Trevor Mcdonell
GPU Acceleration Engineer
Utrecht, Utrecht, Netherlands
Summary
🤩
Rockstar🎓
Top SchoolTrevor Mcdonell is a GPU Acceleration Engineer with 17 years of experience bridging functional programming and high-performance parallel computing, currently applying that expertise at PassiveLogic. He is the developer and maintainer of the Accelerate embedded language for high-performance array computations, contributing deep back-end work on AST, stencil and vector operations that underpin GPU and multicore execution. His background spans academia and industry—from an Assistant Professorship at Utrecht University and a PhD at UNSW to internships at NVIDIA—giving him a rare mix of research rigor and production-focused engineering. Trevor has also improved Haskell bindings for LLVM, solving tricky build and atomic operation issues, which reflects his attention to low-level correctness and portability. Based in Utrecht, he combines language design, compiler internals, and accelerator tuning to extract performance without sacrificing the expressiveness of functional abstractions.
17 years of coding experience
10 years of employment as a software developer
The University of SydneyUNSW Sydney
Github Skills (13)
data-structures10
code-generation10
haskell10
data-structure10
llvm10
linear-algebra9
llvm-ir9
c-language8
build-system8
testing8
architecture8
architectures8
cprogramming-language8
Programming languages (14)
Github contributions (5)
AccelerateHS/accelerate
Apr 2010 - Mar 2022
Embedded language for high-performance array computations
Role in this project:
Back-end Developer
Back-end DeveloperContributions:1 release, 21 reviews, 3113 commits in 12 years
Contributions summary:Trevor's commits indicate their focus on the core logic of the "Accelerate" library. Their contributions include additions to and modifications of the underlying AST, including changes related to Stencil and vector operations. The user also worked on refactoring internal aspects of the library, such as handling for various data types and functions.
cudaarrayembedded-languagegpuhaskell
llvm-hs/llvm-hs
Mar 2017 - Nov 2021
Haskell bindings for LLVM
Role in this project:Back-end Developer
Back-end DeveloperContributions:20 commits, 14 PRs, 12 pushes in 4 years 8 months
Contributions summary:Trevor primarily contributed to the Haskell bindings for LLVM, focusing on fixing bugs and improving the build process. They addressed typos, defaulted to shared-llvm linking, and resolved build issues for different GHC versions. The user also added support for atomic operations on floating-point types and improved testing by adding round-trip tests.
llvm-ircode-generationllvm-hshaskell-bindingshaskell
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.