CUDA setup for Nvidia GPUs (Process can also be used depending on your configuration)

NvidiaCudasetup.txt

Before you check the download, use these links to check the compatibility:
For tensorflow and CUDA compatibility:
https://www.tensorflow.org/install/source#gpu
For CUDA and Nvidia driver compatibility:
https://docs.nvidia.com/deploy/cuda-compatibility/index.html
For cuDNN compatibility:
https://docs.nvidia.com/deeplearning/sdk/cudnn-support-matrix/index.html
https://www.tensorflow.org/install/source#linux


CUDA Download:
https://developer.nvidia.com/cuda-downloads
If you want legacy releases then click on that 
Choose your target system click on deb(local)
Download and follow the base installer insrtuctions

Nvidia Driver download:
https://www.nvidia.com/Download/index.aspx?lang=en-us
search
Download
Use the additional information for download instructions
"""
'dpkg -i nvidia-diag-driver-local-repo-ubuntu1604-384.183_1.0-1_amd64.deb’
apt-get update
apt-get install cuda-drivers
reboot
"""

run nvidia-smi //to make sure that the driver and CUDA is installed properly

cuDNN Download:
https://developer.nvidia.com/rdp/cudnn-archive
Download the runtime cuDNN file based on your OS and CUDA version
sudo dpkg -i {cuDNN.debfile}
sudo dpkg -i {cuDNN-dev.debfile}
sudo dpkg -i {cuDNN-doc.debfile}

// you should see "Test passed"
nvcc -V
or 
cp -r /usr/src/cudnn_samples_v7/ ~
cd ~/cudnn_samples_v7/mnistCUDNN
make clean && make
./mnistCUDNN

Congradulations

To completely remove CUDA, NVIDIA, cuDNN:
//remove files
sudo apt remove nvidia*
sudo apt remove cuda*
sudo apt remove libcudnn*

//removes config files from apt
dpkg -l | grep cuda- | awk '{print $2}' | xargs -n1 sudo dpkg --purge
sudo dpkg -P $(dpkg -l | grep cuda | awk '{print $2}')