C++ AMP function argument

Question

Hello,  I am trying to call functions inside a for loop in main.  My AMP code works fine, but I am not really sure what is going on.  In the following code, am I copying data from host to GPU twice every step of the loop? Both array A and
B are really large. I'm trying to avoid copying as much as I can.  It would be very nice if I can pass arrays on GPU (created by funcX) as funcY arguments.  Is there any way I can pass the array already exist on GPU to funcY?  The arrays I use
are defined as 1-Dimensional on CPU and it is treated as 2-Dimensional on GPU. Thank you in advance for your help.   
void funcX(int* A, int* B ){
array_view<int,2> Aamp(n,m,A);
array_view<int,2> Bamp(n,m,B);
parallel_for_each(…..
//do something here
void funcY(int* A, int* B ){
array_view<int,2> Aamp(n,m,A);
array_view<int,2> Bamp(n,m,B);
parallel_for_each(…..
//do something here
int main(){
For(i=0;i<10;i++){
funcX(A,B);
funcY(A,B);
return 0;
 

Answer 1

Hello pi123 (should it not be 314?:)). For this case, the easiest way to go about things is to use the source-less array_view i.e. do something like this:
array_view<T> tmp(some_size);
parallel_for_each(..., [=](...) restrict(amp) {
// Use tmp here.
That will ensure that the underlying storage lives on the accelerator side, and doesn't get spuriously copied back (unless you explicitly access tmp on the host side). I hope this helps. Cheers!