Nikolay Sakharnykh
Director, Developer Technology at NVIDIA
New York, New York, United States
Summary
🤩
Rockstar🎓
Top SchoolNikolay Sakharnykh is a Director of Developer Technology at NVIDIA with 15+ years of hands-on experience optimizing CPU/GPU performance, designing parallel algorithms, and delivering HPC and AI solutions. He has progressed through engineering and leadership roles at NVIDIA across multiple global locations, combining deep low-level systems expertise with team leadership and product-focused delivery. Nikolay contributes to prominent open-source projects like rapidsai/cudf, where his low-level fixes and refactors improved correctness and performance of multi-column GPU joins. With an MS in Applied Mathematics and Programming from MSU, he blends rigorous academic training with practical kernel- and API-level engineering. Colleagues rely on him to translate hardware architecture insights into performant software and scalable developer tooling.
14 years of coding experience
16 years of employment as a software developer
Lomonosov Moscow State University
Github Skills (11)
kernel10
dataframes10
cuda10
c-language10
cpp-templates10
cprogramming-language10
gpu10
dataframe10
c-templates10
data-structure9
data-structures9
Programming languages (5)
Github contributions (5)
rapidsai/cudf
Apr 2018 - Jun 2019
cuDF - GPU DataFrame Library
Role in this project:
Back-end Developer
Back-end DeveloperContributions:4 reviews, 130 commits, 10 PRs in 1 year 2 months
Contributions summary:Nikolay contributed to the cuDF GPU DataFrame Library, focusing on low-level code improvements related to joining operations. They addressed correctness issues in multi-column joins by replacing `thrust::pair` with a bare `pair` class. Further contributions involved refactoring and improving the multi-column join implementation for up to 3 columns of int32/int64 data types. The user's work included API clean-up and kernel modifications, and their efforts are aimed at enhancing the performance and correctness of the core data processing functionalities within the library.
cudadataframe-librarydata-analysiscppcudf
NVIDIA/CoMD-CUDA
Nov 2013 - Jan 2017
Contributions:27 commits, 1 PR, 11 pushes in 3 years 3 months
molecular-dynamicsproxygpumoleculardynamics
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.