none of the providers can be installed

On following the instructions for installing CUDA to CentOS 8, I get the following errors

Error: 
 Problem: package cuda-10.1.243-1.x86_64 requires cuda-10-1 >= 10.1.243, but none of the providers can be installed
  - package cuda-10-1-10.1.243-1.x86_64 requires cuda-runtime-10-1 >= 10.1.243, but none of the providers can be installed
  - package cuda-runtime-10-1-10.1.243-1.x86_64 requires nvidia-drivers >= 418.87, but none of the providers can be installed
  - package nvidia-driver-3:418.87.00-1.el8.x86_64 requires nvidia-kmod-common = 3:418.87.00, but none of the providers can be installed
  - package cuda-drivers-418.87.00-1.x86_64 requires nvidia-kmod >= 3:418.87.00, but none of the providers can be installed
  - package nvidia-kmod-common-3:418.87.00-1.el8.noarch requires nvidia-kmod = 3:418.87.00, but none of the providers can be installed
  - package cuda-drivers-418.87.01-1.x86_64 requires nvidia-xconfig >= 3:418.87.01, but none of the providers can be installed
  - cannot install the best candidate for the job
  - nothing provides dkms needed by kmod-nvidia-latest-dkms-3:418.87.00-1.el8.x86_64
  - package kmod-nvidia-latest-dkms-3:418.87.01-1.el8.x86_64 is excluded
  - nothing provides dkms needed by kmod-nvidia-latest-dkms-3:418.87.01-1.el8.x86_64
  - package nvidia-driver-3:418.87.01-1.el8.x86_64 is excluded
  - package nvidia-xconfig-3:418.87.01-1.el8.x86_64 is excluded

Bump.

I figured it out. The problem is that a package dkms is not installed. I downloaded it from here, https://centos.pkgs.org/7/epel-x86_64/dkms-2.7.1-1.el7.noarch.rpm.html.

It requires elfutils-libelf-devel.x86_64.rpm be installed. Once that’s done, the instructions for installing cuda on the nvidia webpage will work. I suggest the documentation on the page be fixed.