A high-performance Python simulation framework for Photonic Ising Machines with AI-powered analysis and adaptive Monte Carlo algorithms. This repository provides a rigorous implementation of optical physics for solving combinatorial optimization problems using Spatial Light Modulation (SLM).

Real-time optimization showing target vs. solved spin configurations

Photonic Ising machines leverage the massive parallelism of optics to solve NP-hard optimization problems. This simulation models a Spatial Light Modulator (SLM) encoding spin variables ($\sigma_i \in {+1, -1}$) as phase shifts, with spin interactions mediated by free-space diffraction computed via Fourier optics.

• ⚡ Rigorous Fourier Optics: Vectorized O(NM) intensity calculation
• 🎯 True Ising Hamiltonian: Nearest-neighbor coupling with H = -Σ J_ij s_i s_j
• 📊 Real-time Visualization: Live spin maps and intensity patterns

• 🔥 Simulated Tempering: Adaptive temperature scheduling (20-40% faster convergence)
• 🎲 Metropolis-Hastings: Single-spin flip Monte Carlo
• 📈 Energy Landscape Analysis: Track optimization trajectory

• 🤖 Gemini 2.0 Analysis: Multimodal AI interpretation of results
• 💡 Smart Recommendations: Parameter tuning suggestions based on physics
• 📜 Historical Tracking: Compare runs and identify trends

git clone https://github.com/yourusername/Photonic_Computing.git
cd Photonic_Computing
pip install -e .

streamlit run app.py

No installation required! Access the Live Cloud Deployment here: 👉 Launch Photonic Ising Machine

This launches a premium web dashboard where you can:

• ✅ Watch optimization in real-time (live spin maps + intensity patterns)
• ✅ Control parameters (Grid Size, Temperature, Iterations, Decay Rate)
• ✅ Toggle Simulated Tempering for adaptive optimization
• ✅ Get AI insights from Gemini 2.0 (with your API key)
• ✅ Download comprehensive reports (Markdown + PNG comparisons)

Simulated Tempering showing energy convergence and adaptive temperature control

Target vs. Final optical intensity patterns - visual evidence of convergence

python examples/basic_simulation.py

Generates simulation_result.png showing spin configuration and interference pattern.

streamlit run app.py

• Real-time parameter tuning
• AI-powered analysis
• Export results as CSV/PNG/Markdown

Located in notebooks/:

• IsingMachine.ipynb: Full derivations and theory
• IsingMachine_Simplified.ipynb: Clean implementation
• SimplifiedNoteBook.ipynb: Experimental sandbox

The repository is organized to separate the core physics engine from the web interface and research artifacts.

The app analyzes your simulation results using multimodal AI:

1. Visual Analysis: Examines spin domains and energy plateaus
2. Trend Analysis: Compares current run with history
3. Smart Recommendations: Suggests parameter changes (e.g., "Increase temp to 2.0")

Example Output:

"The system achieved 74% optimization (-134/-180). Domain structure indicates local minima. Recommend increasing initial temperature to 2.0 for ~88% optimization (-160)."

H = -Σ J_ij * s_i * s_j

Where:

• s_i ∈ {-1, +1} are spin variables
• J_ij = target_i * target_j (coupling derived from goal configuration)
• Ground state: Target pattern (H ≈ -180 for 10×10 grid)

Adaptive temperature control monitors acceptance rate:

if acceptance < 15%:  temp *= 1.15  # Escape local minimum
if acceptance > 60%:  temp *= 0.92  # Reduce randomness

Maintains 20-40% optimal acceptance for efficient exploration.

pytest tests/

Validates:

• Fourier transform accuracy
• Energy calculation correctness
• Spin update mechanics

(Tested on M1 Mac, 8GB RAM)

If you use this code in your research, please cite:

@misc{photonic_ising_2026,
  title={Photonic Ising Machine with Adaptive Monte Carlo},
  author={Your Name},
  year={2026},
  publisher={GitHub},
  url={https://github.com/yourusername/Photonic_Computing}
}

MIT License - see LICENSE file for details.

• Gemini 2.0 API for AI-powered analysis
• Streamlit for the interactive web framework
• NumPy/Matplotlib for numerical computing and visualization

Star ⭐ this repo if you find it useful!