Ubuntu 16.04 Environment Setup

1. Check GPU Device Recognition


  $ sudo lspci | grep NVIDIA
  3D controller: NVIDIA Corporation GK210GL [Tesla K80] indicates that it is recognized as K80.
  3D controller: NVIDIA Corporation GP102GL [Tesla P40] (rev a1) indicates that it is P40.

2. Block Open Source Driver Nouveau

Edit the following file:


sudo vim /etc/modprobe.d/blacklist-nouveau.conf

Input the following content:


blacklist nouveau
blacklist lbm-nouveau
options nouveau modeset=0
alias nouveau off
alias lbm-nouveau off

Update and restart:


sudo update-initramfs -u
sudo reboot
sudo apt-get install build-essential pkg-config linux-headers-`uname -r`

3. Install NVIDIA Driver

3.1 Download

Download the appropriate driver from the NVIDIA official website (current version 418.126.02), the url is https://www.nvidia.com/Download/index.aspx?lang=en-us .

You can also download it from UFile, it’s faster. http://gpu.cn-bj.ufileos.com/NVIDIA-Linux-x86_64-418.126.02.run .

3.2 Installation


sudo chmod +x NVIDIA-Linux-x86_64-418.126.02.run
sudo ./NVIDIA-Linux-x86_64-418.126.02.run

3.3 Check Driver Status


$ sudo nvidia-smi

4. Install CUDA Library

4.1 Network Installation


sudo wget https://developer.download.nvidia.com/compute/cuda/repos/ubuntu1604/x86_64/cuda-ubuntu1604.pin

sudo mv cuda-ubuntu1604.pin /etc/apt/preferences.d/cuda-repository-pin-600

sudo apt-key adv --fetch-keys http://developer.download.nvidia.com/compute/cuda/repos/ubuntu1604/x86_64/7fa2af80.pub

sudo add-apt-repository "deb http://developer.download.nvidia.com/compute/cuda/repos/ubuntu1604/x86_64/ /"

sudo apt-get update

sudo apt-get -y install cuda

FAQ

1. Why is the GPU utilization rate 100% when I use nvidia-smi?

This problem is caused by the system’s inaccurate reading of GPU status information. The following command can correct it and make the system read the command correctly. #sudo nvidia-smi -pm 1

2. In addition to self-installation, are there other ways to obtain driver images?

You may submit a ticket or contact staff to obtain an image that contains GPU driver and Cuda environment made by Surfercloud, which saves the time of manual installation.