I want to install nvidia driver and cuda and cuDNN. I am using ubuntu 18.04.6 LTS. I'd like to run deep learning algorithms.
I am a little confused with the ways of installation. Could you help me with that to make sure I do the right thing.
- First way is to follow the steps provided here for installing cuda 11.2 on Ubuntu 20.04. As follows:
sudo apt-get update
sudo apt-get upgrade
# install other import packages
sudo apt-get install g++ freeglut3-dev build-essential libx11-dev libxmu-dev libxi-dev libglu1-mesa libglu1-mesa-dev
# first get the PPA repository driver
sudo add-apt-repository ppa:graphics-drivers/ppa
sudo apt update
# install nvidia driver with dependencies
sudo apt install libnvidia-common-470
sudo apt install libnvidia-gl-470
sudo apt install nvidia-driver-470
- But If I want to install cuda toolkit in this link, as far as I know, this itself can install drivers too.
My questions are:
A) I'd like to know what the difference is between these two methods? and which one is more preferable?
B) Can I run sudo ubuntu-drivers autoinstall
instead of the three last commands? If it's the latest driver is 470, are both the same?
C) In what way can I avoid reinstalling driver after every kernel update?
I'd appreciate it if someone clarify these for me.