I’ve really been trying to avoid posting this, but after having some major driver clashes on a new installation of Ubuntu 22.04, I have now re-installed clean with nothing to clash and using “Software & Updates” I have successfully installed the recommended “nvidia-driver-520-open” (open-kernal), which after rebooting is still in use.
The problem is that I want to continue to conda install cuda/cudnn/tenorflow-gpu, but when I run “nvidia-smi” I am getting:
“No devices were found”?
However, when running “lsmod | grep nvidia”, I’m getting:
nvidia_uvm 1363968 0
nvidia_drm 69632 0
nvidia_modeset 1155072 1 nvidia_drm
nvidia 5238784 5 nvidia_uvm,nvidia_modeset
drm_kms_helper 311296 2 nvidia_drm,i915
drm 622592 10 drm_kms_helper,nvidia,nvidia_drm,i915,ttm
Everywhere I look there seems to be an alternative method for installing these drivers, but I assumed the “Software & Updates” method would be best?
Also, although this suggested the ‘520-open’ is the recommended, after using THIS, I have downloaded “NVIDIA-Linux-x86_64-515.76”, though I am skeptical to install it as it may once again clash with the installed version.
Running “lspci -nnk | grep -iA2 vga” I am getting:
00:02.0 VGA compatible controller [0300]: Intel Corporation TigerLake-H GT1 [UHD Graphics] [8086:9a60] (rev 01)
DeviceName: Onboard - Video
Subsystem: Micro-Star International Co., Ltd. [MSI] TigerLake-H GT1 [UHD Graphics] [1462:12e9]
--
01:00.0 VGA compatible controller [0300]: NVIDIA Corporation GA106M [GeForce RTX 3060 Mobile / Max-Q] [10de:2520] (rev a1)
Subsystem: Micro-Star International Co., Ltd. [MSI] GA106M [GeForce RTX 3060 Mobile / Max-Q] [1462:12e9]
Kernel driver in use: nvidia
Should I purge current drivers and re-attempt with the new 515 version instead?