Kumar Aatish
Analytics Libraries CUDA Software Engineer at NVIDIA
Atlanta, Georgia, United States
Summary
🤩
Rockstar🎓
Top SchoolKumar Aatish is an Analytics Libraries CUDA Software Engineer with 11 years of experience building high-performance parallel computing solutions for GPUs, including sustained contributions to the RAPIDS open-source ecosystem (cuDF and cuGraph). At NVIDIA he develops and tests GPU DataFrame and graph analytics features, notably implementing multi-left join behavior and refactoring core C++/CUDA graph structures to improve performance and correctness. His background spans DARPA- and DOE-funded projects delivering multi-GPU clustering, dynamic graph libraries (Hornet), and FFT/ptychography accelerations that achieved tens of times speedups over CPU baselines. Comfortable across CUDA, OpenCL, and CPU-parallel models like OpenMP and MPI, he combines deep systems-level optimization with hands-on test automation. Based in Atlanta, he pairs academic research experience from Georgia Tech and UC San Diego with production-grade engineering at ArrayFire and NVIDIA. Not obvious from titles: much of his impact comes from improving core data structures and tests that make GPU analytics libraries both faster and more reliable at scale.
11 years of coding experience
4 years of employment as a software developer
University of California San DiegoBachelor of Engineering (B.E.), Bachelor of Engineering (B.E.) at Birla Institute of Technology and Science
English, Hindi
Stackoverflow
101reputation
820reached
2answers
0questions
Github Skills (19)
algorithm10
algorithms10
graph-algorithms10
c-language10
testing10
dataframes10
gpu-programming10
cudf10
data-structure10
dataframe10
cuda10
data-structures10
datastructures-algorithms10
cpp10
cprogramming-language10
Programming languages (11)
Github contributions (5)
rapidsai/cugraph
Mar 2019 - Mar 2022
cuGraph - RAPIDS Graph Analytics Library
Role in this project:
Back-end Developer
Back-end DeveloperContributions:82 reviews, 227 commits, 46 PRs in 3 years
Contributions summary:Kumar's contributions focused on documentation and improvements of the cuGraph RAPIDS Graph Analytics Library, specifically the CPP code base related to the functions.h and core logic for the graph representations. They refactored, improved, and added functionality to manage the graph's data structures.
cudagraph-analysisanalyticsgraph-analyticsgraphml
rapidsai/cudf
Jul 2018 - May 2021
cuDF - GPU DataFrame Library
Role in this project:Back-end Developer & Test Automation Engineer
Back-end Developer & Test Automation EngineerContributions:80 reviews, 185 commits, 42 PRs in 2 years 10 months
Contributions summary:Kumar's contributions centered on developing and testing GPU DataFrame library functionality. They implemented and tested multi-left join features, focusing on code modifications in C++ (cuDF) and leveraging CUDA for GPU processing. Their work included adding and expanding tests to validate the correct functionality of these join operations, ensuring the library's performance and reliability.
cudadataframe-librarydata-analysiscppcudf
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.