Ubuntu 安装CUDA, cuDNN, TensorRT（草稿）

写在前面

自己的测试环境：
Ubuntu20.04,

本文安装的版本：
cuda_11.1.0；cuDNN-8.2.1；TensorRT-8.2.4.2

一、CUDA, cuDNN, TensorRT 三个库的版本的确定

参考如下链接：
[1] initative. CUDA安装&cuDNN、TensorRT版本匹配 [EB/OL]. https://blog.csdn.net/initative/article/details/138213992, 2024-05-14/2024-07-10.

二、解决方法

cudnn:

sudo cp cudnn-11.4-linux-x64-v8.2.4.15/cuda/include/cudnn*.h  /usr/local/cuda-11.4/include
sudo cp cudnn-11.4-linux-x64-v8.2.4.15/cuda/lib64/libcudnn*  /usr/local/cuda-11.4/lib64
sudo cp -P  cudnn-11.4-linux-x64-v8.2.4.15/cuda/lib64/libcudnn*  /usr/local/cuda-11.4/lib64
sudo chmod a+r /usr/local/cuda-11.4/include/cudnn*.h /usr/local/cuda-11.4/lib64/libcudnn*

Tensorrt

wong@wong-laptop:~/Datum/software_package/NVIDIA/TensorRT-8.2.4.2.Linux.x86_64-gnu.cuda-11.4.cudnn8.2/TensorRT-8.2.4.2$ sudo cp -r  -P ./lib/*  /usr/local/TensorRT-8.2.4.2/lib/
wong@wong-laptop:~/Datum/software_package/NVIDIA/TensorRT-8.2.4.2.Linux.x86_64-gnu.cuda-11.4.cudnn8.2/TensorRT-8.2.4.2$ sudo cp -r  -P ./bin/*  /usr/local/TensorRT-8.2.4.2/bin/

现在只有 include 未复制到 /usr/local/TensorRT-8.2.4.2/include/ 下.

参考链接

[1] wangchengjue29. CUDA安装&cuDNN、TensorRT版本匹配 [EB/OL]. https://blog.csdn.net/initative/article/details/138213992, 2024-05-14/ 2024-07-11.

[1] (目前主要参考这个)秋山丶雪绪. 从零开始 TensorRT（1）环境篇：下载、安装、测试 [EB/OL]. https://blog.csdn.net/weixin_43605641/article/details/126744446, 2024-05-11/2024-07-08.
[2] 星光2020. CUDA和cuDNN各版本下载及版本对应关系 [EB/OL]. https://blog.csdn.net/mbdong/article/details/121769951, 2021-12-07/2024-07-08.
[3] stivory.Tensorrt安装及使用（python版本） [EB/OL]. https://blog.csdn.net/stivory/article/details/135847529, 2024-02-28/2024-07-08.
[4] nvidia. cuDNN Archive [EB/OL]. https://developer.nvidia.com/rdp/cudnn-archive, xxxx-xx-xx/2024-07-08.
[5] nvidia. NVIDIA TensorRT 8.x Download [EB/OL]. https://developer.nvidia.com/nvidia-tensorrt-8x-download, xxxx-xx-xx/2024-07-08.

[6] 【深度学习】显卡驱动, cuda, cudnn的关系与版本对应问题 https://blog.csdn.net/All_In_gzx_cc/article/details/126126297

[7] https://docs.nvidia.com/deeplearning/tensorrt/archives/index.html#overview
[8] https://docs.nvidia.com/deeplearning/tensorrt/release-notes/index.html#rel_8-0-3