CUDA returns 0 for all values for simple vector addition

I have an RTX 3070 and installed cuda 12.2 Ubuntu 22.04. I got the code from
The repository supposedly works. But I get zeros for the added vector C.
Can anyone please help.
Thanks in advance.
I am adding my code as well in case if you want to look (3.1 KB)
CMakeLists.txt (288 Bytes)

Your code is not doing proper CUDA error checking. Coupled with that, you are compiling for an incorrect architecture; does not match your GPU. As a result, you are getting a launch error but you don’t know it.

Thanks for the reply, I installed the correct cuda version, i.e., nvcc --version and nvidia-smi shows the same cuda version.

None of that has anything to do with what I suggested.

This is the changed CMakelists.txt, I am using now. And this is giving the correct answer.