linux GPU相关库多版本管理

silasbert

2024-11-15 2024-12-04 约 610 字预计阅读 3 分钟次阅读条评论

注意

本文最后更新于 2024-12-04，文中内容可能已过时。

https://www.cnblogs.com/sddai/p/10278005.html

1 GPU驱动

驱动下载地址 https://www.nvidia.com/en-us/drivers/ 对于4080， Driver Version:550.127.05 Release Date:Tue Oct 22, 2024
查看版本。

bash

$ nvidia-smi
Fri Nov 15 13:02:22 2024
+-----------------------------------------------------------------------------------------+
| NVIDIA-SMI 550.127.05             Driver Version: 550.127.05     CUDA Version: 12.4     |
|-----------------------------------------+------------------------+----------------------+
| 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 RTX 4080        Off |   00000000:17:00.0 Off |                  N/A |
| 50%   31C    P0             55W /  320W |       1MiB /  16376MiB |      0%      Default |
|                                         |                        |                  N/A |
+-----------------------------------------+------------------------+----------------------+
|   1  NVIDIA GeForce RTX 4080        Off |   00000000:B1:00.0 Off |                  N/A |
|  0%   28C    P0             47W /  320W |       1MiB /  16376MiB |      2%      Default |
|                                         |                        |                  N/A |
+-----------------------------------------+------------------------+----------------------+

+-----------------------------------------------------------------------------------------+
| Processes:                                                                              |
|  GPU   GI   CI        PID   Type   Process name                              GPU Memory |
|        ID   ID                                                               Usage      |
|=========================================================================================|
|  No running processes found                                                             |
+-----------------------------------------------------------------------------------------+

其他

2 CUDA

最新版本下载地址 https://developer.nvidia.com/cuda-downloads 例如我选的是 https://developer.nvidia.com/cuda-downloads?target_os=Linux&target_arch=x86_64&Distribution=Ubuntu&target_version=20.04&target_type=runfile_local 。从 https://developer.nvidia.com/cuda-toolkit-archive 可以下载历史版本。
安装完成的提示。

bash

$ sudo bash ./cuda_12.6.2_560.35.03_linux.run
[sudo] password for rui:
===========
= Summary =
===========

Driver:   Installed
Toolkit:  Installed in /usr/local/cuda-12.6/

Please make sure that
 -   PATH includes /usr/local/cuda-12.6/bin
 -   LD_LIBRARY_PATH includes /usr/local/cuda-12.6/lib64, or, add /usr/local/cuda-12.6/lib64 to /etc/ld.so.conf and run ldconfig as root

To uninstall the CUDA Toolkit, run cuda-uninstaller in /usr/local/cuda-12.6/bin
To uninstall the NVIDIA Driver, run nvidia-uninstall
Logfile is /var/log/cuda-installer.log

查看cuda版本。使用命令行查看

bash

$ nvcc --version
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2021 NVIDIA Corporation
Built on Thu_Nov_18_09:45:30_PST_2021
Cuda compilation tools, release 11.5, V11.5.119
Build cuda_11.5.r11.5/compiler.30672275_0

$ ./nvcc --version
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2023 NVIDIA Corporation
Built on Mon_Apr__3_17:16:06_PDT_2023
Cuda compilation tools, release 12.1, V12.1.

$ /usr/local/cuda-12.6/bin/nvcc --version
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2024 NVIDIA Corporation
Built on Thu_Sep_12_02:18:05_PDT_2024
Cuda compilation tools, release 12.6, V12.6.77
Build cuda_12.6.r12.6/compiler.34841621_0

python

import torch

print(torch.__version__)  ##  查看torch当前版本号
print(torch.version.cuda)  ##  编译当前版本的torch使用的cuda版本号
print(torch.cuda.is_available())  ##  查看当前cuda是否可用于当前版本的Torch，如果输出True，则表示可用

多版本共存。安装好你想要的cuda版本以后，在你的~/.bashrc或者~/.zshrc里写入下面这些。

bash

CUDA_HOME=/usr/local/cuda-12.6    ##  这里写你想要的cuda的版本。
export PATH=$CUDA_HOME/bin:$PATH
export LD_LIBRARY_PATH=$CUDA_HOME/lib64:$LD_LIBRARY_PATH

3 CUDNN

4 python cuda torch

5 docker cuda

https://docs.nvidia.com/datacenter/cloud-native/container-toolkit/latest/install-guide.html## installing-the-nvidia-container-toolkit

安装。

bash

curl -fsSL https://nvidia.github.io/libnvidia-container/gpgkey | sudo gpg --dearmor -o /usr/share/keyrings/nvidia-container-toolkit-keyring.gpg \
  && curl -s -L https://nvidia.github.io/libnvidia-container/stable/deb/nvidia-container-toolkit.list | \
sed 's## deb https://#deb [signed-by=/usr/share/keyrings/nvidia-container-toolkit-keyring.gpg] https://#g' | \
sudo tee /etc/apt/sources.list.d/nvidia-container-toolkit.list

sudo sed -i -e '/experimental/ s/^## //g' /etc/apt/sources.list.d/nvidia-container-toolkit.list
sudo apt-get update
sudo apt-get install -y nvidia-container-toolkit

配置

bash

sudo nvidia-ctk runtime configure --runtime=docker
sudo systemctl restart docker

验证

bash

   docker run --rm --runtime=nvidia --gpus all ubuntu nvidia-smi

结束。

目录

linux GPU相关库多版本管理

1 GPU驱动

2 CUDA

3 CUDNN

4 python cuda torch

5 docker cuda