Adam Straw
Senior Software Engineer at NVIDIA
Hillsboro, Oregon, United States
Summary
🤩
Rockstar🎓
Top SchoolAdam Straw is a senior software engineer and performance-focused architect with nine years of experience building and optimizing large-scale C++ systems for deep learning and graphics at companies like Intel, OctoML, and NVIDIA. He specializes in compiler and backend work—improving MLIR/LLVM-based toolchains and accelerating kernels across CPU, GPU, and specialized accelerators—with notable open-source contributions to PlaidML and TVM that improved GEMM performance, profiling, and Vulkan support. A proven technical leader and mentor, he has driven product strategy and cross-organizational forums (nGraph, Maestro) while partnering with customers such as Microsoft and Waymo to deliver real-world solutions. Passionate about making AI efficient and practical, he combines rigorous engineering with a collaborative, people-first approach and a sense of humor that helps build productive teams.
9 years of coding experience
21 years of employment as a software developer
Bachelor of Science (B.S.) with Distinction (Magna Cum Laude), Computer Engineering, Bachelor of Science (B.S.) with Distinction (Magna Cum Laude), Computer Engineering at Iowa State UniversityIntro to Artificial Intelligence - Stanford CS 271, Intro to Artificial Intelligence - Stanford CS 271 at Udacity
Stackoverflow
1reputation
0reached
0answers
0questions
Github Skills (22)
c-language10
matrix-multiplication10
machine-learning10
mlr10
hexagons10
deep-learning10
performance-optimization10
backpropagation10
compiler-optimization10
performance-tuning10
cprogramming-language10
spirv9
vulkan9
affine-transformation9
test-automation9
Programming languages (4)
Github contributions (5)
NervanaSystems/ngraph
Dec 2017 - Jun 2020
nGraph has moved to OpenVINO
Role in this project:
Back-end Developer
Back-end DeveloperContributions:75 commits, 96 PRs, 279 pushes in 2 years 6 months
Contributions summary:Adam made significant contributions to the nGraph library, focusing on implementing and optimizing core functionalities related to graph manipulation and optimization. Their work included creating utility functions for topological sorting and graph cloning, crucial for framework integration. They also implemented convolution backpropagation, addressing build warnings, and integrating auto-differentiation for numerical comparisons. These changes demonstrate a focus on performance and functionality within a deep-learning compiler and runtime.
openvinongraphtensorflowmxnetdeep-learning
plaidml/plaidml
Jul 2020 - May 2021
PlaidML is a framework for making deep learning work everywhere.
Role in this project:Back-end Developer & Performance Engineer
Back-end Developer & Performance EngineerContributions:64 reviews, 22 commits, 30 PRs in 9 months
Contributions summary:Adam's contributions focused on improving the performance and functionality of the PlaidML framework. They implemented and enabled optimization passes, including Localize, ResizeTmps, and Affinex MemRefDataFlowOpt, targeting x86 architectures. Furthermore, the user integrated Vulkan profiling support and improved kernel timing mechanisms to facilitate performance analysis. They also worked on the conversion of AffineLoad and AffineVectorLoad to PXA operations.
deep-learningplaidml
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.