v2.0.0 - High-Performance Modular Engine: 600x Speedup

🚀 v2.0.0: The Production Engine

This release marks the transition from a Python research prototype to a production-grade High-Performance Computing (HPC) library.

💎 Engineering Highlights

• Modular C++17 Architecture: Implemented full separation between IsingLattice (State Management) and MetropolisEngine (Physics Logic).
• Data-Oriented Design (DOD): Replaced slow modulo-based neighbor lookups with a Precomputed Neighbor Table, transforming $O(N)$ boundary calculations into $O(1)$ memory reads.
• Boltzmann Optimization: Integrated a Lookup Table (LUT) for acceptance probabilities, eliminating expensive std::exp calls in the simulation hot-path.
• Hardware-Aware Memory: Unified contiguous int8_t storage ensures maximum L1/L2 cache residency.

📈 Benchmarks (L=200, 40k sites)

• Throughput: ~0.0126 GigaFlips per second (GF/s).
• Latency: 79.25 ns per spin-flip attempt, including real-time magnetization measurements.
• Speedup: ~630x faster than the pure Python baseline.

🧪 Physics Validation

• Confirmed accuracy via Spontaneous Symmetry Breaking analysis ($\langle|m|\rangle \approx 0.61$ at $T_c$).
• Validated scientific protocol by separating Thermalization (Warm-up) from the Production phase.

---

This engine serves as the high-speed backend for the upcoming Phase 3.0: Econophysics and Market Sentiment Analysis.