Memory hierarchy

Appears in 1 tutorial

Tiers from tiny-fast (registers, SRAM/L1) to big-slow (VRAM), then off-GPU (system RAM, disk).

As used in LLM Infrastructure →

Tiers from tiny-fast (registers, SRAM/L1) to big-slow (VRAM), then off-GPU (system RAM, disk). Good kernels keep hot data in fast tiers.