【CUDA-1】centos7安装CUDA

准备

CUDA 版本：10.0
安装包

• cuda_10.0.130_410.48_linux.run
• cudnn-10.0-linux-x64-v7.5.0.56.tgz

检查硬件环境

• 检测系统是否已安装 GPU，执行命令：

lspci | grep -i nvidia

输出类似如下信息表明已安装 GPU ：
03:00.0 VGA compatible controller: NVIDIA Corporation Device 1b06 (rev a1)
03:00.1 Audio device: NVIDIA Corporation Device 10ef (rev a1)

• 检查内核版本

uname -r
# 如果内核版本小于3.10.0-957，则升级至3.10.0-957:
yum install kernel

• 手动设置启动内核：
  grub2-set-default "CentOS Linux (3.10.0-957.1.3.el7.x86_64) 7 (Core)"
  再次重启系统查看改动是否生效。

检查软件环境

• 安装内核头文件

yum install kernel-devel-$(uname -r) kernel-headers-$(uname -r)

• 安装 GCC 、EPEL 源 和 DKMS

yum install gcc gcc-c++
yum install epel-release
yum install --enablerepo=epel dkms

• 禁用 Nouveau 驱动

vi  /etc/modprobe.d/blacklist-nouveau.conf

# 编辑内容
blacklist nouveau
options nouveau modeset=0
保存文件执行
dracut --force

# 运行以下命令查看是否禁用成功
lsmod | grep nouveau
# 如果有输出如下信息，说明禁用失败，可以尝试重启之后，再执行lsmod | grep nouveau。

安装

1. 进入文字模式

init 3

2. 执行安装脚本

# sh cuda_10.0.130_410.48_linux.run

Do you accept the previously read EULA?
accept/decline/quit: accept

Install NVIDIA Accelerated Graphics Driver for Linux-x86_64 387.26?
(y)es/(n)o/(q)uit: y

Do you want to install the OpenGL libraries?
(y)es/(n)o/(q)uit [ default is yes ]: n

Do you want to run nvidia-xconfig?
(y)es/(n)o/(q)uit [ default is no ]: n

Install the CUDA 10.0 Toolkit?
(y)es/(n)o/(q)uit: y
Enter Toolkit Location
[ default is /usr/local/cuda-10.0 ]: /home/default

Do you want to install a symbolic link at /usr/local/cuda?
(y)es/(n)o/(q)uit: y

Install the CUDA 10.0 Samples?
(y)es/(n)o/(q)uit: n
===========
= Summary =
===========
Driver:   Installed
Toolkit:  Installed
Samples:  Not Selected 

3. 检查是否安装成功

# 命令查看 GPU 设备状态，出现一下内容说明安装成功
nvidia-smi 

• 运行测试程序，检测 CUDA ToolKit 是否安装成功:

$ /usr/local/cuda/extras/demo_Suite/deviceQuery
/usr/local/cuda/extras/demo_suite/deviceQuery Starting...

 CUDA Device Query (Runtime API) version (CUDART static linking)

Detected 1 CUDA Capable device(s)

Device 0: "GeForce GTX 1080 Ti"
  CUDA Driver Version / Runtime Version          10.0 / 10.0
  CUDA Capability Major/Minor version number:    6.1
  Total amount of global memory:                 11178 MBytes (11721506816 bytes)
  (28) Multiprocessors, (128) CUDA Cores/MP:     3584 CUDA Cores
  GPU Max Clock rate:                            1582 MHz (1.58 GHz)
  Memory Clock rate:                             5505 Mhz
......
......
......
deviceQuery, CUDA Driver = CUDART, CUDA Driver Version = 10.0, CUDA Runtime Version = 10.0, NumDevs = 1, Device0 = GeForce GTX 1080 Ti
Result = PASS

• 安装 cuDNN 库

sudo cp cuda/include/cudnn.h /usr/local/cuda/include
sudo cp cuda/lib64/libcudnn* /usr/local/cuda/lib64
sudo chmod a+r /usr/local/cuda/include/cudnn.h /usr/local/cuda/lib64/libcudnn*

卸载

如需要更新CUDA或者主动卸载CUDA，请按如下卸载方式执行。
先卸载 CUDA ToolKit，再卸载 NVIDIA 驱动。

/icooper/tools/cuda-10.0/bin/uninstall_cuda_10.0.pl

# 卸载 dmks 中的 nvidia 模块
dkms remove nvidia/396.26 -k 3.10.0-957.1.3.el7.x86_64

nvidia-uninstall

参考

Nvidia CUDA 安装文档：
https://docs.nvidia.com/cuda/cuda-installation-guide-linux/index.html#runfile-installation

Nvidia cuDNN 安装文档:
https://docs.nvidia.com/deeplearning/sdk/cudnn-install/index.html