Learning Resources
CUDA
Programming Massively Parallel Processors (PMPP Book)
Comprehensive guide by NVIDIA Chief Scientist David Kirk and NVIDIA Fellow Wen-mei Hwu.
GPU Mode Lectures
In-depth video lectures on GPU architecture and CUDA programming by the community.
Modal GPU Glossary
Extensive glossary of GPU and CUDA terminology for quick reference.
CUDA C++ Programming Guide
NVIDIA official CUDA C++ programming guide — the definitive reference.
CUDA C++ Best Practices Guide
Official best practices for memory coalescing, bank conflicts, occupancy, and more.
NVIDIA Deep Learning Institute (DLI)
Free hands-on GPU programming labs with cloud GPU — no setup required.
GPU MODE GitHub
Community-maintained GPU learning repository with lecture notes and code examples.
CUTLASS
NVIDIA official CUDA C++ template library for high-performance linear algebra.
Triton
Triton Language Documentation
Official Triton language reference, tutorials, and API docs.
Triton GitHub Repository
Source code, examples, and issues for the Triton compiler and runtime.
Triton Getting Started Tutorials
Hands-on tutorials from OpenAI covering vector add, matmul, and fused kernels.
Introducing Triton (OpenAI Blog)
Original OpenAI blog post announcing Triton — explains the motivation and design.
TileLang
TileLang: A Composable Kernel Language for AI
Original TileLang paper published on arXiv — design philosophy and implementation.
TileLang Official Documentation
Installation guide, API reference, and examples for TileLang.
Deep Learning Compiler (MLC Book)
Comprehensive textbook on ML compiler design from the TVM/MLC community.
Apache TVM Documentation
Official documentation for the TVM compiler stack that TileLang builds upon.