arjun-kava / install-cuda-cudnn

installation process for CUDA and CuDNN

Installation process for CUDA and CuDNN

Verify Capacity of GPU

lspci | grep -i nvidia

Check supported version of linux

uname -m && cat /etc/*release

verify gcc and make installed in system

gcc --version # if failed execute sudo apt install gcc

make --version # if failes execute sudo apt install make

check version of kernel and install headers

uname -r

sudo apt-get install linux-headers-$(uname -r)

blacklist nouveau drivers

echo "blacklist nouveau" >> /etc/modprobe.d/blacklist-nouveau.conf

echo "options nouveau modeset=0" >> /etc/modprobe.d/blacklist-nouveau.conf

Regenerate kernel initrampfs

sudo update-initramfs -u

Remove preinstalled drivers

sudo apt-get purge nvidia-cuda*

sudo apt-get purge nvidia-*

Download Compatible driver .run file

https://developer.nvidia.com/cuda-downloads

Log in with init 3 and install drivers

init 3

cd Downloads

./your-nividia-driver.run # skip OpenGL installation and create cuda link while installation

After Installation Config Path

nano /home/$USER/.bashrc

append following content into file

export PATH=/usr/local/cuda-x.x/bin${PATH:+:${PATH}}

export LD_LIBRARY_PATH=/usr/local/cuda-x.x/lib64${LD_LIBRARY_PATH:+:${LD_LIBRARY_PATH}}

export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/cuda-x.x/lib64

then save file and execute

source /home/$USER/.bashrc

Verify Installation

nvcc --version