How to Install cuDNN: Difference between revisions
No edit summary |
mNo edit summary |
||
(4 intermediate revisions by the same user not shown) | |||
Line 1: | Line 1: | ||
<seo title="NVIDIA CUDA cuDNN | cuDNN | RidgeRun Developer" titlemode="replace" keywords="GStreamer, Linux SDK, Linux BSP, Embedded Linux, Device Drivers, Nvidia, Xilinx, TI, NXP, Freescale, Embedded Linux driver development, Linux Software development, Embedded Linux SDK, Embedded Linux Application development, GStreamer Multimedia Framework, cuDNN, NVIDIA cuDNN,NVIDIA CUDA® Deep Neural Network library, NVIDIA CUDA cuDNN." description="This wiki explains about how to Build TensorFlow with GPU Support."></seo> | |||
= Tested | <table> | ||
<tr> | |||
<td><div class="clear; float:right">__TOC__</div></td> | |||
<td> | |||
{{NVIDIA Preferred Partner logo}} | |||
<td> | |||
<center> | |||
{{ContactUs Button}} | |||
</center> | |||
</tr> | |||
</table> | |||
The NVIDIA CUDA® Deep Neural Network library (cuDNN) is a GPU-accelerated library of primitives for deep neural networks.The following is a summary of the cuDNN Installation guide instructions in [https://docs.nvidia.com/deeplearning/sdk/cudnn-install/index.html NVIDIA's Deep Learning SDK] | |||
== Tested Operating Systems for NVIDIA cuDNN == | |||
These instructions have been tested on: | These instructions have been tested on: | ||
* Elementary Hera 5.1 | * Elementary Hera 5.1 | ||
= Download = | == Download cuDNN== | ||
Download the following packages from [https://developer.nvidia.com/cudnn cuDNN download page]. | Download the following packages from [https://developer.nvidia.com/cudnn cuDNN download page]. | ||
Line 15: | Line 29: | ||
The remainder of the wiki assumes these have been downloaded to **~/Downloads** | The remainder of the wiki assumes these have been downloaded to **~/Downloads** | ||
= Configuration Variables = | == Configuration Variables == | ||
<syntaxhighlight lang=bash> | <syntaxhighlight lang=bash> | ||
Line 21: | Line 35: | ||
</syntaxhighlight> | </syntaxhighlight> | ||
= Installing cuDNN = | == Installing cuDNN == | ||
<syntaxhighlight lang=bash> | <syntaxhighlight lang=bash> | ||
Line 32: | Line 46: | ||
</syntaxhighlight> | </syntaxhighlight> | ||
= Validating Installation = | == Validating cuDNN Installation == | ||
<syntaxhighlight lang=bash> | <syntaxhighlight lang=bash> | ||
Line 42: | Line 56: | ||
./mnistCUDNN | ./mnistCUDNN | ||
</syntaxhighlight> | </syntaxhighlight> | ||
[[Category:Jetson]][[Category:NVIDIA Xavier]] |
Latest revision as of 16:17, 23 August 2022
|
|
The NVIDIA CUDA® Deep Neural Network library (cuDNN) is a GPU-accelerated library of primitives for deep neural networks.The following is a summary of the cuDNN Installation guide instructions in NVIDIA's Deep Learning SDK
Tested Operating Systems for NVIDIA cuDNN
These instructions have been tested on:
- Elementary Hera 5.1
Download cuDNN
Download the following packages from cuDNN download page.
- cuDNN Runtime Library for Ubuntu18.04 (Deb)
- cuDNN Developer Library for Ubuntu18.04 (Deb)
- cuDNN Code Samples and User Guide for Ubuntu18.04 (Deb)
The remainder of the wiki assumes these have been downloaded to **~/Downloads**
Configuration Variables
CUDNN_DIR=~/cuDNN
Installing cuDNN
mkdir -p $CUDNN_DIR mv ~/Downloads/libcudnn* $CUDNN_DIR sudo dpkg -i libcudnn7_7.6.5.32-1+cuda10.2_amd64.deb sudo dpkg -i libcudnn7-dev_7.6.5.32-1+cuda10.2_amd64.deb sudo dpkg -i libcudnn7-doc_7.6.5.32-1+cuda10.2_amd64.deb
Validating cuDNN Installation
cp -r /usr/src/cudnn_samples_v7/ $CUDNN_DIR cd ${CUDNN_DIR}/cudnn_samples_v7/mnistCUDNN make clean && make # Run the test ./mnistCUDNN