Parallel copy from Host to Device ?

Accelerated Computing CUDA CUDA Programming and Performance
1

Hi,

I need to realize parallel copy HostToDevice / DeviceToHost for multiple GPU.

Recently I found that this procedure occurs in sequential.

Whether zero-copying is a way out ?

I want to know whether it is possible to do this in parallel to date?

Thanks for your reply.

2

.

3

Hi,

repeated tests have shown that access HostToDevice/DeviceToHost for multiple GPU in parallel.

Sequential access is artifact
occurrence of which I still do not understand.

Sorry that took your attention.