"Information should not be displayed all at once; let people gradually become familiar with it." - Edward Tufte
Transform GPU monitoring from complex metrics into intuitive visual patterns. Enterprise-grade NVIDIA GPU monitoring with real-time analytics, intelligent alerts, and historical analysis.
Real-time GPU metrics visualized for instant comprehension
Do you find yourself:
- Parsing dense terminal output when you should be focusing on your work?
- Missing critical temperature spikes or memory leaks?
- Lacking historical data for performance analysis?
- Needing better alerting for GPU health?
- Managing multiple GPUs across different workloads?
- Real-time visual dashboard
- Color-coded temperature ranges
- Resource utilization patterns
- Multi-GPU support
- Dark/light mode
- Configurable thresholds
- Temperature spike detection
- Resource bottleneck warnings
- Email/webhook notifications
- Historical performance data
- Time-series analysis
- Usage pattern recognition
- Power efficiency metrics
- RESTful API
- Supabase time-series storage
- Zero-config deployment
- Multi-user access
Transform dense metrics into intuitive patterns
Clear resource utilization during ML model training
Immediate visual alerts during intensive workloads
Intuitive color-coding for instant recognition:
- 🔴 ≥85°C: Critical
- 🟠 75-84°C: Warning
- 🟡 65-74°C: Normal
- 🟢 50-64°C: Optimal
- 🔵 <50°C: Cool
# Clone repository
git clone https://github.com/jackccrawford/gpu-sentinel-pro.git
# Start services
./backend/src/service/run_service.sh
./frontend/run_frontend.sh
# Access dashboard
http://localhost:3055See Installation Guide for detailed setup instructions.
- Requirements & User Stories
- Technical Architecture
- API Reference
- Installation Guide
- Contributing Guide
- Security Policy
MIT License - Free for personal and commercial use.
Built for Orphaned NVIDIAs