Ubuntu Environment

New Ubuntu 18.04 install on x86_64 workstation

Install Ubuntu 18.04 Alongside With Windows 10 in Dual Boot

Download Ubuntu 18.04 ISO from https://www.ubuntu.com/#download
Follow the procedure on https://www.itzgeek.com/how-tos/linux/ubuntu-how-tos/how-to-install-ubuntu-18-04-alongside-with-windows-10-or-8-in-dual-boot.html

New Ubuntu 18.04 install on x86_64 MSI GS65 laptop

Install Ubuntu 18.04 Alongside With Windows 10 in Dual Boot

Use ubuntu 18.04 boot USB created from workstation install
Follow the procedure on https://askubuntu.com/questions/1063910/installing-ubuntu-on-a-msi-gs65-freezes-at-login-screen

Installing system libraries and tools

Update Ubuntu


$ sudo apt-get update
$ sudo apt-get full-upgrade
$ sudo ubuntu-drivers autoinstall
# Disable auto resize of windows at top
$ gsettings set org.gnome.mutter edge-tiling false

Install the NVIDIA 440 driver

Get the latest supported drivers from the Ubuntu repository


$ sudo add-apt-repository ppa:graphics-drivers/ppa
$ sudo apt update

Open Softare & Updates program from the application menu. Click the Additional Drivers tab. You can see what driver is being used for NVIDIA and a list of proprietary drivers.

As you can see nvidia-driver-440 is available for the GeForce GTX 1080 Ti card. Select nvidia-driver-440 and click Apply Changes button to install the driver. Reboot the system to apply the changes.

After reboot, use the follow command to check the driver version


$ nvidia-smi

Check CPU temp with Psensor GUI tool

Install the needed tools


$ sudo apt install lm-sensors hddtemp

Start the detection of your hardware sensors:


$ sudo sensors-detect

Check the operations


$ sensors

Install Psensor


$ sudo apt install psensor

Development Tools

CLion IDE - ubuntu software app
Google Chrome Web Browser
Go to https://www.google.com/chrome. Click the Download Chrome button. Select the first option (64 bit .deb for Debian/Ubuntu), click Accept and Install to download the installer google-chrome-stable_current_amd64.deb. Double click on the install file to install.
CMake - ubuntu software app
KDiff3 (File And Directory Comparison And Merge Tool) - ubuntu software app

Doxygen


$ sudo apt-get install doxygen
$ sudo apt-get install doxygen-gui
$ sudo $HOME/Projects/pdp/scripts/add-doxygenMenuItem.sh

Typora


$ sudo apt-key adv --keyserver keyserver.ubuntu.com --recv-keys BA300B7755AFCFAE
$ sudo add-apt-repository 'deb https://typora.io/linux ./'
$ sudo apt-get update
$ sudo apt-get install typora

Subversion - command line (svn)


$ sudo apt-get install subversion

SmartSVN

Download the latest version smartsvn-linux-11_0_0.tar.gz from https://www.smartsvn.com/downloads/smartsvn/smartsvn-linux-11_0_0.tar.gz
unpack to $HOME/Developer/smartsvn


 $ cd $HOME/Developer/smartsvn/bin
 $ sudo ./add-menuitem.sh

7Zip


$ sudo apt install p7zip-full p7zip-rar

Build essential tools


   $ sudo apt-get install build-essential g++

libjpeg


$ sudo apt-get install libjpeg-dev

libturbojpeg


$ sudo apt-get install libturbojpeg-dev

libz


$ sudo apt-get install libz-dev

liblzma


$ sudo apt-get install liblzma-dev

EGL


$ sudo apt-get install libegl1-mesa
$ sudo apt-get install libegl1-mesa-dev

GLES


$ sudo apt-get install libgles2-mesa
$ sudo apt-get install libgles2-mesa-dev
$ sudo apt-get install freeglut3-dev

X11 development


sudo apt-get install xorg-dev

Cleanup


$ sudo apt-get update
$ sudo apt autoremove

Nvidia Libraries

CUDA

Download the CUDA 10.2 Ubuntu installer


$ cd ~/Developer/Projects
$ mkdir installers
$ cd installers
$ wget http://developer.download.nvidia.com/compute/cuda/10.2/Prod/local_installers/cuda_10.2.89_440.33.01_linux.run

Install CUDA 10.2 using the installer


x
$ sudo sh cuda_10.2.89_440.33.01_linux.run
Choose "Continue" when prompted about the existing package manager
Type "accept" when prompted for acceptance of the EULA
Turn off the Driver install on the following options screen (440 driver installed above)
CUDA Installer
- [ ] Driver
     [ ] 440.33.01
+ [X] CUDA Toolkit 10.2
  [X] CUDA Samples 10.2
  [X] CUDA Demo 10.2
  [X] CUDA Documentation 10.2
Then choose "Install"

Add CUDA 10 to the development search path by adding the following lines to .bashrc


$ nano .bashrc
  # NVIDIA CUDA Toolkit
  if [ -d "/usr/local/cuda-10.2/bin/" ]; then
    export PATH=/usr/local/cuda-10.2/bin${PATH:+:${PATH}}
    export LD_LIBRARY_PATH=/usr/local/cuda-10.2/lib64${LD_LIBRARY_PATH:+:${LD_LIBRARY_PATH}}
  fi
$ source ~/.bashrc

Check NVIDIA CUDA Compiler
```
$ nvcc --version
```

CuDNN

Download the cuDNN pre-built libraries for Linux, CUDA version 10.2 cudnn-10.2-linux-x64-v7.6.5.32.tgz from https://developer.nvidia.com/rdp/cudnn-download using the cuDNN Library for Linux link. Move cudnn-10.2-linux-x64-v7.6.5.32.tgz to $HOME/Developer/Projects/installers

Add the header files and pre-built libraries to the CUDA install


$ cd $HOME/Developer/Projects/installers
$ tar -xvf cudnn-10.2-linux-x64-v7.6.5.32.tgz
$ cd cuda
$ sudo cp -P include/* /usr/local/cuda/include/
$ sudo cp -P lib64/* /usr/local/cuda/lib64/
$ cd ..
$ sudo rm -rf cuda

Video Codec

Download the Video Codec SDK 9.1 pre-built libraries for Linux, Video_Codec_SDK_9.1.23.zip from https://developer.nvidia.com/nvidia-video-codec-sdk/download to $HOME/Developer/Projects/installers and unzip to Video_Codec_SDK_9.1.23

Add the header files and pre-built libraries to the CUDA install


$ cd $HOME/Developer/Projects/installers
$ sudo cp include/* /usr/local/cuda/include/
$ sudo cp Lib/linux/stubs/x86_64/* /usr/local/cuda/lib64/

OptiX - Ray Tracing Engine

Download the OptiX SDK 7.0 pre-built libraries for Linux, NVIDIA-OptiX-SDK-7.0.0-linux64.sh from https://developer.nvidia.com/designworks/optix/download to $HOME/Developer/Projects

Install OptiX SDK 7.0 using the installer


$ cd $HOME/Developer/Projects
$ sh NVIDIA-OptiX-SDK-7.0.0-linux64.sh
$ mv NVIDIA-OptiX-SDK-7.0.0-linux64.sh installers/.
$ mv NVIDIA-OptiX-SDK-7.0.0-linux64 NVIDIA-OptiX-7.0.0

Install Third Party Libraries

Download the following third party libraries for Ubuntu 18.04 from the BAP development website at Third party libraries downloads (Will need your username and password to access the site)


x
boost 1.73.0
glfw 3.3.2
libpng 1.6.37
itk 5.1.0
pdp 2.6.4
rnifti 1.1.0
sqlite 3.32.1

Move the downloaded files to $HOME/Developer/Projects directory and and extract with the 7-Zip Extract Here command from the file browser.

Qt 5.15.0

Download the latest Qt unified windows installer from https://www.qt.io/download-qt-installer
Run the unified installer using open source user login. You will need to register as a QT Open Source user at the Qt website.
Select $HOME/Developer/Projects/QT as the install directory
Select Qt 5.15.0 packages Desktop gcc 64-bit and QtCharts and run install

Create symbolic link for install directory


$ cd ~/Developer/Projects
$ ln -s ~/Developer/Projects/QT/5.15.0/gcc_64 qt-5.15.0

Install Datasets

Download the following Datasets for the BAP development website at Data sets (Will need your username and password to access the site)


BRATS PBT voxel classification
MNIST hand written numbers

Move the downloaded files to $HOME/data/ImageData directory and and extract with the 7-Zip Extract Here command from the file browser.

Install Neural Network models

Download the following models for the BAP development website at Models (Will need your username and password to access the site)


BRATS PBT voxel classification
MNIST hand written numbers

Move the downloaded files to $HOME/networks directory and and extract with the 7-Zip Extract Here command from the file browser.

CLion build

Checkout the source

Use either the SmartSVN or TortoiseSVN svn clients to checkout the BAP source code from svn://3.14.90.147/bap/trunk to `$HOME/Projects/bap (Will need your username and password to access this repository).

Create the CLion project

Create directory `$HOME/Projects/bap-clion
...