On Dienstag 09 Juni 2009, Chris Heuser wrote:
> Hey everybody,
>
> I know that when a call is made to a CUDA kernel, control immediately
> returns to the calling function or program. But if I then immediately call
> a memcpy from the changed gpu array to an array on the host, will the call
> wait for the kernel to finish before executing the memcpy?

It will. Memory transfers don't overlap with kernel execution *within the same 
stream*.

Andreas

Attachment: signature.asc
Description: This is a digitally signed message part.

_______________________________________________
PyCuda mailing list
[email protected]
http://tiker.net/mailman/listinfo/pycuda_tiker.net

Reply via email to