配置环境
创建新用户
$ sudo useradd jiangcm -d /home/jiangcm/ -m
# 为新用户添加密码
$ sudo passwd jiangcm
# 为新用户增加sudoer权限
$ su root
$ sudo vim /etc/sudoers
在文本中添加命令
# 添加jiangcm ALL=(ALL) ALL
jiangcm ALL=(ALL:ALL) ALL
Nouveau禁用
查看是否安装nouveau:
ubuntu@ubuntu:~$ lsmod | grep nouveau
nouveau 1708032 0
video 45056 1 nouveau
ttm 106496 2 ast,nouveau
i2c_algo_bit 16384 3 igb,ast,nouveau
mxm_wmi 16384 1 nouveau
drm_kms_helper 172032 2 ast,nouveau
drm 401408 5 drm_kms_helper,ast,ttm,nouveau
wmi 24576 2 mxm_wmi,nouveau
禁用:
# 打开blacklist文件
$ sudo vim /etc/modprobe.d/blacklist.conf
# 将nouveau添加到命令中
blacklist nouveau
options nouveau modeset=0
# 保存文件后,更新
$ sudo update-initramfs -u
# 重新启动服务器
$ reboot
检查是否禁用
$ lspci | grep nouveau
安装gcc和cmake
# 安装
$ sudo apt-get install build-essential
$ sudo apt-get install cmake
# 检测是否安装成功
$ gcc --version
$ cmake --version
根据型号下载并安装nvidia驱动,网上下载需要的版本:
查看GPU型号:
jiangcm@ubuntu:~/software$ lspci | grep -i vga
04:00.0 VGA compatible controller: NVIDIA Corporation Device 1e04 (rev a1)
05:00.0 VGA compatible controller: NVIDIA Corporation Device 1e04 (rev a1)
06:00.0 VGA compatible controller: NVIDIA Corporation Device 1e04 (rev a1)
07:00.0 VGA compatible controller: NVIDIA Corporation Device 1e04 (rev a1)
08:00.0 VGA compatible controller: NVIDIA Corporation Device 1e04 (rev a1)
0b:00.0 VGA compatible controller: NVIDIA Corporation Device 1e04 (rev a1)
0c:00.0 VGA compatible controller: NVIDIA Corporation Device 1e04 (rev a1)
0d:00.0 VGA compatible controller: NVIDIA Corporation Device 1e04 (rev a1)
0e:00.0 VGA compatible controller: NVIDIA Corporation Device 1e04 (rev a1)
0f:00.0 VGA compatible controller: NVIDIA Corporation Device 1e04 (rev a1)
12:00.0 VGA compatible controller: ASPEED Technology, Inc. ASPEED Graphics Family (rev 30)
根据PCI ID Repository查看版本,输入1e04,得到是2080Ti.
nvidia、cuda、cudnn和tensorflow版本的选择:
tensorflow与cuda版本对照表:
详细地址:tensorflow-cuda
cuda与nvidia驱动对照表
地址:cuda-nvidia
选用的版本如下:
NVIDIA-Linux-x86_64-430.64.run
cuda_10.1.105_418.39_linux.run
cudnn-10.1-linux-x64-v7.6.0.64.tgz
考虑到安装tensorflow1.X版本,后续会安装不同版本的cuda,到时候会切换不同版本的cuda,便于调用tensorflow1.X。
安装nvidia驱动
# Ctrl+Alt+F1 进入tty1终端模式,并且输入下面命令关闭显示管理:
$ sudo service lightdm stop
# 安装驱动
sudo ./NVIDIA-Linux-x86_64-430.64.run --no-opengl-files
步骤如下:
查看安装结果:
$ nvidia-smi
安装cuda
安装
下载指定版本的cuda和cudnn,我这里下载的是cuda_10.1.105_418.39_linux.run
和cudnn-10.1-linux-x64-v7.6.0.64.tgz
# cd进入cuda安装包(.run文件)所在的目录下,打开终端
$ sudo sh cuda_10.1.105_418.39_linux.run
第一选项为accept.
之后为取消Driver的安装,因为之前已经安装过了。通过Enter
键可以取消。
然后在选择Install,进行安装。
配置环境
# 打开文件~/.bashrc
$ vvim ~/.bahsrc
# 添加命令到文本中
export PATH="/usr/local/cuda/bin:$PATH"
export LD_LIBRARY_PATH="/usr/lcoal/cuda/lib64:$LD_LIBRARY_PATH"
# 保存后激活
$ source ~/.bashrc
同时在/etc/profile
中也配置相同的路径。
检测是否安装成功
$ cd /usr/local/cuda/samples/1_Utilities/deviceQuery
$ sudo make
$ ./deviceQuery
有如下结果表示安装成功:
jiangcm@ubuntu:/usr/local/cuda/samples/1_Utilities/deviceQuery$ ./deviceQuery
./deviceQuery Starting...
CUDA Device Query (Runtime API) version (CUDART static linking)
Detected 10 CUDA Capable device(s)
Device 0: "GeForce RTX 2080 Ti"
CUDA Driver Version / Runtime Version 10.1 / 10.1
CUDA Capability Major/Minor version number: 7.5
Total amount of global memory: 11019 MBytes (11554717696 bytes)
(68) Multiprocessors, ( 64) CUDA Cores/MP: 4352 CUDA Cores
GPU Max Clock rate: 1545 MHz (1.54 GHz)
Memory Clock rate: 7000 Mhz
Memory Bus Width: 352-bit
L2 Cache Size: 5767168 bytes
Maximum Texture Dimension Size (x,y,z) 1D=(131072), 2D=(131072, 65536), 3D=(16384, 16384, 16384)
Maximum Layered 1D Texture Size, (num) layers 1D=(32768), 2048 layers
Maximum Layered 2D Texture Size, (num) layers 2D=(32768, 32768), 2048 layers
Total amount of constant memory: 65536 bytes
Total amount of shared memory per block: 49152 bytes
Total number of registers available per block: 65536
Warp size: 32
Maximum number of threads per multiprocessor: 1024
Maximum number of threads per block: 1024
Max dimension size of a thread block (x,y,z): (1024, 1024, 64)
Max dimension size of a grid size (x,y,z): (2147483647, 65535, 65535)
Maximum memory pitch: 2147483647 bytes
Texture alignment: 512 bytes
Concurrent copy and kernel execution: Yes with 3 copy engine(s)
Run time limit on kernels: No
Integrated GPU sharing Host Memory: No
Support host page-locked memory mapping: Yes
Alignment requirement for Surfaces: Yes
Device has ECC support: Disabled
Device supports Unified Addressing (UVA): Yes
Device supports Compute Preemption: Yes
Supports Cooperative Kernel Launch: Yes
Supports MultiDevice Co-op Kernel Launch: Yes
Device PCI Domain ID / Bus ID / location ID: 0 / 4 / 0
Compute Mode:
< Default (multiple host threads can use ::cudaSetDevice() with device simultaneously) >
...
安装cudnn
解压cudnn
$ tar -xvf cudnn-10.1-linux-x64-v7.6.0.64.tgz
将文件复制到指定cuda的路径中,并更改权限
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*
查看cudnn版本,确认安装成功
cat /usr/local/cuda/include/cudnn.h | grep CUDNN_MAJOR -A 2
安装Anaconda
下载anaconda到本地.
# 安装
$ bash ./Anaconda3-5.3.1-Linux-x86_64.sh
...
pyopenssl
A thin Python wrapper around (a subset of) the OpenSSL library.
kerberos (krb5, non-Windows platforms)
A network authentication protocol designed to provide strong authentication for client/server applications by using secret-key cryptography.
cryptography
A Python library which exposes cryptographic recipes and primitives.
Do you accept the license terms? [yes|no]
>>> yes
回车使用默认目录
Anaconda3 will now be installed into this location:
/home/jiangcm/anaconda3
- Press ENTER to confirm the location
- Press CTRL-C to abort the installation
- Or specify a different location below
[/home/jiangcm/anaconda3] >>>
是否把anaconda的bin添加到用户的环境变量中?输入yes
Do you wish the installer to initialize Anaconda3
in your /home/jiangcm/.bashrc ? [yes|no]
[no] >>> yes
是否安装VSCode,输入no
To install Visual Studio Code, you will need:
- Administrator Privileges
- Internet connectivity
Visual Studio Code License: https://code.visualstudio.com/license
Do you wish to proceed with the installation of Microsoft VSCode? [yes|no]
>>> no
重新打开一个窗口,输入
conda activte base
激活默认环境。安装成功。
ed:
- Administrator Privileges
- Internet connectivity
Visual Studio Code License: https://code.visualstudio.com/license
Do you wish to proceed with the installation of Microsoft VSCode? [yes|no]
no
重新打开一个窗口,输入
```shell
conda activte base
激活默认环境。安装成功。