# Install CUDA
https://koding.work/how-to-install-cuda-and-cudnn-to-ubuntu-20-04/

## 若已裝錯版本，請依照以下步驟執行
1.重新安裝driver

$ sudo apt remove '^nvidia'
$ sudo apt autoremove
$ sudo reboot
$ sudo apt install nvidia-driver-xxx

2. 修改~/.bashrc

$ vim ~/.bashrc
找到export PATH=/usr/local/cuda-11.4/bin${PATH:+:${PATH}}並刪除 (dd 刪除整行)
找到export LD_LIBRARY_PATH=/usr/local/cuda11.4/lib64${LD_LIBRARY_PATH:+:${LD_LIBRARY_PATH}}並刪除 (dd 刪除整行)
:wq
$ sudo reboot

## Driver Error (reinstall driver)
https://clay-atlas.com/us/blog/2022/07/29/solved-nvidia-smi-has-failed-because-it-couldnt-communicate-with-the-nvidia-driver-make-sure-that-the-latest-nvidia-driver-is-installed-and-running/

NVIDIA-SMI has failed because it couldn't communicate with the NVIDIA driver. Make sure that the latest NVIDIA driver is installed and running.

sudo apt purge nvidia-*
sudo apt search nvidia-driver*
sudo apt install nvidia-driver<NVIDIA DRIVER VERSION>

## CUDA 11.7 Installation
1. Upgrade the system, and reboot if required.

$ sudo apt update
$ sudo apt upgrade
$ sudo reboot

2. Install CUDA Toolkit from official site.

$ wget https://developer.download.nvidia.com/compute/cuda/repos/ubuntu2004/x86_64/cuda-ubuntu2004.pin
$ sudo mv cuda-ubuntu2004.pin /etc/apt/preferences.d/cuda-repository-pin-600
$ wget https://developer.download.nvidia.com/compute/cuda/11.7.0/local_installers/cuda-repo-ubuntu2004-11-7-local_11.7.0-515.43.04-1_amd64.deb
$ sudo dpkg -i cuda-repo-ubuntu2004-11-7-local_11.7.0-515.43.04-1_amd64.deb
$ sudo cp /var/cuda-repo-ubuntu2004-11-7-local/cuda-*-keyring.gpg /usr/share/keyrings/
$ sudo apt-get update
$ sudo apt-get -y install cuda

3. Add CUDA into the PATH.

$ echo 'export PATH=/usr/local/cuda-11.7/bin${PATH:+:${PATH}}' >> ~/.bashrc
$ echo 'export LD_LIBRARY_PATH=/usr/local/cuda11.7/lib64${LD_LIBRARY_PATH:+:${LD_LIBRARY_PATH}}' >> ~/.bashrc

4. Reboot and verify the installation.

$ sudo reboot
$ nvcc --version
$ nvidia-smi

5. Install Pytorch
https://pytorch.org/

6. CPU/GPU memory usage

!ln -sf /opt/bin/nvidia-smi /usr/bin/nvidia-smi
!pip install gputil
!pip install psutil
!pip install humanize
import psutil
import humanize
import os, time
import GPUtil as GPU

def worker():
  while True:
    process = psutil.Process(os.getpid())
    print("Gen RAM Free: " + humanize.naturalsize( psutil.virtual_memory().available `enter code here`), " I Proc size: " + humanize.naturalsize( process.memory_info().rss))
    #  print("CPU Memory usage(MB): ", process.memory_info().rss/ 1024 ** 2, " MB")
    gpu = GPU.getGPUs()[0]
    print("GPU RAM Free: {0:.0f}MB | Used: {1:.0f}MB | Util {2:3.0f}% | Total {3:.0f}MB".format(gpu.memoryFree, gpu.memoryUsed, gpu.memoryUtil*100, gpu.memoryTotal))
    time.sleep(6)

import threading
t = threading.Thread(target=worker, name='Monitor')
t.start()