in Portland, he excels at translating domain expert needs into practical, layered software architectures across diverse architectures and use cases. An uncommon strength is his end-to-end fluency from VHDL/ASIC modeling to kernel and user-space optimization, enabling pragmatic decisions that align hardware, firmware, and software roadmaps.
15 years of coding experience
30 years of employment as a software developer
University of Kansas
French, Irish, Welsh, Scottish Gaelic, English
Github Skills (14)
lua10
c-language10
parallel-computing10
cprogramming-language10
performance-optimization10
llvm9
compiler-design9
memory-management9
code-optimization8
architecture8
freebsd8
multithreading8
architectures8
build-system7
Programming languages (6)
Github contributions (5)
StanfordLegion/legion
Apr 2015 - Feb 2022
The Legion Parallel Programming System
Role in this project:
Back-end Developer & Performance Engineer
Back-end Developer & Performance EngineerContributions:10 releases, 1628 commits, 80 PRs in 6 years 10 months
Contributions summary:Sean primarily focused on optimizing and extending the functionality of the Realm parallel programming system, specifically in the area of data transfer operations. They implemented solutions for handling empty copy cases and ensuring proper execution of multi-dimensional data transfers within the system's DMA channel. The user addressed issues related to HDF5 file transfers, indicating strong familiarity with parallel I/O frameworks. Their contributions also include efforts to enhance performance through improvements to memory allocation and instance layout.
system-programminglegionparallel-programmingparallel
terralang/terra
Dec 2015 - Jan 2018
Terra is a low-level system programming language that is embedded in and meta-programmed by the Lua programming language.
Role in this project:Back-end Developer
Back-end DeveloperContributions:18 commits, 8 PRs, 3 comments in 2 years 1 month
Contributions summary:Sean primarily contributed to fixing compiler warnings and addressing code issues within the Terra programming language project. Their work involved modifying source code files, including `lparser.cpp`, `lobject.cpp`, and `tcompiler.cpp`, to improve code quality and address potential problems. Further, the user made adjustments to internal LuaJIT code used by Terra, and fixed prototype issues for the bytecode loading functions. Additional commits include fixing signedness warnings, and fixing build errors for FreeBSD.
luac99system-programmingmetacompiler
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.