二、Docker安装nvidia-docker (Ubuntu16.04 系统)

官方安装指导

应用场景

• 需要gup训练网络
• 需要gup实现网络部署

1. 检查显卡驱动版本和Docker版本

junli@ubuntu16:~$ ubuntu-drivers devices
== /sys/devices/pci0000:00/0000:00:1d.0/0000:3c:00.0 ==
manual_install: True
model    : GM108M [GeForce 940MX]
modalias : pci:v000010DEd0000134Dsv000017AAsd000039C8bc03sc02i00
vendor   : NVIDIA Corporation
driver   : nvidia-418 - third-party free
driver   : nvidia-410 - third-party free
driver   : nvidia-430 - third-party free recommended
driver   : nvidia-415 - third-party free
driver   : nvidia-384 - distro non-free
driver   : xserver-xorg-video-nouveau - distro free builtin

junli@ubuntu16:~$ docker --version
Docker version 20.10.7, build f0df350

可安装CUDA11.0以下的版本

2. 安装nvidia-docker

junli@ubuntu16:~$ curl -s -L https://nvidia.github.io/nvidia-docker/gpgkey | sudo apt-key add -
junli@ubuntu16:~$ distribution=$(. /etc/os-release;echo $ID$VERSION_ID)
junli@ubuntu16:~$ curl -s -L https://nvidia.github.io/nvidia-docker/$distribution/nvidia-docker.list | sudo tee /etc/apt/sources.list.d/nvidia-docker.list
junli@ubuntu16:~$ sudo apt-get update
junli@ubuntu16:~$ sudo apt-get install -y nvidia-docker2

正在读取软件包列表... 完成
正在分析软件包的依赖关系树       
正在读取状态信息... 完成       
将会同时安装下列软件：
  libnvidia-container-tools libnvidia-container1 nvidia-container-toolkit
下列【新】软件包将被安装：
  ......
/sbin/ldconfig.real: /usr/local/cuda-10.2/targets/x86_64-linux/lib/libcudnn.so.7 is not a symbolic link


junli@ubuntu16:~$ sudo systemctl restart docker

3. 验证nvidia-docker安装

根据主机条件获取镜像网址

junli@ubuntu16:~$ docker pull nvidia/cuda:8.0-cudnn6-devel-ubuntu16.04
junli@ubuntu16:~$ docker run --runtime=nvidia --rm nvidia/cuda:8.0-cudnn6-devel-ubuntu16.04 nvidia-smi
Tue Aug  2 06:38:49 2022       
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 450.57       Driver Version: 450.57       CUDA Version: 11.0     |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |
|                               |                      |               MIG M. |
|===============================+======================+======================|
|   0  GeForce 940MX       Off  | 00000000:3C:00.0 Off |                  N/A |
| N/A   41C    P0    N/A /  N/A |      0MiB /  2004MiB |      0%      Default |
|                               |                      |                  N/A |
+-------------------------------+----------------------+----------------------+
                                                                               
+-----------------------------------------------------------------------------+
| Processes:                                                                  |
|  GPU   GI   CI        PID   Type   Process name                  GPU Memory |
|        ID   ID                                                   Usage      |
|=============================================================================|
|  No running processes found                                                 |
+-----------------------------------------------------------------------------+