Elliot Gorokhovsky
Undergraduate Research Assistant
New York, New York, United States
Summary
🤩
Rockstar🎓
Top SchoolElliot Gorokhovsky is a software engineer and undergraduate researcher with 11 years of hands-on experience in performance-oriented backend development and applied research. Based in New York and affiliated with Arizona State University's SEFCOM lab, he contributes to angr by expanding its simulated Linux syscall library and brings deep low-level expertise from work on real-time compression (facebook/zstd) and the Triton compiler. His contributions center on multi-threading, memory management, and performance tuning—skills demonstrated by implementing memory-limits for dictionary training and refining TMA descriptor handling to improve GPU code generation. Elliot also has a strong theoretical background in algorithms and computational economics, having proved the first hardness-of-approximation result for min-sum k-clustering and developed reductions for networked market equilibria. Comfortable bridging research and production, he combines rigorous mathematical thinking with practical systems engineering on widely used open-source projects.
11 years of coding experience
1 year of employment as a software developer
Bachelor's degree, Mathematics, Bachelor's degree, Mathematics at Arizona State University
Stackoverflow
3,630reputation
182kreached
15answers
112questions
Github Skills (19)
benchmark10
python10
multithreading10
compression-algorithm10
memory-management10
benchmarking10
c1110
c1710
nvidia10
triton10
performance-optimization10
cuda10
compiler-design9
scheme6
sicp6
Programming languages (10)
Github contributions (5)
facebook/zstd
Dec 2021 - Jan 2023
Zstandard - Fast real-time compression algorithm
Role in this project:
Back-end Developer & Performance Engineer
Back-end Developer & Performance EngineerContributions:132 reviews, 102 commits, 119 PRs in 1 year 1 month
Contributions summary:Elliot primarily contributed to optimizing the Zstandard compression algorithm, specifically focusing on improving the memory management and performance of multi-threaded compression. They implemented a memory limit option for dictionary training, and refactored the MT memory management code. Their work also included bug fixes and performance improvements related to the largeNbDicts benchmark. The user's contributions suggest a deep understanding of compression algorithms and multi-threading optimization.
deflatezstandardbzip2compressionzstd
triton-lang/triton
Feb 2023 - Feb 2025
Development repository for the Triton language and compiler
Role in this project:Backend Developer & Performance Engineer
Backend Developer & Performance EngineerContributions:25 reviews, 8 PRs, 3 pushes in 2 years
Contributions summary:Elliot contributed to the Triton language and compiler by fixing bugs related to TMA descriptor cache management and optimizing performance. They addressed issues with the ptxas compiler, ensuring the correct repro steps and information were available during ptxas segfaults. Additionally, the user implemented support for passing TMA descriptors by-value, which improved performance, and made the TMA tests compatible with older CUDA toolchains. Furthermore, they fixed a bug in the persistent matrix multiplication tutorial.
compilerprogramming-languagecode-generationtriton
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.