lspci | grep -i nvidia
uname -m && cat /etc/*release
#for case1: original driver installed by apt-get: sudo apt-get remove --purge nvidia* #for case2: original driver installed by runfile: sudo chmod +x *.run sudo ./NVIDIA-Linux-x86_64-384.59.run --uninstall
sudo gedit /etc/modprobe.d/blacklist.conf
在文本最后添加:html
blacklist nouveau options nouveau modeset=0
而后执行:linux
sudo update-initramfs -u
重启以后,能够查看nouveau有没有运行:ubuntu
lsmod | grep nouveau # 没输出表明禁用生效
sudo service lightdm stop #这会关闭图形界面
按Ctrl-Alt+F1进入命令行界面,输入用户名和密码登陆。bash
驱动网址https://www.nvidia.cn/Download/index.aspx?lang=cnide
#给驱动run文件赋予执行权限: sudo chmod +x NVIDIA-Linux-x86_64-384.59.run #后面的参数很是重要,不可省略: sudo ./NVIDIA-Linux-x86_64-384.59.run –no-opengl-files
nvidia-smi #若列出GPU的信息列表,表示驱动安装成功 nvidia-settings #若弹出设置对话框,亦表示驱动安装成功
网址http://developer.nvidia.com/cuda-downloads 选择runfile安装post
sudo sh cuda_<version>_linux.run
开始安装以后,须要阅读说明,可使用Ctrl + C直接阅读完成,或者使用空格键慢慢阅读。下面为安装选项:测试
(是否赞成条款,必须赞成才能继续安装) accept/decline/quit: accept (这里不要安装驱动,由于已经安装最新的驱动了,不然可能会安装旧版本的显卡驱动,致使重复登陆的状况) Install NVIDIA Accelerated Graphics Driver for Linux-x86_64 410.48? (y)es/(n)o/(q)uit: n Install the CUDA 10.0 Toolkit?(是否安装CUDA 10 ,这里必需要安装) (y)es/(n)o/(q)uit: y Enter Toolkit Location(安装路径,使用默认,直接回车就行) [ default is /usr/local/cuda-10.0 ]: Do you want to install a symbolic link at /usr/local/cuda?(赞成建立软连接) (y)es/(n)o/(q)uit: y Install the CUDA 10.0 Samples?(不用安装测试,自己就有了) (y)es/(n)o/(q)uit: n Installing the CUDA Toolkit in /usr/local/cuda-10.0 ...(开始安装)
sudo gedit ~/.bashrc
末尾加入ui
export PATH=/usr/local/cuda-8.0/bin:$PATH export LD_LIBRARY_PATH=/usr/local/cuda/lib64:$LD_LIBRARY_PATH
更新命令行
source ~/.bashrc
查看cuda版本code
nvcc -V
CUDA Sample测试:
#编译并测试设备 deviceQuery: cd /usr/local/cuda-8.0/samples/1_Utilities/deviceQuery make ./deviceQuery #编译并测试带宽 bandwidthTest: cd ../bandwidthTest make ./bandwidthTest
若是这两个测试的最后结果都是Result = PASS,说明CUDA安装成功。
在命令行中输入
sudo apt-get remove cuda sudo apt-get autoclean sudo apt-get remove cuda*
而后在目录切换到usr/local/下
cd /usr/local/ sudo rm -r cuda-9.1
下载对应版本cuDNN https://developer.nvidia.com/cudnn
tar xvzf cudnn-9.2-linux-x64-v7.1 sudo cp -P cuda/include/cudnn.h /usr/local/cuda/include sudo cp -P cuda/lib64/libcudnn* /usr/local/cuda/lib64 sudo chmod a+r /usr/local/cuda/include/cudnn.h /usr/local/cuda/lib64/libcudnn* sudo ldconfig