cuda.core.utils.discard_prefetch#

cuda.core.utils.discard_prefetch(targets, location=None, *, options=None, stream)#

Discard one or more managed-memory ranges and prefetch them to a target location.

Parameters:

  • targets (Buffer | Sequence[Buffer]) – One or more managed allocations to discard and re-prefetch.

  • location (Location | Device | int | Sequence[…]) – Target location(s). A single location applies to all targets; a sequence must match len(targets).

  • options (None) – Reserved for future per-call flags. Must be None.

  • stream (Stream | GraphBuilder) – Stream for the asynchronous operation (keyword-only).

Raises:

NotImplementedError – On a CUDA 12 build of cuda.core. Discard-and-prefetch requires CUDA 13+.

On this page