源码安装cuDNN v7.3.0 for CUDA 10.0，并测试cuDNN是否安装成功

参考：

https://docs.nvidia.com/deeplearning/sdk/cudnn-install/index.html#axzz4qYJp45J2

https://blog.csdn.net/u010480194/article/details/54287727

 

前提：CUDA 10.0 已经安装完成！！

 

1、下载cuDNN：

下载地址：https://developer.nvidia.com/rdp/cudnn-download#a-collapse704-9

选择其他版本：

 

2、源码安装：选择： cuDNN v7.3.0 Library for Linux

下载后，解压，直接将对应头文件和库拷贝到CUDA的安装目录即可，记得修改库文件的文件权限

1. Unzip the cuDNN package.

   $ tar -xzvf cudnn-9.0-linux-x64-v7.tgz

2. Copy the following files into the CUDA Toolkit directory, and change the file permissions.

   $ sudo cp cuda/include/cudnn.h /usr/local/cuda/include

   $ sudo cp cuda/lib64/libcudnn* /usr/local/cuda/lib64

   $ sudo chmod 777 /usr/local/cuda/include/cudnn.h /usr/local/cuda/lib64/libcudnn*

3. update link

     sudo chmod +r libcudnn.so.7.3.0 # 自己查看.so的版本

    sudo ln -sf libcudnn.so.7.3.0 libcudnn.so.7

    sudo ln -sf libcudnn.so.7 libcudnn.so

    sudo ldconfig

 

3、测试cuDNN

1. Copy the cuDNN sample to a writable path.

   $sudo cp -r /usr/src/cudnn_samples_v7/ ~/

2. Go to the writable path.

   $ cd ~/cudnn_samples_v7/mnistCUDNN

3. Compile the mnistCUDNN sample.

   $sudo make clean && make

4. Run the mnistCUDNN sample.

   $ ./mnistCUDNN

   If cuDNN is properly installed and running on your Linux system, you will see a message similar to the following:如果安装成功 测试完成后显示：

   Test passed!