NVIDIA Data Science Workbench
*IMPORTANT NOTE: A completely revised Workbench release is planned. To learn more, watch GTC Session S51283 – “A Path to Democratized Data Science in Any Cloud”.
What is NVIDIA Data Science Workbench?
- An unobtrusive desktop application to increase productivity for data scientists, data engineers, and AI developers
What does Workbench do?
- Simplifies and orchestrates data science tasks on GPU-enabled workstations
What are the key features in Workbench?
- Data science software installation and updates
- Single-click access to popular tools and CLIs
- Reproducing/containerizing code for experimentation
- Quick download of NGC containers for GPU-optimized code
- Curated RSS feed for relevant articles and blogs
Getting Started
Here are installation instructions for Workbench. If you need support along the way,
please reach out to NVIDIA via the email address provided at the end of the Installation section.
What you will need:
- NVIDIA GPU with 8GB or more of vRAM, such as NVIDIA RTX A6000, RTX A5000, RTX 8000, RTX6000, RTX5000
or GeForce 3090 Ampere GPUs. (16GB of vRAM recommended) - Ubuntu 18.04 or 20.04 OS, a terminal window and latest Chrome web browser (version 91.0 for example).
What to expect:
You will be running command line tools such as git, pip, nvidia-smi, ngc, kaggle, aws-cli in your favorite shell (usually bash). You will be asked to enter your sudo password during the initial install of Workbench during the pip3 installation. Workbench automatically clones the NVIDIA Data Science Stack (GitHub - NVIDIA/data-science-stack: NVIDIA Data Science stack tools) to facilitate NVIDIA Linux driver installation.
Your current NVIDIA driver will be upgraded to support the most up-to-date NGC Deep Learning framework (TensorFlow, PyTorch) containers. NVIDIA updates these containers once a month.
When Workbench updates, older versions of the containers will still be available in docker images reporting. After Workbench is installed and running, look for an NVIDIA icon in the upper right-hand side of the Ubuntu toolbar near the networking icon. To start Workbench, click on the NVIDIA icon and a pulldown menu of data science tasks will appear.
Installation steps:
Steps to successfully loading NVIDIA Data Science Workbench:
-
sudo apt install python3 pip (# may be required if pip3 is not already installed)
-
pip3 install nvdsw (# sudo passwd may be required)
-
nvdsw-setup
-
reboot
After a successful reboot, the NVIDIA Data Science Workbench logo will appear in the upper right-hand corner of Ubuntu desktop toolbar.
For support and feedback, reach out to us by email at NVDSWTechSupport@nvidia.com.