If we want to copy $3000$ bytes of data from host array A$\_$h $($A$\_$h is a pointer

to element $0$ of the source array$)$ to device array A$\_$d $($A$\_$d is a pointer to

element $0$ of the destination array$),$ what would be an appropriate API call

for this data copy in CUDA?

$($A$)$ cudaMemcpy$(3000,$ A$\_$h$,$ A$\_$d$,$ cudaMemcpyHostToDevice$)$;

$($B$)$ cudaMemcpy$($A$\_$h$,$ A$\_$d$,3000,$ cudaMemcpyDeviceTHost$)$;

$($C$)$ cudaMemcpy$($A$\_$d$,$ A$\_$h$,3000,$ cudaMemcpyHostToDevice$)$;

$($D$)$ cudaMemcpy$(3000,$ A$\_$d$,$ A$\_$h$,$ cudaMemcpyHostToDevice$)$;