Deep learning environment configuration (I) installation of CUDA and cudnn

0. choice CUDA and cuDNN edition
   tensorflow Corresponding CUDA and cuDNN edition 、cuda The corresponding graphics card driver version is as follows :
   
   Also available at NVIDIA Control panel - system information - View supported in component CUDA edition :
   My configuration is GTX 1050(411.31 drive )+CUDA10.0+cuDNN7.4.2 for CUDA10.0+tensorflow2.0.0-gpu.
   1. Install the appropriate version of the graphics card driver ： Driving the genie app（ The lower part can be stable / newest ） You can install older drivers .
1. download CUDA:https://developer.nvidia.com/cuda-toolkit-archive
   download cuDNN:https://developer.nvidia.com/rdp/cudnn-archive
2. install CUDA
3. install cuDNN
   decompression cuDNN Compressed package , You can see bin、include、lib Catalog , Copy the files in the directory to C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA In the corresponding directory of .
4. Add environment variables
   In environment variable - Of the system variable Path Click under ’ edit - newly build ’ Add a path ( The following is the default installation path ):

C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v10.0\bin
C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v10.0\lib\x64

6. Will file C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v10.0\extras\CUPTI\libx64\cupti64_100.dll Copy to folder C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v10.0\bin in .

7. verification
   cmd in cd Enter the installation directory , function bandwidthTest.exe and deviceQuery.exe, If all Result = PASS, Installation succeeded .

cd C:\Program Files\NVIDIA GPU Computing Toolkit\CUDA\v10.0\extras\demo_suite
bandwidthTest.exe
deviceQuery.exe