Curated GitHub list of CUDA programming books from beginner to advanced, covering C++, Python, architecture, and 2024-2026 releases.
Key Takeaways
Canonical starting points: CUDA by Example (Sanders & Kandrot, 2010) for beginners; Programming Massively Parallel Processors 3rd Ed. (Kirk & Hwu, 2022) as the architecture reference.
Python track covered via Numba/CuPy (Hands-On GPU Programming with Python and CUDA, Tuomanen 2018) and modern pybind11 interop (Motta 2024).
2024-2026 shelf includes dedicated titles on kernel optimization, Nsight debugging, Tensor Cores, multi-GPU, and CUDA 12.6/13 – useful for staying current post-Hopper.
Maintainer recommends pairing any book with the free official CUDA C++ Programming Guide (v13.x, 2026) given how fast the API evolves.
Library depth includes cuBLAS, cuFFT, Thrust, NPP coverage (Soyata 2018) and production patterns like streams and multi-GPU (Cheng et al. 2014).
