Dmitry Zarukin
Deep Learning Software Engineer at Intel Corporation
Hillsboro, Oregon, United States
Summary
🤩
Rockstar🎓
Top SchoolDmitry Zarukin is a Deep Learning Software Engineer with seven years of experience building high-performance numerical libraries at Intel, currently contributing to the oneAPI Deep Neural Network Library. With a strong mathematical foundation from a Master's at Novosibirsk State University, he specializes in C++, performance optimization, and low-level algorithmic improvements for convolution, layer normalization, and matrix multiplication. His work on oneDNN emphasizes code refactoring and IO/helper abstractions to squeeze kernel efficiency and remove dead code, reflecting an engineer-focused approach to sustainable performance gains. Prior experience on the Intel Math Kernel Library rounds out his expertise in optimized linear algebra and computational kernels for semiconductor-grade workloads. Based in Hillsboro, Oregon, he blends rigorous academic training with hands-on production engineering for large-scale ML runtimes. Colleagues appreciate that he looks for small structural changes that produce outsized performance improvements.
7 years of coding experience
3 years of employment as a software developer
Master's degree, Mathematics, 4.3, Master's degree, Mathematics, 4.3 at Novosibirsk State University (NSU)
Stackoverflow
1reputation
0reached
0answers
0questions
Github Skills (17)
normalize10
c-language10
matrix-multiplication10
refactor10
normalization10
computer-engineering10
performance-optimization10
convolution10
normalizing10
refactoring10
cprogramming-language10
numerical9
algorithm9
bfd9
algorithms9
Programming languages (4)
Github contributions (5)
uxlfoundation/oneDNN
Mar 2019 - Jan 2023
oneAPI Deep Neural Network Library (oneDNN)
Role in this project:
Back-end Developer
Back-end DeveloperContributions:867 reviews, 1524 commits, 218 PRs in 3 years 11 months
Contributions summary:Dmitry's commits primarily involve code refactoring and optimization within the oneDNN library. The code changes focus on migrating read and write operations to an io_helper to improve kernel efficiency, as well as removing dead code. Additionally, the user's work appears to be focused on optimizing existing code, specifically in the areas of convolution, layer normalization, and matrix multiplication to improve performance.
bfloat16sse41avx512openmpopencl
dzarukin/oneDNN
Nov 2020 - Sep 2024
oneAPI Deep Neural Network Library (oneDNN)
Contributions:68 pushes, 41 branches in 3 years 11 months
deep-learningdeep-neural-networkoneapineural-networkonednn
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.