Lessons
GPU basics and how inference actually runs
intermediate · 18 min
A grid of SMs. Each one is a tiny warp-wide processor.