Selected Experience
NVIDIA | Senior CUDA Math Libraries Engineer | 2023–Present
- Optimize cuFFT performance for Blackwell, Rubin, Feynman, and future GPU architectures.
- Developed an AI-agent framework for CUDA kernel generation, profiling, and autotuning.
- Designed AI-agent harnesses and tools (MCPs, CLIs, ...) to accelerate development.
- Developed CUDA Tile integration and autotuning across SIMT, cuTile Python, NVCC, and LTO IR.
- Optimized ML inference in PyTorch using internal research techniques.
Innovative Computing Laboratory @ UTK | Research Assistant (HPC) | 2019–2023
- Developed MAGMA and SLATE GPU kernels; optimized MAGMA BLAS for AMD GPUs.
- Designed SABATH for reproducible scientific ML following FAIR principles.
Qardian Labs | Machine Learning Consultant | 2020
- Trained and evaluated an ML model for heart-disease identification and anomaly detection, with a web interface for risk assessment.
Oak Ridge National Laboratory | Research Intern (HPC) | 2017–2018
- Built FractalExplorer, a distributed GPU fractal visualizer for the SimpleSummit Jetson cluster.
Selected Publications
Brown et al. Design, Optimization, and Benchmarking of Dense Linear Algebra Algorithms on AMD GPUs. IEEE HPEC, 2020.
Luszczek & Brown. Surrogate ML/AI Model Benchmarking for FAIR Principles’ Conformance. IEEE HPEC, 2022.
Selected Projects
mycc: C-subset compiler using Bison, Flex, and LLVM, with an optional ORC JIT path.
kscript: Dynamic language implemented in C, with arrays, linear algebra, FFTs, and a WASM REPL.
Skills
Languages: C/C++, CUDA, Python, Rust, TypeScript.
Machine Learning: PyTorch, JAX, Triton, cuTile Python, LLVM/MLIR, JIT/LTO.
Agentic Engineering: Claude, Codex, Cursor, OpenCode, Pi, MCPs, harnesses, CLIs.