Nvidia Drivers and CUDA
Installing CUDA 10.0 on Ubuntu 16.04
General Info
check nvidia drivers and cuda version:
nvcc --version nvidia-smi
Download the latest version of CUDA which is compatible with your version of drivers installed. For example, if you have installed a driver of 410 version, CUDA 10 can be installed, for 375 CUDA 8 should be installed.
CUDA 10.0 (10.0.130) >= 410.48 driver
Avoid CUDA 10.1, it conflicts with tensorflow-gpu
Steps:
1. Uninstall previous versions of CUDA
sudo apt-get --purge remove cuda-10.1 sudo apt-get remove nvidia-cuda-toolkit
2. Then install:
sudo dpkg -i cuda-repo-ubuntu1604-10-0-local-10.0.130-410.48_1.0-1_amd64.deb sudo apt-key add /var/cuda-repo-10-0-local-10.0.130-410.48/7fa2af80.pub sudo apt-get update sudo apt-get install cuda-10-0
Steps and installation file downloaded from here:
https://developer.nvidia.com/cuda-10.0-download-archive?target_os=Linux&target_arch=x86_64&target_distro=Ubuntu&target_version=1604&target_type=deblocal
3. Specify your cuda path (cuda 10.0 example)
sudo nano /home/interactionstation/.bashrc
locate the two lines regarding cuda9 (in case it was previously installed), comment them out and include instead:
export PATH=/usr/local/cuda-10.0/bin:$PATH export LD_LIBRARY_PATH=/usr/local/cuda-10.0/lib64:$LD_LIBRARY_PATH
4. Save the file
5. Restart the computer
More info: https://medium.com/@muskulpesent/install-cuda-on-ubuntu-16-04-67ec087f8561 https://developer.nvidia.com/cuda-downloads?target_os=Linux&target_arch=x86_64&target_distro=Ubuntu&target_version=1604&target_type=deblocal
Installing CUDA 9.0 on Ubuntu 16.04 (Old! Using 10.0, Nov 2019)
- wget http://developer.download.nvidia.com/compute/cuda/repos/ubuntu1604/x86_64/cuda-repo-ubuntu1604_9.0.176-1_amd64.deb
- wget http://developer.download.nvidia.com/compute/machine-learning/repos/ubuntu1604/x86_64/libcudnn7_7.0.5.15-1+cuda9.0_amd64.deb
- wget http://developer.download.nvidia.com/compute/machine-learning/repos/ubuntu1604/x86_64/libcudnn7-dev_7.0.5.15-1+cuda9.0_amd64.deb
- wget http://developer.download.nvidia.com/compute/machine-learning/repos/ubuntu1604/x86_64/libnccl2_2.1.4-1+cuda9.0_amd64.deb
- wget http://developer.download.nvidia.com/compute/machine-learning/repos/ubuntu1604/x86_64/libnccl-dev_2.1.4-1+cuda9.0_amd64.deb
- sudo dpkg -i cuda-repo-ubuntu1604_9.0.176-1_amd64.deb
- sudo dpkg -i libcudnn7_7.0.5.15-1+cuda9.0_amd64.deb
- sudo dpkg -i libcudnn7-dev_7.0.5.15-1+cuda9.0_amd64.deb
- sudo dpkg -i libnccl2_2.1.4-1+cuda9.0_amd64.deb
- sudo dpkg -i libnccl-dev_2.1.4-1+cuda9.0_amd64.deb
- sudo apt-get update
- sudo apt-get install cuda=9.0.176-1
- sudo apt-get install libcudnn7-dev
- sudo apt-get install libnccl-dev
- sudo reboot now
- export PATH=/usr/local/cuda-9.0/bin${PATH:+:${PATH}}
- export LD_LIBRARY_PATH=/usr/local/cuda-9.0/lib64${LD_LIBRARY_PATH:+:${LD_LIBRARY_PATH}}
- sudo nano .bashrc
- Add the two last export lines at the end of the file. Save and reboot.
Resources used:
- https://askubuntu.com/questions/61396/how-do-i-install-the-nvidia-drivers
- https://medium.com/@bbloks/a-machine-learning-environment-with-ubuntu-and-gpu-acceleration-in-5-steps-765608325356
- https://yangcha.github.io/CUDA90/
Installing cuDNN (for ubuntu 16.04, cuda 10-0)
1.Download the following:
- runtime:
- developer
- samples
2. and then, navigate to your <cudnnpath> directory containing cuDNN Debian file.
3. Install the runtime library, for example:
sudo dpkg -i runtimefile.deb
4. Install the developer library, for example:
sudo dpkg -i devlib.deb
5. Install the code samples and the cuDNN Library User Guide, for example:
sudo dpkg -i codesamples.deb