Mark Harris
Data Analyst
New South Wales, Australia
Summary
🤩
RockstarMark Harris is a seasoned data analyst and GPU computing specialist with 16 years of experience building high-performance, CUDA-accelerated data pipelines and libraries. Based in New South Wales, he has contributed deeply to the RAPIDS ecosystem (cudf, rmm, cuML, cuGraph, cuSpatial, RAFT) and maintained NVIDIA developer samples and Thrust tests, demonstrating rare expertise across both algorithm design and low-level GPU memory management. He focuses on optimizing parallel algorithms, memory allocation patterns, and robust testing—work that has influenced widely used open-source GPU data tooling. Comfortable across C++, CUDA, Python bindings, and HPC workflows, he brings a pragmatic mix of systems-level rigor and data-focused engineering. An interesting strength: he routinely turns build and synchronization pitfalls into maintainable solutions, preventing subtle performance regressions at scale.
16 years of coding experience
Spanish, English
Stackoverflow
27,033reputation
3.4mreached
235answers
2questions
gpu
top-1%
gpgpu
top-1%
nvidia
top-1%
cuda
top-1%
Github Skills (44)
algorithm10
unit-testing10
thrust10
algorithms10
graph-algorithms10
c-language10
gpgpu10
testing10
memory-management10
gpu-programming10
machine-learning10
cmake10
data-structure10
kernel10
nvidia10
Programming languages (11)
Github contributions (5)
NVIDIA-developer-blog/code-samples
Mar 2012 - Nov 2021
Source code examples from the Parallel Forall Blog
Role in this project:
Backend Developer
Backend DeveloperContributions:7 reviews, 68 commits, 17 PRs in 9 years 10 months
Contributions summary:Mark implemented and refined CUDA code examples for the NVIDIA developer blog. Their primary focus was on developing and optimizing parallel algorithms, specifically for numerical methods like Jacobi relaxation. The contributions involved integrating OpenACC and OpenMP directives to improve performance and readability. They also added example code related to CUDA C++ and Fortran.
gpucudaparallel
rapidsai/rmm
Aug 2018 - Oct 2022
RAPIDS Memory Manager
Role in this project:Back-end Developer
Back-end DeveloperContributions:1 release, 991 reviews, 774 commits in 4 years 3 months
Contributions summary:Mark appears to be focused on the implementation and maintenance of the RAPIDS Memory Manager within the `rmm` repository. Their commits primarily involve adding memory manager-related files, test files, and adapting existing tests to use the new RMM system. The user also contributed to the python bindings and added functionality for the numba cuda.device_array.
cudamemory-managementmemorycpppython
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.