ubuntu(28):ubuntu系统多版本conda和多版本cuda共存

0. cuda(包括cudnn)、conda安装照常

注意：

（1）多个conda不要安装到一个目录了，可以见下面的示例目录；

（2）cuda（包括cudnn）不用纠结是否添加超链接

1. 需要修改环境配置--示例

sudo gedit ~/.bashrc

bash 复制代码

# cuda version & cuda version
alias condaenv='export PATH="/home/xiu/anaconda3/bin:$PATH"; . /home/xiu/anaconda3/etc/profile.d/conda.sh; export CUDA_HOME=/usr/local/cuda-11.3; export PATH=$CUDA_HOME/bin:$PATH; export LD_LIBRARY_PATH=$CUDA_HOME/lib64:$LD_LIBRARY_PATH'
alias condaenv3='export PATH="/home/xiu/anaconda3/bin:$PATH"; . /home/xiu/anaconda3/etc/profile.d/conda.sh; export CUDA_HOME=/usr/local/cuda-11.8; export PATH=$CUDA_HOME/bin:$PATH; export LD_LIBRARY_PATH=$CUDA_HOME/lib64:$LD_LIBRARY_PATH'
alias condaenv2='export PATH="/home/xiu/anaconda3_2024/bin:$PATH"; . /home/xiu/anaconda3_2024/etc/profile.d/conda.sh; export CUDA_HOME=/usr/local/cuda-11.3; export PATH=$CUDA_HOME/bin:$PATH; export LD_LIBRARY_PATH=$CUDA_HOME/lib64:$LD_LIBRARY_PATH'
alias condaenv4='export PATH="/home/xiu/anaconda3_2024/bin:$PATH"; . /home/xiu/anaconda3_2024/etc/profile.d/conda.sh; export CUDA_HOME=/usr/local/cuda-11.8; export PATH=$CUDA_HOME/bin:$PATH; export LD_LIBRARY_PATH=$CUDA_HOME/lib64:$LD_LIBRARY_PATH'

source ~/.bashrc

3. 确认

后续在终端输出condaenv,condaenv3,condaenv2,condaenv4会启动不同的cuda和conda，可以使用nvcc -V和
conda --verison查看，如：

bash 复制代码

xiu@xiu:~/Learning/$ condaenv4
xiu@xiu:~/Learning/$ python
Python 3.12.7 | packaged by Anaconda, Inc. | (main, Oct  4 2024, 13:27:36) [GCC 11.2.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> exit()
xiu@xiu:~/Learning/$ nvcc -V
nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2022 NVIDIA Corporation
Built on Wed_Sep_21_10:33:58_PDT_2022
Cuda compilation tools, release 11.8, V11.8.89
Build cuda_11.8.r11.8/compiler.31833905_0