
Your First CUDA Program: Add a Million Numbers on a GPU
Write a CUDA kernel in Python, learn how threads and blocks divide the work, and see whether GPU speed survives copying the data there and back.
Repair it. Mod it. Build it. Code it. Practical tutorials for people who would rather understand technology than just consume it.

Write a CUDA kernel in Python, learn how threads and blocks divide the work, and see whether GPU speed survives copying the data there and back.

Parallel processing divides work across computing resources. Learn how CPUs and GPUs use it, what limits speedups, and when more workers help.

CUDA cores are arithmetic resources, not miniature CPUs. Learn how they work with SMs, warps, Tensor Cores, and memory.