操作系统： Ubuntu 16.04/18.04
装置 Nvidia Driver
举荐应用 graphics drivers PPA 装置 Nvidia 驱动。
sudo add-apt-repository ppa:graphics-drivers/ppasudo apt update检测举荐的 Nvidia 显卡驱动：
ubuntu-drivers devices装置 Nvidia 驱动（以下是 RTX2060 上的状况）：
# Ubuntu 16.04 only search 430 for CUDA < 10.2apt-cache search nvidiasudo apt install nvidia-430# Ubuntu 18.04 could search 440 for CUDA <= 10.2apt-cache search nvidia | grep ^nvidia-driversudo apt install nvidia-driver-440驱动对应的 CUDA 版本，请见 CUDA Compatibility 。
最初， sudo reboot 重启。之后，运行 nvidia-smi 输入 Nvidia 驱动信息:
$ nvidia-smiFri Apr 17 07:31:55 2020+-----------------------------------------------------------------------------+| NVIDIA-SMI 440.82 Driver Version: 440.82 CUDA Version: 10.2 ||-------------------------------+----------------------+----------------------+| GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC || Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. ||===============================+======================+======================|| 0 GeForce RTX 2060 Off | 00000000:01:00.0 Off | N/A || N/A 40C P8 5W / N/A | 263MiB / 5934MiB | 3% Default |+-------------------------------+----------------------+----------------------++-----------------------------------------------------------------------------+| Processes: GPU Memory || GPU PID Type Process name Usage ||=============================================================================|| 0 1560 G /usr/lib/xorg/Xorg 144MiB || 0 1726 G /usr/bin/gnome-shell 76MiB || 0 2063 G ...uest-channel-token=10544833948196615517 39MiB |+-----------------------------------------------------------------------------+如果装置 CUDA Toolkit ，请先理解 CUDA Compatibility 。装置 CUDA Toolkit 时，留神其携带的驱动版本，最好将其与驱动别离进行装置。而驱动从官网上间接找适合的版本。
装置 Docker
# update the apt package indexsudo apt-get update# install packages to allow apt to use a repository over HTTPSsudo apt-get install apt-transport-https ca-certificates curl gnupg2 software-properties-common# add Docker’s official GPG keycurl -fsSL https://download.docker.com/linux/ubuntu/gpg | sudo apt-key add -# set up the stable repositorysudo add-apt-repository \ "deb [arch=amd64] https://mirrors.tuna.tsinghua.edu.cn/docker-ce/linux/ubuntu \ $(lsb_release -cs) \ stable"# update the apt package indexsudo apt-get update# install the latest version of Docker Engine and containerdsudo apt-get install docker-ce docker-ce-cli containerd.io之后，将 Docker 设为 non-root 用户可用：
sudo groupadd dockersudo usermod -aG docker $USER参考
- Install Docker Engine on Ubuntu
- Docker CE 清华源
- Post-installation steps for Linux
装置 Nvidia Docker
# add the package repositoriesdistribution=$(. /etc/os-release;echo $ID$VERSION_ID)curl -s -L https://nvidia.github.io/nvidia-docker/gpgkey | sudo apt-key add -curl -s -L https://nvidia.github.io/nvidia-docker/$distribution/nvidia-docker.list | sudo tee /etc/apt/sources.list.d/nvidia-docker.listsudo apt-get updatesudo apt-get install -y nvidia-container-toolkitsudo systemctl restart docker应用
#### Test nvidia-smi with the latest official CUDA imagedocker run --gpus all nvidia/cuda:10.0-base nvidia-smi# Start a GPU enabled container on two GPUsdocker run --gpus 2 nvidia/cuda:10.0-base nvidia-smi# Starting a GPU enabled container on specific GPUsdocker run --gpus '"device=1,2"' nvidia/cuda:10.0-base nvidia-smidocker run --gpus '"device=UUID-ABCDEF,1"' nvidia/cuda:10.0-base nvidia-smi# Specifying a capability (graphics, compute, ...) for my container# Note this is rarely if ever used this waydocker run --gpus all,capabilities=utility nvidia/cuda:10.0-base nvidia-smi$ docker run --gpus all nvidia/cuda:10.2-base-ubuntu16.04 nvidia-smiUnable to find image 'nvidia/cuda:10.2-base-ubuntu16.04' locally10.2-base-ubuntu16.04: Pulling from nvidia/cuda976a760c94fc: Pull completec58992f3c37b: Pull complete0ca0e5e7f12e: Pull completef2a274cc00ca: Pull complete708a53113e13: Pull complete7dde2dc03189: Pull complete2d21d4aba891: Pull completeDigest: sha256:1423b386bb4f950d12b3b0f3ad51eba42d754ee73f8fc4a60657a1904993b68cStatus: Downloaded newer image for nvidia/cuda:10.2-base-ubuntu16.04Fri Apr 24 08:17:26 2020+-----------------------------------------------------------------------------+| NVIDIA-SMI 440.82 Driver Version: 440.82 CUDA Version: 10.2 ||-------------------------------+----------------------+----------------------+| GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC || Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. ||===============================+======================+======================|| 0 GeForce RTX 2060 Off | 00000000:01:00.0 Off | N/A || N/A 38C P8 10W / N/A | 523MiB / 5934MiB | 21% Default |+-------------------------------+----------------------+----------------------++-----------------------------------------------------------------------------+| Processes: GPU Memory || GPU PID Type Process name Usage ||=============================================================================|+-----------------------------------------------------------------------------+参考
- nvidia-docker
- nvidia/cuda
结语
Go coding!
分享 Coding 中实用的小技巧、小常识！欢送关注，独特成长！