A JupyterLab extension for visualizing GPU usage dashboards directly within the JupyterLab interface.
JupyterLab NVdashboard is a JupyterLab extension that provides real-time dashboards for monitoring GPU usage and system hardware metrics. It visualizes key performance indicators like GPU compute utilization, memory consumption, and throughput directly within JupyterLab, helping users optimize and debug GPU-intensive workloads.
Data scientists, machine learning engineers, and developers working with GPU-accelerated computing in JupyterLab who need to monitor resource utilization during model training, data processing, or scientific simulations.
It integrates GPU monitoring seamlessly into the JupyterLab interface, eliminating the need for external tools and providing interactive, theme-aware visualizations with features like time-series brushing and synced tooltips for detailed analysis.
A JupyterLab extension for displaying dashboards of GPU usage.
Open-Awesome is built by the community, for the community. Submit a project, suggest an awesome list, or help improve the catalog on GitHub.
Displays live charts for GPU compute utilization, memory consumption, PCIe, and NVLink throughput directly in JupyterLab, enabling immediate performance assessment during GPU-intensive workflows.
Features brushing for selecting time ranges and synced tooltips across charts, allowing detailed exploration and correlation of metrics, as shown in the demo GIFs and screenshots.
Automatically adapts to JupyterLab's light or dark themes, ensuring a cohesive visual experience without manual adjustments, as demonstrated in the theme compatibility images.
Installs as a full extension with both frontend and server components for JupyterLab v4, providing native dashboard access within the development environment with minimal setup.
Only compatible with JupyterLab v4 and later, forcing upgrades for users on older versions and requiring maintenance of a separate branch for v3 support, as noted in the version compatibility section.
Defers adding cell execution markers to charts due to asynchronous complexities, limiting the ability to correlate GPU usage with specific notebook cells for enhanced debugging, as mentioned in future improvements.
Relies on pynvml for GPU metrics, making it unsuitable for systems with AMD, Intel, or other non-NVIDIA GPUs, and it lacks support for broader hardware monitoring beyond GPU-focused metrics.