> ./paths _ > ./modules > ./articles > ./papers

# GPU Programming

Master GPU programming from architecture basics to writing custom CUDA kernels and PyTorch extensions.

└─ status: in-progress

2 sections · 4 modules

│ Understanding GPU architecture and writing your first kernel

├─ 1.1 GPU Architecture Overview

How GPUs differ from CPUs, the SIMT execution model, and memory hierarchy fundamentals.

└─ 1.2 Writing Your First Kernel

Write, compile, and launch a CUDA kernel — from thread indexing to shared memory basics.

cuda · hands-on · setup

│ Real-world GPU programming patterns

├─ 2.1 Matrix Multiplication on GPU

Implement tiled matrix multiplication with shared memory and benchmark against cuBLAS.

cuda · optimization · math · hands-on

└─ 2.2 PyTorch Custom CUDA Extensions

Bridge Python and CUDA by writing custom PyTorch operators with the C++ extension API.

cuda · pytorch · hands-on