Hi
On a Dell R740 with a dual GV100GL Tesla V100, and Linux 4.18.0-193.6.3.el8_2.x86_64
with Centos 8.2
Following https://docs.nvidia.com/cuda/cuda-installation-guide-linux/ and up to step “3.3. RHEL8/CentOS8”.
$ sudo dnf install cuda
NOTE: Skipping kernel installation since no NVIDIA driver kernel module package kmod-nvidia-450.51.06-4.18.0-193.6.3 for kernel kernel-4.18.0-193.6.3.el8_2.x86_64 and driver nvidia-driver-3:450.51.06-1.el8.x86_64 could be found
Package cuda-11.0.2-1.x86_64 is already installed.
$ sudo dnf install nvidia-driver:latest-dkms
NOTE: Skipping kernel installation since no NVIDIA driver kernel module package kmod-nvidia-450.51.06-4.18.0-193.6.3 for kernel kernel-4.18.0-193.6.3.el8_2.x86_64 and driver nvidia-driver-3:450.51.06-1.el8.x86_64 could be found
No match for argument: nvidia-driver:latest-dkms
Error: Unable to find a match: nvidia-driver:latest-dkms
$ sudo dnf install nvidia-driver
NOTE: Skipping kernel installation since no NVIDIA driver kernel module package kmod-nvidia-450.51.06-4.18.0-193.6.3 for kernel kernel-4.18.0-193.6.3.el8_2.x86_64 and driver nvidia-driver-3:450.51.06-1.el8.x86_64 could be found
Package nvidia-driver-3:450.51.06-1.el8.x86_64 is already installed.
Dependencies resolved.
Nothing to do.
“dnf list installed | grep nvidia-driver” shows it is installed:
nvidia-driver.x86_64 3:450.51.06-1.el8 @cuda-rhel8-x86_64
nvidia-driver-NVML.x86_64 3:450.51.06-1.el8 @cuda-rhel8-x86_64
nvidia-driver-NvFBCOpenGL.x86_64 3:450.51.06-1.el8 @cuda-rhel8-x86_64
nvidia-driver-cuda.x86_64 3:450.51.06-1.el8 @cuda-rhel8-x86_64
nvidia-driver-cuda-libs.x86_64 3:450.51.06-1.el8 @cuda-rhel8-x86_64
nvidia-driver-devel.x86_64 3:450.51.06-1.el8 @cuda-rhel8-x86_64
nvidia-driver-libs.x86_64 3:450.51.06-1.el8 @cuda-rhel8-x86_64
This was a clean install and have tried the above a few times. This looks like the same error that this person reported here:
Missing Nvidia Drivers while Installing CUDA-11 on RHEL8.1
https://forums.developer.nvidia.com/t/missing-nvidia-drivers-while-installing-cuda-11-on-rhel8-1/139664
The device files are there:
$ ls -l /dev/nvidia*
crw-rw-rw-. 1 root root 195, 0 Jul 27 11:45 /dev/nvidia0
crw-rw-rw-. 1 root root 195, 1 Jul 27 11:45 /dev/nvidia1
crw-rw-rw-. 1 root root 195, 255 Jul 27 11:45 /dev/nvidiactl
crw-rw-rw-. 1 root root 195, 254 Jul 27 11:45 /dev/nvidia-modeset
crw-rw-rw-. 1 root root 234, 0 Jul 27 11:45 /dev/nvidia-uvm
crw-rw-rw-. 1 root root 234, 1 Jul 27 11:45 /dev/nvidia-uvm-tools
Thanks
Mike