GPU Profiler
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.
Quick Start
# 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!
Live Demo
Try GPU Profiler - Upload your profiling files instantly
Demo
Features
- 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
How It Works
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
- Upload - Drop your NVIDIA profiling files (.nvprof, .nsys-rep, .json)
- Parse - Client-side processing extracts kernel launches and GPU metrics
- Visualize - Interactive timeline, flame graphs, and utilization heatmaps
- Analyze - AI-powered bottleneck detection and optimization hints
- Share - Export results or generate shareable links
Installation
Requirements
- Node.js 18+
- NPM or Yarn
- Modern web browser (Chrome, Firefox, Safari, Edge)
Setup
# 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
Architecture
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
Key Technologies
- 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
Performance
- 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
Privacy & Security
- 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
Contributing
We welcome contributions! Please see our Contributing Guide for details.
Development Workflow
- Fork the repository
- Create a feature branch:
git checkout -b feature/amazing-feature
- Make your changes
- Run tests:
npm test
- Commit:
git commit -m 'Add amazing feature'
- Push:
git push origin feature/amazing-feature
- Open a Pull Request
Reporting Issues
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)
Roadmap
v1.1 (Next Release)
- 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.
v1.2 (Future)
- 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
Enterprise
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
Performance Features
GPU Profiler provides comprehensive performance insights:
| Category |
Features |
| Kernel Analysis |
Execution time, occupancy, register usage, memory throughput |
| Memory Profiling |
Transfer patterns, bandwidth utilization, cache efficiency |
| Bottleneck Detection |
AI-powered optimization suggestions and performance hints |
| Visual Analytics |
Interactive charts, zoom capabilities, data export |
Contributing
We welcome contributions! See CONTRIBUTING.md for guidelines.
License
MIT License - free for personal, educational, and commercial use.
Links
Made with love by the RightNow AI Team