Or do I need to perform cudaMemset() if I want to make sure the array contains all 0? I can't find it in the documentation. Thank you.
The cudaMalloc
documentation says:
Allocates
size
bytes of linear memory on the device and returns in*devPtr
a pointer to the allocated memory. The allocated memory is suitably aligned for any kind of variable. The memory is not cleared.
So, you will need cudaMemset
to initialize the memory.