deep-roots learning program · module 12
Module 12 — GPU & AI accelerator memory
Every chapter: first principles, What insiders know, a See it yourself lab, connections, and a quiz with worked answers.
02
Coalescing & the roofline
Warp memory access as cache-line physics; tiling; compute-vs-memory bound.
→
03
Allocation & the real OOM
GMMU, the caching allocator, fragmentation — why free ≥ request still fails.
→
the exam