Lesley's Digital Garden

Home

❯

Notes

❯

CUDA memory and cache architecture

CUDA memory and cache architecture

Apr 08, 2024

  • parent : CUDA
  • computing/cuda/memory

Different types of memory

  • R/W per-thread register
  • R/W per-thread local memory
  • R/W per-block shared memory and L1 cache
  • R/W per-grid global memory
  • Read-only per-grid constant memory (read-only on device)

Host code can write to constant memory


Graph View

Backlinks

  • CUDA

Created with Quartz v4.5.2 © 2026. All notes are written in Obsidian.

  • GitHub Repository
  • Mastodon