Data scientists, data engineers, and AI developers wrestle with many operational tasks that can slow development and impact productivity, including software stack management, installation, and updates. Modern workflows are complex, including the need to reproduce state-of-the-art assets to accelerate development.

Master Your Data Science Environment for Workstations

NVIDIA Data Science Workbench simplifies and orchestrates tasks so you can manage your data science software environment on GPU-enabled workstations for greater productivity and ease-of-use. Finally, users have fast and convenient access to data science tools, key assets are just a click away. Workbench provides the following benefits:

  • Easier Manageability: Quick and easy setup. Manage NVIDIA Data Science Stack (DSS) software versions. Leverage tools that provide the best GPU optimized frameworks for productivity. Get access to drivers, CUDA, NVIDIA Container Toolkit, and other tools.

  • Greater Reproducibility: Easily containerize GitHub content to build quality models faster within your Jupyter environment. Use the Kaggle CLI to easily run a Kaggle container and kernel.

  • Enhanced Productivity: Benefit from easy software and driver installation and updates, as well as quick access to the Jupyter notebook, software assets, Kaggle notebooks, GitHub, and more. Easy access to NVIDIA GPU Cloud (NGC) containers for GPU-optimized code that also runs in AWS.

Easier Development Workflows

Building models begins with research, knowledge discovery, browsing Kaggle, or all of the above. You retrieve data, perform ETL operations, and then engineer features. The real challenge begins when you actually set up, build, train, and optimize your model.

This is where NVIDIA Data Science Workbench makes things easier. Do you need the current software stack and updated drivers for your GPU-accelerated workstation? Need to quickly grab NGC containers and GPU-optimized tools? Want to dockerize GitHub content for Jupyter in seconds? With Workbench, you can do all of this in more productive model building. Plus, Workbench can easily move your model to AWS or GPU-enabled servers.