Menu
Reinstall NVIDIA drivers without opengl Ubuntu 16.04 GTX 1080ti
(Last time, NVIDIA driver 390 was installed in the old tutorial) There are only few steps to install the GPU driver for GeForce GTX 1080Ti in Ubuntu 18.04.(Sik-Ho Tsang @ Medium)First, remove any. Remove installed nvidia driver & reset to default graphics driver. On Ubuntu 15.10, after installing the nvidia proprietary driver, it didn't quite work as expected.going down to older version made it work, but graphics were slow. Run the below to reset back to the original graphics setup. Sign up for free to join this conversation on GitHub.
reinstall.sh
# Download installers |
mkdir ~/Downloads/nvidia |
cd~/Downloads/nvidia |
wget https://developer.nvidia.com/compute/cuda/8.0/Prod2/local_installers/cuda_8.0.61_375.26_linux-run |
wget http://us.download.nvidia.com/XFree86/Linux-x86_64/384.59/NVIDIA-Linux-x86_64-384.59.run |
sudo chmod +x NVIDIA-Linux-x86_64-384.59.run |
sudo chmod +x cuda_8.0.61_375.26_linux-run |
./cuda_8.0.61_375.26_linux-run -extract=~/Downloads/nvidia/ |
# Uninstall old stuff |
sudo apt-get --purge remove nvidia-* |
sudo nvidia-uninstall |
# Reboot |
sudo shutdown -r now |
# In grub boot menu hit `e` and add `nouveau.modeset=0` to the end of the line beginning with `linux` |
# F10 to boot |
# CTRL+ALT+F1 and log in |
sudo service lightdm stop |
sudo ./NVIDIA-Linux-x86_64-384.59.run --no-opengl-files |
sudo ./cuda-linux64-rel-8.0.61-21551265.run --no-opengl-libs |
# Verify installation |
nvidia-smi |
cat /proc/driver/nvidia/version |
# Start jupyter |
xvfb-run -s '-screen 0 1400x900x24' jupyter notebook |
commented Oct 26, 2017
This is AMAZING. Thank you so much! |
commented Jan 1, 2018
Thanks a lot, worked like a charm on a - google cloud Tesla K80 gpu ubuntu 16.04 - wm. But I don't need the xvfb-run -s '-screen 0 1400x900x24' part, strange?By the way, did you manage to get universe working? |
commented Feb 1, 2018
Yes! It works and it's great :D But even I don't need to use xvfb-run for rendering gym, I used a -X flag for ssh Any idea what might be happening? |
commented Feb 1, 2018
Also, if you could add another step about installing cuDNN libraries too, it would be useful for other people https://stackoverflow.com/questions/31326015/how-to-verify-cudnn-installation This helps! |
Sign up for freeto join this conversation on GitHub. Already have an account? Sign in to comment