Skip to content

Instantly share code, notes, and snippets.

@Mahedi-61
Created December 6, 2017 11:15
Show Gist options
  • Star 4 You must be signed in to star a gist
  • Fork 2 You must be signed in to fork a gist
  • Save Mahedi-61/7b1f737e328fe19f4063f08a009c82c4 to your computer and use it in GitHub Desktop.
Save Mahedi-61/7b1f737e328fe19f4063f08a009c82c4 to your computer and use it in GitHub Desktop.
Installing CUDA toolkit v8..0 and cuDNN 6.0 on Ubuntu 16.04
#!/bin/bash
## This gist is a step by step instructions to install cuda v8.0 and cudnn 6.0 on ubuntu 16.04
## official guide: http://docs.nvidia.com/cuda/cuda-installation-guide-linux/index.html
### steps ####
# verify the system has a cuda-capable gpu
# verify the system has gcc installed
# download and install the nvidia cuda toolkit
# download cudnn
# setup environment variables
# verify the system
###
### checking your NVIDIA gpu is cuda enable or not got to https://developer.nvidia.com/cuda-gpus
### to verify your gpu is cuda enable enter in to the comman line
lspci | grep -i nvidia
### gcc compiler is required for development using the cuda toolkit. to verify the version of gcc install enter
gcc --version
# to download go to https://developer.nvidia.com/cuda-downloads
# set (linux -> x86_64 -> Ubuntu -> 16.04 -> deb (network))
CUDA_REPO_PKG="cuda-repo-ubuntu1604_8.0.61-1_amd64.deb"
wget http://developer.download.nvidia.com/compute/cuda/repos/ubuntu1604/x86_64/${CUDA_REPO_PKG}
sudo dpkg -i ${CUDA_REPO_PKG}
sudo apt-get update
# to install cuda toolkit set cuda-8-0 (..not cuda or cuda-8.0)
sudo apt-get install cuda-8-0
# install cuDNN v6.0
# in order to download cuDNN you have to regeistered here https://developer.nvidia.com/developer-program/signup
# then download cuDNN v6.0 form https://developer.nvidia.com/cudnn
CUDNN_TAR_FILE="cudnn-8.0-linux-x64-v6.0.tgz"
wget http://developer.download.nvidia.com/compute/redist/cudnn/v6.0/${CUDNN_TAR_FILE}
tar -xzvf ${CUDNN_TAR_FILE}
#copy the following files into the cuda toolkit directory.
sudo cp -P cuda/include/cudnn.h /usr/local/cuda-8.0/include
sudo cp -P cuda/lib64/libcudnn* /usr/local/cuda-8.0/lib64/
sudo chmod a+r /usr/local/cuda-8.0/lib64/libcudnn*
# set environment variables permanently on ~/.bashrc file
export PATH=/usr/local/cuda-8.0/bin${PATH:+:${PATH}}
export LD_LIBRARY_PATH=/usr/local/cuda-8.0/lib64${LD_LIBRARY_PATH:+:${LD_LIBRARY_PATH}}
# need to logout once to get the effect of these variables
### verification of instllation.
# to verify nvidia driver version
cat /proc/driver/nvidia/version
# to verify current cuda toolkit version
nvcc -V
# to check which processes are running on gpu
nvidia-smi
# if above command shows output. then we are done !!
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment