Data Science Stack on Ubuntu
Set up ML environments with ease on your AI workstation using an out-of-the-box solution for data science.
Try the Beta Learn more about data science tools on Ubuntu ›
Get started with data science on your workstation or public cloud
![Google Cloud](https://res.cloudinary.com/canonical/image/fetch/f_auto,q_auto,fl_sanitize,w_382,h_312/https://assets.ubuntu.com/v1/928e43c9-google-cloud-logo.png)
![Lenovo](https://res.cloudinary.com/canonical/image/fetch/f_auto,q_auto,fl_sanitize,w_203,h_312/https://assets.ubuntu.com/v1/daf5ad79-lenovo-logo.png)
![Microsoft Azure](https://res.cloudinary.com/canonical/image/fetch/f_auto,q_auto,fl_sanitize,w_95,h_312/https://assets.ubuntu.com/v1/6bdb9d14-microsoft-azure-2021.png)
![AWS](https://res.cloudinary.com/canonical/image/fetch/f_auto,q_auto,fl_sanitize,w_152,h_312/https://assets.ubuntu.com/v1/18492406-aws-logo.png)
![HP](https://res.cloudinary.com/canonical/image/fetch/f_auto,q_auto,fl_sanitize,w_134,h_312/https://assets.ubuntu.com/v1/34dacc75-hp-logo.png)
![Nvidia](https://res.cloudinary.com/canonical/image/fetch/f_auto,q_auto,fl_sanitize,w_310,h_312/https://assets.ubuntu.com/v1/5fdeba6f-nvidia-logo.png)
Why choose Ubuntu for Data Science?
- Get started on your workstation to develop models. Scale as you upskill and deploy in production when needed using an MLOps platform.
- Benefit from long-term support (LTS), which is released every 2 years, with 5 years of standard support extended up to 12 years with an Ubuntu Pro Desktop subscription.
- Access secure and supporting data science and ML packages such as Python, Tensorflow, PyTorch or MLflow.
- Ubuntu is the target platform for NVIDIA AI Workbench and Canonical Data Science Stack. It enables accelerated data science workloads to run locally from multiple GPU silicon vendors, including NVIDIA or Intel.
Get leading open source ML tools seamlessly integrated
![MLFlow](https://res.cloudinary.com/canonical/image/fetch/f_auto,q_auto,fl_sanitize,w_222,h_312/https://assets.ubuntu.com/v1/8aa72b35-mlflow-logo.png)
What is Data Science Stack?
Get started with data science using a few commands.
- Get an ML environment ready within minutes on any Linux distribution
- Streamline the complexity of GPU configuration and quickly attach it run containerised workloads
- Managed multiple machine-learning environments with an intuitive CLI and UI
- Access leading open source ML tooling such as Jupyter Notebook or MLflow
What's inside Data Science Stack?
Data science stack includes tools that will help you get started easily:
- JupyterLab for ETL, model training and experimentation
- MLFlow for experiment tracking and model registry
- ML frameworks by default, include PyTorch or TensorFlow
- GPU support for different types and easy enablement
Fully configure your chosen stack to your specific needs.
Why choose
Data Science Stack?
Improve developer productivity
Easy to use on any AI workstation
Run your ML workloads in a secure environment
Begin your AI journey on Ubuntu
One vendor to support your AI stack
Scale your AI workloads with an MLOps platform
Machine learning operations (MLOps) is a practice that enables data scientists and ML engineers to develop and deploy models in a reproducible and repeatable manner.
Charmed Kubeflow is an MLOps platform that covers the entire ML lifecycle. It is a cloud-native application that runs anywhere, whether in a private or public cloud, supporting even hybrid or multi-cloud scenarios.
Download the MLOps guide What is Kubeflow?
Open source AI resources
Learn how to select your data science tools and quickly get your environment ready on Ubuntu.
Upgrade your data science workflows with Ubuntu WSL
Learn how to upgrade your data science workflows with WSL.
Kubeflow vs MLFlow
Choosing a suitable machine learning tool can often be challenging. Understand the differences between the most famous open source solutions.
An overview of machine learning security risks
Be on top of the security risks for machine learning projects and learn how to mitigate them.
Guide to Managed AI Infrastructure
Discover how managed AI infrastructure can accelerate deployments and increase security in our latest executive guide.