Optimize For Gpu

This is a comprehensive GPU acceleration skill that transforms CPU-bound Python code to run on NVIDIA GPUs, often delivering 10x to 1000x speedups. It covers the full RAPIDS ecosystem (cuDF, cuML, cuGraph) plus CuPy, Numba CUDA, and Warp for physics simulation and differentiable rendering. The skill intelligently picks the right tool based on your workload: CuPy for NumPy array operations, cuDF for pandas dataframes, cuML for scikit-learn pipelines, and Warp for particle systems and mesh operations. It handles everything from vector search and geospatial analysis to medical imaging and GPUDirect Storage file I/O. The decision framework is solid, and it triggers proactively when it spots CPU-bound code that would benefit from GPU acceleration, even if you didn't explicitly ask for it.

skills.sh installs: 2.1K
GitHub stars: 56.6K
last updated: Aug 21, 2026

View on GitHub

Install

$ npx -y skills add k-dense-ai/scientific-agent-skills --skill optimize-for-gpu --agent claude-code

Installs into .claude/skills of the current project.

Files

GPU Optimization for Python with NVIDIA

Treat GPU acceleration as an evidence-driven optimization, not an automatic rewrite. Preserve the user's numerical and algorithmic contract, measure with representative data, and keep the GPU version only when synchronized end-to-end benchmarks show a useful improvement.

Reviewed against RAPIDS 26.08, CuPy 14.2, Numba-CUDA 0.30.4, and Warp 1.17. The references contain illustrative GPU examples: source/API review is not execution on CUDA hardware. Validate them on the user's target GPU before reporting correctness or speedup. Do not use a moving latest documentation page to infer compatibility with a pinned release.