Install CUDA 10.1 in Ubuntu 18.04

Pytorch를 사용해야해서 CUDA 10.1를 설치해야했는데, CUDA  공식 사이트의 다운로드 페이지에서 나와있는 설명으로 따라해도 번번히 실패를 하곤 해서 구글링을 통해 해결하였다.

도움은 깃허브 이곳[gist.github.com/Mahedi-61/2a2f1579d4271717d421065168ce6a73#file-cuda_10-1_installation_on_ubuntu_18-04-L6] 에서 받았고, 후에 사용해야해서 기록을 위해 적어놓는다.

 

### steps ###

# verity the system has a cuda-capable gpu

# download and install the nvidia cuda toolkit and cudnn

# setup environmental variables

# verify the installation

###

 

### If you have previous installation remove it first.

sudo apt-get purge nvidia*

sudo apt remove nvidia-*

sudo rm /etc/apt/sources.list.d./cuda*

sudo apt-get autoremove && sudo apt-get autoclean

sudo rm -rf /usr/local/cuda*

 

### to verify your gpu is cuda enable check

lspci | grep -i nvidia

 

### gcc compiler is required for development using the cuda toolkit. to verify the version of gcc install enter

gcc --version

 

# system update

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-key adv --fetch-keys http://developer.download.nvidia.com/compute/cuda/repos/ubuntu1804/x86_64/7fa2af80.pub

echo "deb https://developer.download.nvidia.com/compute/cuda/repos/ubuntu1804/x86_64 /" | sudo tee /etc/apt/sources.list.d/cuda.list

sudo apt-get update

 

# installing CUDA-10.1

sudo apt-get -o Dpkg::Options::="--force-overwrite" install cuda-10-1 cuda-drivers

 

# setup your paths

echo 'export PATH=/usr/local/cuda-10.1/bin:$PATH' >> ~/.bashrc

echo 'export LD_LIBRARY_PATH=/usr/local/cuda-10.1/lib64:$LD_LIBRARY_PATH' >> ~/.bashrc

source ~/.bashrc

sudo ldconfig

 

# install cuDNN v7.6

# in order to download cuDNN you have to be regeistered here https://developer.nvidia.com/developer-program/signup

# then download cuDNN v7.6 form https://developer.nvidia.com/cudnn

CUDNN_TAR_FILE="cudnn-10.1-linux-x64-v7.6.5.32.tgz"

wget https://developer.nvidia.com/compute/machine-learning/cudnn/secure/7.6.5.32/Production/10.1_20191031/cudnn-10.1-linux-x64-v7.6.5.32.tgz

tar -xzvf ${CUDNN_TAR_FILE}

 

# copy the following files into the cuda toolkit directory.

sudo cp -P cuda/include/cudnn.h /usr/local/cuda-10.1/include

sudo cp -P cuda/lib64/libcudnn* /usr/local/cuda-10.1/lib64/

sudo chmod a+r /usr/local/cuda-10.1/lib64/libcudnn*

 

# Finally, to verify the installation, check

nvidia-smi

nvcc -V