Tetrahedron Optimization 3.0

Small project to test how many recursion levels of Sierpiński Tetrahedron, GPU can draw using OpenGL. This project is a personal experiment in real-time rendering and GPU optimization using OpenGL.

📚 Table of Contents

• 🚀 Features
• 🛠 Requirements
• 🔧 Build Instructions
• 📁 Project Structure
• 🧠 Rendering & Optimization Techniques
• 📊 Performance Benchmark
• 📜 License
• 👤 Author

🚀 Features

• 🔺 Recursive Sierpiński Tetrahedron (Recursion levels: 0–14)
• 🎨 Real-time background and object color customization
• 🌀 Interactive scene rotation (X / Y axes)
• 🖱️ Clean, minimal UI built with Dear ImGui
• ⚡ High-performance rendering using:
  • Instancing
  • SSBO (Shader Storage Buffer Objects)
  • Geometry Shader-based rendering
  • Recursive generation from previous tetrahedrons
  • GPU buffer mapping with memory barriers (GLsync)
  • Data is sent to shaders only when changed
• ✅ Fully compatible with OpenGL 4.3+

🛠 Requirements

• C++17 or newer
• OpenGL 4.3+
• Dependencies (via CPM.cmake or manual inclusion):
  • GLFW
  • GLAD
  • GLM
  • Dear ImGui
  • spdlog
  • freetype
  • stb_image

🔧 Build Instructions

Step-by-step

git clone https://github.com/Muppetsg2/OpenGL_Tetrahedron_Optimization.git
cd OpenGL_Tetrahedron_Optimization
cmake -B build -DCMAKE_BUILD_TYPE=Release
cmake --build build --parallel
./build/OpenGL_Tetrahedron_Optimization

NOTE for Windows 10/11 users:
Make sure Developer Mode is enabled in your system settings.
Go to: Settings > Privacy & security > For developers > Developer Mode

📁 Project Structure

├── cmake/                     # CMake utilities
│   ├── CPM.cmake              # CPM settings
│   └── global_settings.cmake  # CMake global settings
├── git_images/                # Images for Readme
├── res/                       # Resources
│   ├── shaders/
│   └── textures/
├── src/                       # Main application source
│   ├── CMakeLists.txt
│   ├── main.cpp
│   ├── pch.hpp
│   ├── Shader.cpp
│   ├── Shader.hpp
│   ├── Texture.cpp
│   └── Texture.hpp
├── thirdparty/                # External libraries
├── CMakeLists.txt
├── CMakeSettings.json
├── LICENSE
└── README.md

🧠 Rendering & Optimization Techniques

This project uses several GPU-friendly techniques to maintain real-time performance even at high recursion depths:

• Instancing – One draw call for thousands/millions of tetrahedrons.
• Shader Storage Buffer Objects (SSBO) – Store instance transformation data efficiently on the GPU.
• Geometry Shader – Dynamically constructs tetrahedron geometry from a single point per instance.
• Recursive Construction – New tetrahedrons are built from the output of the previous recursion level.
• Mapped Buffer Updates – Efficient CPU-to-GPU data transfer using glMapBufferRange.
• GLsync and Memory Barriers – Ensures safe access to GPU buffers across frames.
• Delta Data Uploading – Only update GPU buffers when instance data changes.

📊 Performance Benchmark

Frame rates measured at different recursion levels on multiple GPUs:

GPU	Recursion 8	Recursion 12	Recursion 13	Recursion 14
NVIDIA RTX 3060 Laptop	~1700 FPS	~76 FPS	~19 FPS	~3 FPS
NVIDIA RTX 4070 Ti SUPER	~1700 FPS	~130 FPS	~37 FPS	~9 FPS

⚠️ Performance is affected by resolution, drivers, system load, and whether debug mode is enabled.

Number of tetrahedrons generated at each recursion level (4^(r-1)):

Recursion Level	Number of Tetrahedrons
8	16 384
12	4 194 304
13	16 777 216
14	67 108 864

ℹ️ r — recursion level, where r = 0 means no object is rendered.

✨ Authors

👤 Marceli Antosik (Muppetsg2)

💡 If you have any questions or suggestions, feel free to contact me!

📜 License

📝 This project is MIT Licensed.

📖 MIT License Overview:

• ✅ Free to use, modify, and distribute.
• ✅ Can be used in commercial and non-commercial projects.
• ❗ Must include original license and copyright.

See the LICENSE file for more details.