A comprehensive collection of GPU kernel examples demonstrating essential parallel computing techniques for modern GPU programming. This project supports both NVIDIA CUDA and AMD ROCm platforms, focusing on the most in-demand GPU programming skills required in industry today.

This repository contains examples of how to accelerate common Python data science libraries using NVIDIA GPUs. Each notebook demonstrates a different library and shows how to enable GPU (CUDA) acceleration with minimal code changes.

CUDA/PTX Static Analysis & Guidance Engine - GPU-free static analyzer for NVIDIA PTX assembly that surfaces warp divergence, register spills, and occupancy bottlenecks before you ever touch a profiler.