krassowski / jupyterlab-nvdashboard

A JupyterLab extension for displaying dashboards of GPU usage.

JupyterLab NVdashboard

NVDashboard is a JupyterLab extension for displaying GPU usage dashboards. It enables JupyterLab users to visualize system hardware metrics within the same interactive environment they use for development. Supported metrics include:

• GPU-compute utilization
• GPU-memory consumption
• PCIe throughput
• NVLink throughput

Demo

Table of Contents

• New Features
  • Brush for Time Series Charts
  • Synced Tooltips
  • Theme Compatibility
• Version Compatibility
• Requirements
• Installation
• Troubleshoot
• Contributing
• Future Improvements

New Features

JupyterLab-nvdashboard v4 brings a host of new features, improved backend architecture, and enhanced frontend components for an even better user experience. Explore the exciting updates below.

Brush for Time Series Charts

Introducing a powerful brushing feature for time series charts. Users can easily inspect past events by selecting a specific time range, providing more granular control over data exploration.

Synced Tooltips

For pages with multiple charts, JupyterLab-nvdashboard now offers synchronized tooltips for timestamps across all charts. This feature enhances the user's ability to analyze data cohesively and understand relationships between different data points.

Theme Compatibility

Seamless integration with JupyterLab themes is now a reality. The extension adapts its colors and aesthetics based on whether the user is in a light or dark theme, ensuring a consistent and visually appealing experience.

Light Theme

Dark Theme

Version Compatibility

JupyterLab-nvdashboard v4 is designed exclusively for JupyterLab v4 and later versions. To ensure continued support for JupyterLab v3 users, we will maintain the previous version separately (branch-0.9).

Requirements

• JupyterLab >=4
• pynvml
• psutil

Installation

Conda

# nightly version (for jupyterlab>=4)
conda install -c rapidsai-nightly -c conda-forge jupyterlab-nvdashboard

# stable version (for jupyterlab<4)
conda install -c rapidsai -c conda-forge jupyterlab-nvdashboard

PyPI

# nightly version (for jupyterlab>=4)
pip install --extra-index-url https://pypi.anaconda.org/rapidsai-wheels-nightly/simple --pre jupyterlab_nvdashboard

# stable version (for jupyterlab<4)
pip install jupyterlab_nvdashboard

Troubleshoot

If you are seeing the frontend extension, but it is not working, check that the server extension is enabled:

jupyter server extension list

If the server extension is installed and enabled, but you are not seeing the frontend extension, check the frontend extension is installed:

jupyter labextension list

Contributing Developers Guide

For more details, check out the contributing guide.

Future Improvements

While we've introduced a range of exciting features in this release, we understand that there are always opportunities for improvement. We have noted a request to add cell execution markers to the charts. Due to the complexities associated with asynchronous cells, we have decided to defer this feature to a future update. Rest assured, we will explore this enhancement in subsequent releases.