显卡驱动+cuda+cudnn

ubuntu16.04，本机显卡型号: GeForce GTX 1060 显卡驱动的安装

参考教程：https://blog.csdn.net/pursuit_zhangyu/article/details/79362128

一，安装NVIDIA驱动下载链接。

## 1，F10进入BIOS，将secure boot设置为disable。
##2，通过下面命令确保电脑上之前的nvidia驱动已卸载干净（可选）。
sudo apt-get remove --purge nvidia*
#### 3，通过下面命令卸载之前下载的驱动安装包（runfile格式）（可选）。
sudo sh /home/xxx/驱动名.run --uninstall  
#### 4，屏蔽开源驱动nouveau，这里选择blacklist，首先创建新文件。
sudo vim /etc/modprobe.d/blacklist-nouveau.conf
or
sudo gedit /etc/modprobe.d/blacklist-nouveau.conf
#内容如下：
blacklist nouveau
options nouveau modeset=0
#再通过如下命令更新：
sudo update-initramfs -u
#通过如下命令，nouveau驱动是否被禁用（无输出则被禁用）。
lsmod | grep nouveau
sudo chmod a+x NVIDIA-Linux-x86_64-384.130.run
# 禁用X服务（最好有一台备用电脑，因为这一步会黑屏...）:
sudo service lightdm stop
进入控制台模式:ctrl+alt+f1
sudo ./NVIDIA-Linux-x86_64-384.130.run -no-opengl-files
#-no-x-check -no-nouveau-check 可选

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

Install NVIDIA Accelerated Graphics Driver for Linux-x86_64 410.48?
(y)es/(n)o/(q)uit: no     //一定选择no，不安装推荐的驱动，因为我们已经安装好了显卡驱动，这里安装推荐的驱动可能会出现驱动问题

Install the CUDA 10.0 Toolkit?
(y)es/(n)o/(q)uit: yes     //确定安装

Enter Toolkit Location
 [ default is /usr/local/cuda-10.0 ]:     //直接按回车选择默认安装路径

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

Install the CUDA 10.0 Samples?
(y)es/(n)o/(q)uit: yes                    //确认安装

Enter CUDA Samples Location
 [ default is /home/zhangman ]:           //直接按回车选择默认安装路径

Installing the CUDA Toolkit in /usr/local/cuda-10.0 ...

Installing the CUDA Samples in /home/zhangman ...
Copying samples to /home/zhangman/NVIDIA_CUDA-10.0_Samples now...
Finished copying samples.
Driver:   Not Selected
Toolkit:  Installed in /usr/local/cuda-10.0
Samples:  Installed in /home/zhangman

Please make sure that
 -   PATH includes /usr/local/cuda-10.0/bin
 -   LD_LIBRARY_PATH includes /usr/local/cuda-10.0/lib64, or, add /usr/local/cuda-10.0/lib64 to /etc/ld.so.conf and run ldconfig as root

To uninstall the CUDA Toolkit, run the uninstall script in /usr/local/cuda-10.0/bin
//cuda卸载方式
Please see CUDA_Installation_Guide_Linux.pdf in /usr/local/cuda-10.0/doc/pdf for detailed information on setting up CUDA.

***WARNING: Incomplete installation! This installation did not install the CUDA Driver. A driver of version at least 384.00 is required for CUDA 10.0 functionality to work.                           
//这个警告是因为我们没安装cuda推荐的驱动，因为我们已经安装好可驱动，忽略即可
To install the driver using this installer, run the following command, replacing <CudaInstaller> with the name of this run file:
    sudo <CudaInstaller>.run -silent -driver

nvidia-smi

cuda的安装与卸载

卸载

cd /usr/local/cuda-8.0/bin
sudo ./uninstall_cua_8.0.pl
cd ..
cd ..
sudo rm -rf cuda-8.0

安装链接

sudo chmod a+x cuda_10.0.130_410.48_linux.run 
# 1. 安装命令
sudo sh cuda_8.0.61_375.26_linux.run
# 2.一堆协议说明，可直接按q退出协议说明
# 3.是否接受协议
accept/decline/quit: accept
# 4.是否安装显卡驱动包，如果已经安装显卡驱动，选择n
Install NVIDIA Accelerated Graphics Driver for Linux-x86_64 375.26? 
y)es/(n)o/(q)uit: n 
# 5.是否安装工具包，选择y
Install the CUDA 8.0 Toolkit?
(y)es/(n)o/(q)uit: y 
# 6.工具包安装地址，默认回车即可
Enter Toolkit Location
[ default is /usr/local/cuda-8.0 ]: 
# 7. 添加链接, 注意:如果已经安装过一个版本，选y则cuda指向当前这个新的版本，选n则不创建软链接
Do you want to install a symbolic link at /usr/local/cuda?
(y)es/(n)o/(q)uit: y
# 8.安装样例
Install the CUDA 8.0 Samples?
(y)es/(n)o/(q)uit: y 
# 9.样例安装地址默认即可
Enter CUDA Samples Location
 [ default is /root ]:  
# 10. 出现安装信息

环境变量

sudo gedit /etc/profile
#加入：
export PATH=/usr/local/cuda-8.0/bin/:$PATH
export LD_LIBRARY_PATH=/usr/local/cuda-8.0/lib64/

#编译并测试设备 deviceQuery：
cd /usr/local/cuda-8.0/samples/1_Utilities/deviceQuery
sudo make
sudo ./deviceQuery #之后重启

安装cudnn

需要先注册安装NVIDIA帐号下载链接，并通过如下命令解压（在下载好的目录里）。

tar -xvf cudnn-10.1-linux-x64-v7.3.1.20.tgz

解压下载的文件，可以看到cuda文件夹，在当前目录打开终端，执行如下命令。

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
sudo chmod a+r /usr/local/cuda/lib64/libcudnn*

并查看cudnn版本。

cat /usr/local/cuda/include/cudnn.h | grep CUDNN_MAJOR -A 2