How can I know if NCCL is installed?

Accelerated Computing GPU-Accelerated Libraries
1

I have a very simple question. I have just recently got access to a multi node machine and I have to do some NCCL tests. In the readme it says

If CUDA is not installed in /usr/local/cuda, you may specify CUDA_HOME. Similarly, if NCCL is not installed in /usr, you may specify NCCL_HOME.

I can see that CUDA is installed but

How can I know if NCCL is installed ? and where?

I have done

find /usr -name "libnccl.so*" 2>/dev/null

and I found this file. However when I did

find /usr -name "nccl.h" 2>/dev/null

it was not found. Obviously I could not build even the simplest

#include <stdio.h>
#include <nccl.h>

int main() {
    printf("NCCL version: %d\n", NCCL_VERSION_CODE);
    return 0;
}

(Btw, I think the OS is CentOS)

2

this may be of interest

Like many software packages, nccl includes packages that provide runtime-only support (e.g. libnccl2) and packages that provide development support (e.g. libnccl-dev).

the .so file will be a part of runtime support, the .h file will be part of development support

you can use standard methods to determine if a package is installed, or not. specifics may depend on your OS and package manager.

If you have installed manually via a tar archive, you will need to keep track of that. There is no method to determine whether/where it has been installed.