Quantum Collider Sandbox¶
Real-Time GPU Particle Physics Simulation with PDG-accurate particle catalog.
40 observable particles with real masses, lifetimes, decay channels, and proper relativistic kinematics. Built on Taichi for GPU acceleration (Vulkan/CUDA/CPU backends).
Getting Started
Usage & Features
Configuration
- Configuration Parameters Reference
- Configuration Guide
- Configuration Overview
- File Structure
- Common Configuration Scenarios
- Phase 1 Trail Optimization Guide
- Tuning for Different GPU Capabilities
- Force Configuration Examples
- Physics Accuracy vs. Performance
- Camera & Rendering Configuration
- Color & Appearance
- Particle Appearance
- Programmatic Configuration
- Preset-Based Configuration
- Testing Configuration Changes
- Configuration Checklist
- Performance Tuning & Optimization
- Phase 1: Trail Rendering Optimization
- Recommended Configurations
- Configuration Parameters
- Hardcoded Skip Conditions (Kernel-level)
- Other Optimization Opportunities
- Resolution & Rendering
- Physics Timestep Tuning
- Particle Rendering
- Measuring Performance
- Performance Testing Checklist
- Troubleshooting Low FPS
- Summary: Phase 1 Benefits
Development & Reference
Quick Start¶
git clone https://github.com/ml3m/quantum-collider-sandbox.git
cd quantum-collider-sandbox
python -m venv .venv && source .venv/bin/activate
make install && make run
On Windows, activate with .venv\Scripts\activate instead.
Key Features¶
- 🎯 Accurate Physics
40 PDG particles (leptons, bosons, Higgs, mesons, baryons)
Relativistic 2-5 body decays with Lorentz boosts
Quantum numbers (flavor, color, spin) tracking
Time dilation (special + general relativity)
- ⚡ GPU Acceleration
Real-time simulation at 1k-2k particles
Taichi backend (Vulkan/CUDA/CPU)
Phase 1 trail optimization: 10x vertex reduction
Optimized for modern GPUs
- 🎨 Advanced Visualization
Particle trails with motion filtering
Collision flashes and decay events
Black hole accretion disk & photon ring
Gravitational lensing effects
3D starfield background
- 🔧 Configurable Physics
10 simulation presets
Coulomb, Newtonian, strong, and magnetic forces
Force constants editable in real-time
Custom particle spawning
- 📊 Data I/O
HDF5 state export/import
JSONL physics event logging
CSV collision data
CMS/ATLAS format support
System Requirements¶
- Hardware
GPU with Vulkan support (NVIDIA, AMD, Intel)
2GB+ VRAM (4GB+ recommended for 1k+ particles)
Multi-core CPU (4+ cores)
- Software
Python 3.10 – 3.12 (3.12.2 recommended)
Vulkan drivers installed and working
- Tested Platforms
Linux (main development platform)
Windows 10/11
macOS 12+