Open-source web-based GPU performance visualization tool that transforms NVIDIA profiling data into interactive insights for CUDA engineers.

Part of the RightNow AI ecosystem. Want 180x more powerful AI assistance for your entire codebase? Try RightNow AI Code Editor.

# Clone and install
git clone https://github.com/RightNow-AI/gpu-profiler.git
cd gpu-profiler
npm install

# Start development server
npm run dev

# Open http://localhost:3000 and drop your .nvprof files!

Try GPU Profiler - Upload your profiling files instantly

• Interactive Visualizations - Timeline, flame graphs, and heatmaps powered by D3.js
• Multiple Format Support - .nvprof, .nsys-rep, and .json profiling files
• Real-time Analysis - Client-side processing for instant results
• Bottleneck Detection - AI-powered hints for performance optimization
• Shareable Results - Generate links to share profiling insights
• Export Capabilities - JSON, CSV, and Chrome Tracing format export
• Professional UI - Clean, modern interface with JetBrains typography

graph LR
    A[Upload .nvprof/.nsys] --> B[Parse Profile Data]
    B --> C[Generate Visualizations]
    C --> D[Timeline View]
    C --> E[Flame Graph]
    C --> F[Heatmap]
    D --> G[Bottleneck Analysis]
    E --> G
    F --> G
    G --> H[Export & Share]
    
    style A fill:#e1f5fe
    style H fill:#c8e6c9
    style G fill:#fff3e0

1. Upload - Drop your NVIDIA profiling files (.nvprof, .nsys-rep, .json)
2. Parse - Client-side processing extracts kernel launches and GPU metrics
3. Visualize - Interactive timeline, flame graphs, and utilization heatmaps
4. Analyze - AI-powered bottleneck detection and optimization hints
5. Share - Export results or generate shareable links

• Node.js 18+
• NPM or Yarn
• Modern web browser (Chrome, Firefox, Safari, Edge)

# Clone repository
git clone https://github.com/RightNow-AI/gpu-profiler.git
cd gpu-profiler

# Install dependencies
npm install

# Start development server
npm run dev

# Build for production
npm run build
npm start

# Verify installation at http://localhost:3000

app/
├── components/           # React components
│   ├── visualizations/  # D3.js visualization components
│   ├── file-upload.tsx  # Drag & drop upload
│   └── profiler-interface.tsx
├── lib/
│   ├── parsers/         # File format parsers
│   ├── utils.ts         # Utility functions
│   ├── sharing.ts       # URL sharing logic
│   └── export.ts        # Export functionality
├── store/               # Zustand state management
├── types/               # TypeScript definitions
└── globals.css          # Tailwind + custom styles

• Next.js 15 - React framework with App Router
• TypeScript - Type safety and developer experience
• D3.js - Data-driven visualizations
• Tailwind CSS - Utility-first styling
• Zustand - Lightweight state management
• Lucide React - Clean, consistent icons

• Client-side processing - No data leaves your browser
• Efficient parsing - Handles large profiling files (up to 100MB)
• Optimized renders - Smooth interactions even with thousands of kernels
• Memory efficient - Smart data compression for sharing

• 100% client-side - Your profiling data never leaves your device
• No tracking - We don't collect any analytics or personal data
• Secure sharing - Shared links use URL-safe encoding
• Open source - Full transparency, audit the code yourself

We welcome contributions! Please see our Contributing Guide for details.

1. Fork the repository
2. Create a feature branch: git checkout -b feature/amazing-feature
3. Make your changes
4. Run tests: npm test
5. Commit: git commit -m 'Add amazing feature'
6. Push: git push origin feature/amazing-feature
7. Open a Pull Request

Found a bug or have a feature request? Open an issue with:

• Clear description of the problem
• Steps to reproduce
• Expected vs actual behavior
• Sample profiling file (if applicable)

• Live profiling - Connect to ncu/nsys running locally
• WebGPU support - Chrome WebGPU trace compatibility
• Multi-file comparison - Compare multiple profiling sessions
• Advanced filters - Filter kernels by occupancy, duration, etc.

• AI optimization - Integration with RightNow AI's CUDA analyzer
• Code mapping - Upload kernel source code for line-by-line analysis
• Collaborative features - Team workspaces and annotations
• API access - Programmatic access for CI/CD integration

Need advanced features for your team? RightNow AI offers:

• Live inline profiling in the AI-powered code editor
• Custom optimization rules for your specific kernels
• Advanced AI analysis with natural language explanations
• Team collaboration features and workspace management

GPU Profiler provides comprehensive performance insights:

We welcome contributions! See CONTRIBUTING.md for guidelines.

MIT License - free for personal, educational, and commercial use.

• Website: www.rightnowai.co
• GPU Profiler: profiler.rightnowai.co
• Documentation: docs.rightnowai.co
• Discord: Join our community
• Twitter/X: @rightnowai_co

Made with love by the RightNow AI Team