Tutorial

How to Setup a Windows Laptop for Data Science

powershell, WSL, Python, Cuda and more!

Benedict Neo

Published in

bitgrit Data Science Publication

5 min readJun 3, 2024

I’ve been a Mac user ever since I started my journey into data science and ML.

But as part of the Z by HP Global Ambassadors program, I was lucky enough to get my hands on powerful HP hardware, specifically the HP Z6 G5 Tower Workstation. Here’s the specs:

Processor: AMD Ryzen Threadripper PRO 7965WX 24-Cores 4.20 GHz
Installed RAM: 256 GB (255 GB usable)
System type: 64-bit operating system, x64-based processor
Graphics: NVIDIA RTX 6000 Ada 48 GB 4DP Graphics

I’ve never had compute power of this scale to train models so I’m super excited to get started.

To get everything setup, I did some research and reading and decided to write this article which outlines what I did to get my PC up and running for doing data science work.

Note: this is by no means the best way to setup your windows, it is just personal perference. Please leave your thoughts and suggestions in the comments!

Let’s dive in!

Step 1: Install PowerShell 7 Using Winget

First, we need to install PowerShell 7. This can be done easily using winget, the Windows Package Manager.

winget install --id Microsoft.Powershell --source winget
winget install --id Microsoft.Powershell.Preview --source winget

Step 2: Install WSL

Windows Subsystem for Linux (WSL) allows you to run a Linux distribution alongside your Windows installation. Here’s how to set it up:

1. Enable the “Virtual Machine Platform” feature on Windows:

Go to “Turn Windows features on or off”.
Check the checkbox near “Virtual Machine Platform”.

2. Install WSL:

wsl --install

3. Set WSL2 version

wsl --set-default-version 2

4. Check the WSL version:

wsl -l -v

Install Ubuntu

1. Install Ubuntu:

wsl - install -d Ubuntu

2. Run Ubuntu:

wsl - distribution Ubuntu

3. Shut down Ubuntu:

wsl -t Ubuntu

4. Check OS status:

wsl - list - verbose

5. Exit WSL:

exit

In the Terminal app, set Ubuntu as default.

upgrade packages

sudo apt update && sudo apt upgrade

Find more linux commands

Step 4: Install ZSH and Oh My Zsh

ZSH offers smarter tab auto-completion, aliases, and other features that make it a great shell for development.

1. Install ZSH:

sudo apt-get install zsh

2. Install Oh My Zsh:

sh -c "$(curl -fsSL https://raw.github.com/ohmyzsh/ohmyzsh/master/tools/install.sh)"

Find themes and plugins to customize your ZSH setup.

Step 5: Setup Git

1. Generate a new SSH key and add it to the ssh-agent

ssh-keygen -t ed25519 -C "your_email@example.com"

2. Add a new SSH key to your GitHub account

3. Configure your username and email

git config --global user.name "Your Name"
git config --global user.email "your_email@example.com"

Step 6: Setup CUDA

CUDA allows you to leverage the power of NVIDIA GPUs for computational tasks.

Install the GCC compiler:

sudo apt install gcc --fix-missing

Install Nvidia CUDA toolkit:

sudo apt install nvidia-cuda-toolkit

check the Driver and CUDA versions:

nvidia-smi
nvcc -V

Step 7: Install Python with Mamba

Mamba is a package manager that’s a drop-in replacement for conda and is generally faster at resolving dependencies

Miniforge distribution >= Miniforge3-22.3.1-0

wget https://github.com/conda-forge/miniforge/releases/latest/download/Miniforge3-Linux-x86_64.sh
bash Miniforge3-Linux-x86_64.sh

Example usage: Create a new environment and install JupyterLab:

mamba create -n myjlabenv jupyterlab -c conda-forge
mamba activate myjlabenv  # activate our environment
jupyter lab               # this will start up jupyter lab and open a browser