0PricingLogin
CUDA Academy · Lesson

Pinned Memory with cudaMallocHost

Page-locked buffers for fast DMA.

Meet Pinned Memory

Pinned memory is host memory the operating system promises never to move or swap out. Because its physical address is fixed, the GPU can read it directly.

Also Called Page-Locked

You will see the term page-locked used as a synonym for pinned. Both mean the same thing: pages locked in place so no swapping can ever happen.

All lessons in this course

  1. Why Pageable Memory Is Slow
  2. Pinned Memory with cudaMallocHost
  3. cudaMemcpyAsync in a Stream
  4. The Double-Buffering Pipeline
← Back to CUDA Academy