Black screen boot after installing CUDA 10.0 on versions 18.04.3 / 18.04 / 16.04.6

I am trying to install CUDA 10.0 because tensorflow 1.13/1.14 only supports this version.

I have just installed a clean version of Ubuntu 18.04.3 (same issue occurs with 18.04 and 16.04.6) following these instructions, downloaded form here. I chose the minimal install option so Ubuntu doesn’t install any drivers.

The first time I log in, I open a terminal and run the following

$ sudo apt update
$ sudo apt upgrade
$ sudo apt-get install linux-headers-$(uname -r)
$ sudo apt install build-essential
$ sudo apt-get install manpages-dev

At this stage, there are no nvidia drivers and no cuda software installed (to make sure, I also purge nvidia and cuda).

I download CUDA 10.0 from here and proceed to run it via (I am following the nvidia instructions here)

$ sudo dpkg -i cuda-repo-ubuntu.x.deb
$ sudo apt-key add /var/cuda-repo-10.0-local/7fa2af80.pub
$ sudo apt update
$ sudo apt install cuda
$ export PATH=/usr/local/cuda-10.0/bin${PATH:+:${PATH}}
$ export LD_LIBRARY_PATH=/usr/local/cuda-10.0/lib64${LD_LIBRARY_PATH:+:${LD_LIBRARY_PATH}}
$ sudo reboot

When I reboot however, I get a black screen with the top line giving some information about my nvme internal drive partition

/dev/nvme0n1p2: clean, 185049/30498816 files, 5606916/121965056 blocks

I can’t even use CTRL+ALT+F2 here because it seems like the screen keeps jumping and doesn’t allow me to type.

This only happens after installing cuda 10.0 specifically. If I try a separate run where I only install a nvidia driver without having installed cuda 10.0, I can reboot my computer multiple times and it works fine (this issue was suggested here). This is strange because the nvidia driver automatically installs cuda 10.1 (this issue is also mentioned at the end of this post) as displayed by the nvidia-smi command.

How can I fix the black boot screen issue and still have cuda 10.0 installed?

The GPU is a GeFroce RTX 2070 Super (list of supported nvidia drivers here)