HOW TO : Install Cuda and Cudnn with GTX 1080.

HOW TO : Install Cuda and Cudnn with GTX 1080.

1.INSTALL DRIVER FOR  GTX 1080

The first time that you install the card in the computer and connect the HDMI to it  the resolution will be the worst, so u need to install the drivers

sudo add-apt-repository ppa:graphics-drivers/ppa
sudo apt-get update
sudo apt install nvidia-367
sudo reboot

Go settigns and now , you can configure your resolution.

2.INSTALL CUDA.

Go to his link CUDA LINK .    in my case i download the version for UBUNTU 16.04 – runfile(local).

Go to the console and run :

sudo sh cuda_8.0.61_375.26_linux.run

Follow the instructions.

Remember that you already install the driver for nvidia so , u should write no.

 

3.INSTALL CUDNN

CUDNN is the library for neural network of nvidia, with this library you can train your own neural network with framework like caffe , tensor flow or darknet.

Register and Download CUDNN in the following link DOWNLOAD CUDNN

$ cd folder/extracted/contents
$ sudo cp -P include/cudnn.h /usr/include
$ sudo cp -P lib64/libcudnn* /usr/lib/x86_64-linux-gnu/
$ sudo chmod a+r /usr/lib/x86_64-linux-gnu/libcudnn*

Path environment variables

cd
sudo nano ./bashrc

paste the following lines in the end of .bashrc

export PATH=$PATH:/usr/local/cuda/bin
export LD_LIBRAY_PATH=$LD_LIBRARY_PATH:/usr/local/cuda/lib64
export CUDA_HOME=/usr/local/cuda
export LD_LIBRARY_PATH=/usr/local/cuda/lib
if [ -z $LD_LIBRARY_PATH ]; then
  LD_LIBRARY_PATH=/usr/local/cuda/lib64:/usr/local/cuda/lib
else
  LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/cuda/lib64:/usr/local/cuda/lib
fi
export LD_LIBRARY_PATH
export LD_LIBRARY_PATH=/usr/local/cuda8.0/lib64${LD_LIBRARY_PATH:+:${LD_LIBRARY_PATH}}

save and close the terminal.

Now check the nvcc

nvcc --version

nvcc: NVIDIA (R) Cuda compiler driver
Copyright (c) 2005-2016 NVIDIA Corporation
Built on Sun_Sep__4_22:14:01_CDT_2016
Cuda compilation tools, release 8.0, V8.0.44

Check the card

$ nvidia-smi     
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 375.51                 Driver Version: 375.51                    |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |
|===============================+======================+======================|
|   0  GeForce GTX 1080    Off  | 0000:01:00.0      On |                  N/A |
| 27%   34C    P8     7W / 180W |    448MiB /  8110MiB |      1%      Default |
+-------------------------------+----------------------+----------------------+
                                                                               
+-----------------------------------------------------------------------------+
| Processes:                                                       GPU Memory |
|  GPU       PID  Type  Process name                               Usage      |
|=============================================================================|
|    0       983    G   /usr/lib/xorg/Xorg                             338MiB |
|    0      1767    G   compiz                                         105MiB |
|    0      2311    G   /usr/lib/firefox/firefox                         2MiB |
+-----------------------------------------------------------------------------+

refence links

http://developer.download.nvidia.com/compute/cuda/7.5/Prod/docs/sidebar/CUDA_Installation_Guide_Linux.pdf

https://askubuntu.com/questions/799184/how-can-i-install-cuda-on-ubuntu-16-04

https://askubuntu.com/questions/767269/how-can-i-install-cudnn-on-ubuntu-16-04

, , , , ,
ZH-CN NL EN DE PT RU ES