RHEL 7.5

1. Create a repository on the MN node installing the CUDA Toolkit:

   # For cuda toolkit name: /path/to/cuda-repo-rhel7-9-2-local-9.2.64-1.ppc64le.rpm
   # extract the contents from the rpm
   mkdir -p /tmp/cuda
   cd /tmp/cuda
   rpm2cpio /path/to/cuda-repo-rhel7-9-2-local-9.2.64-1.ppc64le.rpm | cpio -i -d
   
   # Create the repo directory under xCAT /install dir for cuda 9.2
   mkdir -p /install/cuda-9.2/ppc64le/cuda-core
   cp /tmp/cuda/var/cuda-repo-9-2-local/*.rpm /install/cuda-9.2/ppc64le/cuda-core
   
   # Create the yum repo files
   createrepo /install/cuda-9.2/ppc64le/cuda-core
   

2. The NVIDIA CUDA Toolkit contains rpms that have dependencies on other external packages (such as DKMS). These are provided by EPEL. It’s up to the system administrator to obtain the dependency packages and add those to the cuda-deps directory:

   mkdir -p /install/cuda-9.2/ppc64le/cuda-deps
   
   # Copy the DKMS rpm to this directory
   cp /path/to/dkms-2.4.0-1.20170926git959bd74.el7.noarch.rpm /install/cuda-9.2/ppc64le/cuda-deps
   
   # Execute createrepo in this directory
   createrepo /install/cuda-9.2/ppc64le/cuda-deps