Hello,
we have a new hp apollo 6500 working station with A100 GPUs using Ubunto 20.04 LTS and have installed the driver 465 for cuda 11.1 , e. g.
using the repos Index of /compute/cuda/repos/ubuntu2004/x86_64
and e. g. apt install -y cuda-runtime-11-1 / apt install -y cuda-11-1
The output of dpkg -l | grep nvidia confirms that these packages are installed.
For our kernel 5.4.0-73.82-generic 5.4.106 a nvidia.ko module can also be found under the directory /var/lib/dkms/nvidia/465.19.01/5.4.0-73-generic/x86_64/module.
A ‘modprobe nvidia’ shows, that nvidia could not inserted: No such device and also nvidia-smi failed, because it couldn’t communicate with the NVIDIA driver. It is not loaded.
The usual precaution steps like blacklist the nouveau module (/lib/modprobe.d/nvidia-graphics-drivers.conf) and not enabling secure boot are done.
We are not sure if the use of ubuntu-drivers command was counterproductive. The output of ubuntu-drivers devices is:
== /sys/devices/pci0000:11/0000:11:00.0/0000:12:00.0/0000:13:00.0/0000:14:00.0 ==
modalias : pci:v000010DEd000020F1sv000010DEsd0000145Fbc03sc02i00
vendor : NVIDIA Corporation
manual_install: True
driver : nvidia-driver-450-server - distro non-free
driver : nvidia-driver-455 - third-party non-free
driver : nvidia-driver-460 - third-party non-free recommended
driver : nvidia-driver-460-server - distro non-free
driver : xserver-xorg-video-nouveau - distro free builtin.
nvidia-bug-report.log.gz (5.3 MB)
Can someone help us further?