This is my first post for my daily notes about my machine learning activity. I plan to put a machine learning system inside my laptop. My first objective is to put CUDA engine into this laptop because it already has NVIDIA Graphic Card which is 960 series. For the reference, I would like to use an article from pysearch image blog that in my perspective it is very clear and really specific.
Installing CUDA dependency for Ubuntu.
we need to make sure that our system is up to date
$ sudo apt-get update
$ sudo apt-get upgrade
And then we have to install all dependency based on this blog
$ sudo apt-get install build-essential cmake git unzip pkg-config
$ sudo apt-get install libjpeg-dev libtiff5-dev libjasper-dev libpng12-0
$ sudo apt-get install libavcodec-dev libavformat-dev libswscale-dev libv4l-dev
$ sudo apt-get install libxvidcore-dev libx264-dev
$ sudo apt-get install libgtk-3-dev
$ sudo apt-get install libhdf5-serial-dev graphviz
$ sudo apt-get install libopenblas-dev libatlas-base-dev gfortran
$ sudo apt-get install python-tk python3-tk python-imaging-tk
We will use python for coding in the future, so we will have to install this
$ sudo apt-get install python2.7-dev python3-dev
This step is to prepare when we want to switch to NVIDIA-CUDA for default driver
$ sudo apt-get install linux-image-generic linux-image-extra-virtual
$ sudo apt-get install linux-source linux-headers-generic
INSTALL CUDA TOOLKIT
Let us check up first for the CUDA compatibility on this
wikipedia page. My GPU is 960M which is using maxwell micro-architecture that supported by CUDA 9 SDK. So I decide to download CUDA SDK from
NVIDIA web pages .
For the first CUDA 9.2 is using Nvidia 396 driver, so we need to install this driver and activate before we continue to install Cuda 9.2
For ubuntu 18.04, you cannot install nvidia 396 using command line, you need to set that using Software and Update GUI
Please use the local runfile. In my case when we are using the local deb, it would be automatically asked your system to install and switch your device driver into nvidia 396.37. But unfortunately, this driver is only available for Tesla GPU not for G-Force.
Next Follow this instruction
sudo sh cuda_9.2.88_396.26_linux.run
After executing the above command, you'll see the EULA ( you can skip with Ctrl-C )
You are attempting to install on an unsupported configuration. Do you wish to continue?
(y)es/(n)o [ default is no ]: y
Install NVIDIA Accelerated Graphics Driver for Linux-x86_64 396.37?
(y)es/(n)o/(q)uit: n
Install the CUDA 9.2 Toolkit?
(y)es/(n)o/(q)uit: y
Enter Toolkit Location
[ default is /usr/local/cuda-9.2 ]:
Do you want to install a symbolic link at /usr/local/cuda?
(y)es/(n)o/(q)uit: y
Install the CUDA 9.2 Samples?
(y)es/(n)o/(q)uit: y
Enter CUDA Samples Location
[ default is /home/kinghorn ]: /usr/local/cuda-9.2
Also don't forget to install the update patch
Komentar