CentOS7

System Version：CentOS 7.9.2009

内核版本：Linux localhost.localdomain 3.10.0-1160.el7.x86_64 #1 SMP Mon Oct 19 16:18:59 UTC 2020 x86_64 x86_64 x86_64 GNU/Linux

条件

禁止 Nouveau

安装 kernel-devel 或 kernel-source、kernel-header；

安装 binutils，路径/usr/bin/ld;

1.安装依赖

yum -y install epel-release

yum -y install gcc binutils wget

yum -y install kernel-devel

2.禁用Nouveau 2.1.检查是否开启Nouveau

lsmod | grep nouveau

注意：无信息输出表示已被禁用无需在 *** 作以下步骤；

2.2.修改配置

echo -e “blacklist nouveau\noptions nouveau modeset=0” > /etc/modprobe.d/blacklist.conf

2.3.备份img

mv /boot/initramfs- ( u n a m e − r ) . i m g / b o o t / i n i t r a m f s − (uname -r).img /boot/initramfs- (uname−r).img/boot/initramfs−(uname -r).img.bak

2.4.重建

dracut /boot/initramfs-$(uname -r).img $(uname -r)

2.5.重启系统

reboot

2.6.检查是否关闭

lsmod | grep nouveau

注意：无任何信息输出表示禁用成功；

3.检查驱动 3.1.安装elrepo源

rpm --import https://www.elrepo.org/RPM-GPG-KEY-elrepo.org

rpm -Uvh http://www.elrepo.org/elrepo-release-7.0-4.el7.elrepo.noarch.rpm

或者

yum -y install https://www.elrepo.org/elrepo-release-7.0-4.el7.elrepo.noarch.rpm

3.2.安装nvidia-detect

yum -y install nvidia-detect

3.3.检测显卡驱动

nvidia-detect -v

Probing for supported NVIDIA devices...
[10de:1b06] NVIDIA Corporation GP102 [GeForce GTX 1080 Ti]
This device requires the current 510.60.02 NVIDIA driver kmod-nvidia

4.驱动安装 4.1.下载驱动

wget https://us.download.nvidia.cn/XFree86/Linux-x86_64/510.68.02/NVIDIA-Linux-x86_64-510.68.02.run

注意：如果检测出来版本号与我的不一致可以自行替换部分

建议：自行在英伟达官网下载到U盘，拷贝到服务器

注意：英伟达只会提供最新版本的，可以向下兼容所以我这里安装的是510.68.02，不是510.60.02

4.2.授权

chmod +x NVIDIA-Linux-x86_64-510.68.02.run

这里会报错，需要关闭 X 服务

# 查看是否为gdm（一共有两种，本无服务器属于gdm）
systemctl --all|grep gdm
whereis gdm

systemctl stop gdm.service
# 安装驱动
systemctl start gdm.service

4.3.安装

sh ./NVIDIA-Linux-x86_64-510.68.02.run -s

4.4.查看显卡信息

nvidia-smi

注意：信息输出表示显卡驱动已经安装完成
另外：我还配套安装了
python 3.9.11
pytorch1.11.0
tensorflow-gpu 2.7.0
transformers 4.18.0
cuda 11.3
cudnn 8.2.0
都是最新版，并成功试运行了

+-----------------------------------------------------------------------------+
| NVIDIA-SMI 510.68.02    Driver Version: 510.68.02    CUDA Version: 11.6     |
|-------------------------------+----------------------+----------------------+
| 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  NVIDIA GeForce ...  Off  | 00000000:01:00.0  On |                  N/A |
| 49%   82C    P2   246W / 250W |   8944MiB / 11264MiB |     99%      Default |
|                               |                      |                  N/A | 
| Processes:                                                                  |
|  GPU   GI   CI        PID   Type   Process name                  GPU Memory |
|        ID   ID                                                   Usage      |
|=============================================================================|
|    0   N/A  N/A     10400      G   /usr/bin/X                         84MiB |
|    0   N/A  N/A     23147      G   /usr/bin/gnome-shell               84MiB |
|    0   N/A  N/A     29312      C   python                           8771MiB |
+-----------------------------------------------------------------------------+

5.卸载驱动 5.1.卸载安装

nvidia-uninstall

5.2.清理安装

dkms remove

注意：需要安装“yum -y install dkms”

6.常见错误

1.安装时报错“ERROR: Unable to find the kernel source tree for the currently running kernel. Please make sure you have installed the kernel source files for your kernel and that they are properly configured; on

Red Hat Linux systems, for example, be sure you have the ‘kernel-source’ or ‘kernel-devel’ RPM installed. If you know the correct kernel source files are installed, you may specify the

kernel source path with the ‘–kernel-source-path’ command line option.”

解决办法：

安装内核库

yum -y install epel-release

yum -y install kernel-devel

内核版本对比

rpm -qa |grep kernel

uname -r

安装驱动

./NVIDIA-Linux-x86_64-510.68.02.run --kernel-source-path=/usr/src/kernels/3.10.0-1160.42.2.el7.x86_64 -k $(uname -r)