cuda安装

下载cuda文件

自行选择版本

https://developer.nvidia.com/cuda-downloads

image.png

安装

下载合适的版本，下载到本地

通过nvidia-detect -v命令查看驱动的版本

[admin@A01-R15-I47-118-0825868 http-cnn-rnn]$ nvidia-detect -v
Probing for supported NVIDIA devices...
[10de:1b38] NVIDIA Corporation Device 1b38
This device requires the current 418.43 NVIDIA driver kmod-nvidia
[10de:1b38] NVIDIA Corporation Device 1b38
This device requires the current 418.43 NVIDIA driver kmod-nvidia
[1a03:2000] ASPEED Technology, Inc. ASPEED Graphics Family
[10de:1b38] NVIDIA Corporation Device 1b38
This device requires the current 418.43 NVIDIA driver kmod-nvidia
[10de:1b38] NVIDIA Corporation Device 1b38
This device requires the current 418.43 NVIDIA driver kmod-nvidia
WARNING: Xorg log file /var/log/Xorg.0.log does not exist
WARNING: Unable to determine Xorg ABI compatibility
WARNING: The driver for this device does not support the current Xorg version

通过上述命令说明我们的驱动需要下载418.43的版本，我们只需要找对应418.43 驱动的cuda就行
通过nvidia-msi也能查看

[admin@A01-R15-I47-121-0825870 users]$ nvidia-smi
Sun Mar 10 13:58:41 2019       
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 418.43       Driver Version: 418.43       CUDA Version: 10.1     |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |
|===============================+======================+======================|
|   0  Tesla P40           Off  | 00000000:02:00.0 Off |                    0 |
| N/A   24C    P0    51W / 250W |      0MiB / 22919MiB |      0%      Default |
+-------------------------------+----------------------+----------------------+
|   1  Tesla P40           Off  | 00000000:03:00.0 Off |                    0 |
| N/A   25C    P0    51W / 250W |      0MiB / 22919MiB |      0%      Default |
+-------------------------------+----------------------+----------------------+
|   2  Tesla P40           Off  | 00000000:83:00.0 Off |                    0 |
| N/A   26C    P0    51W / 250W |      0MiB / 22919MiB |      0%      Default |
+-------------------------------+----------------------+----------------------+
|   3  Tesla P40           Off  | 00000000:84:00.0 Off |                    0 |
| N/A   26C    P0    51W / 250W |      0MiB / 22919MiB |      2%      Default |
+-------------------------------+----------------------+----------------------+
                                                                               
+-----------------------------------------------------------------------------+
| Processes:                                                       GPU Memory |
|  GPU       PID   Type   Process name                             Usage      |
|=============================================================================|
|  No running processes found                                                 |
+-----------------------------------------------------------------------------+
通过上述的结果，得知我们需要下载10.1的cuda

wget https://developer.download.nvidia.cn/compute/cuda/10.1/secure/Prod/local_installers/cuda_10.1.105_418.39_linux.run?JxiB7NQ4_56qMViu4uHKi1iR8c5bHTXGvWeOBrUjhnUzji2Z_fWMX3g4qEFSZZwMLad7TRSb2b-WB7SfaEd_ooUe41yV4K-hH7oM8oydaRWaz129kM_VJfEQ6QP6RdarY3wUichr2wAVLRRAElpSwyc4Pxi3pi5H6EjWoARIeOQ1w0RxPHEHb6lkBYg

执行脚本

sudo sh cuda_10.1.105_418.39_linux.run
[admin@A01-R15-I47-121-0825870 software]$ sudo sh cuda_10.1.105_418.39_linux.run

lqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqq
xqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqx
x Do you accept the above EULA? (accept/decline/quit):                         x
x                                                                              x
mqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqqj

按提示操作

对于上面的提示  ，输入accept
后面还有一次回车，然后自动安装

检测是否成功

cd /usr/local/cuda/bin
nvcc --version

加入环境变量

export CUDA_HOME=/usr/local/cuda
export PATH="/usr/local/cuda/bin:/home/admin/anaconda3/bin:$PATH"
LD_LIBRARY_PATH=/usr/local/cuda/lib64:/home/admin/anaconda3/lib:$LD_LIBRARY_PATH
export LD_LIBRARY_PATH

cuda卸载

cd /usr/local/cuda/bin
./cuda-uninstaller