Jetson Monitor

View on GitHub

Jetson Monitor is a real-time web-based monitoring dashboard specifically designed for NVIDIA Jetson devices. This Flask-based application provides comprehensive system monitoring with a focus on the unique capabilities and thermal characteristics of Jetson hardware.

The project emerged from the need to monitor Jetson devices in real-time, particularly their thermal behavior and GPU utilization, which are critical for maintaining optimal performance in edge computing applications.

dashboard

Features

System Monitoring

  • CPU & Memory: Real-time CPU usage, memory consumption, and swap utilization
  • Storage: Disk space monitoring with visual indicators
  • Network: Live network throughput tracking (sent/received data rates)
  • Uptime: System uptime tracking and display

GPU Monitoring

  • GPU Utilization: Real-time GPU usage percentages
  • GPU Temperature: Continuous thermal monitoring
  • Power Consumption: GPU and total system power draw tracking
  • Performance Metrics: Comprehensive GPU performance data via NVIDIA ML

Thermal Management

  • Thermal Status: Real-time thermal state monitoring
  • Throttling Detection: CPU and GPU throttling alerts
  • Visual Indicators: Color-coded status system (Green/Yellow/Red)
  • Event Reporting: Detailed thermal event logging

Memory Pressure Analysis

  • Pressure Scoring: Intelligent memory pressure calculation (0-100 scale)
  • Weighted Algorithm:
    • Memory usage (60% weight)
    • Swap usage (30% weight)
    • Available memory (10% weight)
  • Visual Alerts: Color-coded pressure indicators

Technical Implementation

Backend Architecture

The application is built with Flask 3.0.2 and leverages several key libraries:

  • psutil 5.9.8: System and process monitoring
  • nvidia-ml-py3 7.352.0: NVIDIA GPU monitoring via NVML
  • Real-time Updates: WebSocket-like updates via AJAX polling

Frontend Design

  • Modern Dark Theme: Responsive design optimized for monitoring scenarios
  • Interactive Charts: Real-time data visualization with historical tracking
  • Mobile-Friendly: Responsive layout that works on various screen sizes
  • Configurable Updates: Adjustable refresh intervals (1-60 seconds)

Data Visualization

  • Real-time Charts: Interactive graphs showing the last 60 data points
  • Historical Tracking: Persistent data visualization across sessions
  • Performance Metrics: Visual representation of all monitored parameters

Use Cases

This monitoring solution is particularly valuable for:

  • Edge Computing: Monitoring Jetson devices in remote or embedded applications
  • Development: Real-time performance analysis during application development
  • Thermal Analysis: Understanding thermal behavior under different workloads
  • System Administration: Remote monitoring of Jetson-based systems
  • Research: Performance analysis for AI/ML workloads on edge devices

Back