aras-p/ToyPathTracer
aras-p/ToyPathTracer has 1,117 GitHub stars and 113 forks. GitHot measured 0 stars in 24 hours, -1 in 7 days, and -2 in 15 days. Last checked 2026-11-22 12:00 UTC. Last push 2025-04-12.
Crecimiento de estrellas medido
- in 24h
- 0
- in 7 days
- -1
- in 15 days
- -2
- in 1 month
- Recopilando historial
Cambios netos en períodos que se solapan, no totales diarios. Los datos no disponibles se dejan en blanco. Las estrellas indican interés, no calidad.
Acerca de
Toy path tracer for my own learning purposes (CPU/GPU, C++/C#, Win/Mac/Wasm, DX11/Metal, also Unity)
AprendizajeREADME.md
OriginalToy Path Tracer 
Toy path tracer I did in 2018 for my own learning purposes. Somewhat based on Peter Shirley's Ray Tracing in One Weekend minibook (highly recommended!), and on Kevin Beason's smallpt.

I decided to write blog posts about things I discover as I do this:
- Part 0: Intro
- Part 1: Initial C++ and walkthrough
- Part 2: Fix stupid performance issue
- Part 3: C#, Unity and Burst
- Part 4: Correctness fixes and Mitsuba
- Part 5: simple GPU version via Metal
- Part 6: simple GPU version via D3D11
- Part 7: initial C++ SIMD & SoA
- Part 8: SSE SIMD for HitSpheres
- Part 9: ryg optimizes my code
- Part 10: Update all implementations to match
- Part 11: Buffer-oriented approach on CPU
- Part 12: Buffer-oriented approach on GPU D3D11
- Part 13: GPU thread group data optimization
- Part 14: Make it run on iOS
- Part 15: A bunch of path tracing links
- Part 16: Unity C# Burst optimization
- Part 17: WebAssembly
Note: it can only do spheres, no bounding volume hierachy of any sorts, a lot of stuff hardcoded.
Performance numbers in Mray/s on a scene with ~50 spheres and two light sources, running on the CPU:
| Language | Approach | Ryzen 5950 | AMD TR1950 | MBP 2021 | MBP 2018 | MBA 2020 | iPhone 11 | iPhone X | iPhone SE |
|---|---|---|---|---|---|---|---|---|---|
| C++ | SIMD Intrinsics | 281.0 | 187.0 | 105.4 | 74.0 | 32.3 | 26.4 | 12.9 | 8.5 |
| Scalar | 141.2 | 100.0 | 84.8 | 35.7 | 15.9 | ||||
| WebAssembly (no threads, no SIMD) | 8.4 | 5.0 | 8.1 | 5.6 | |||||
| C# | Unity Burst "manual" SIMD | 227.2 | 133.0 | 103.7 | 60.0 | 29.7 | |||
| Unity Burst | 82.0 | 36.0 | |||||||
| Unity (Editor) | 6.5 | 3.4 | |||||||
| Unity (player Mono) | 6.7 | 3.5 | |||||||
| Unity (player IL2CPP) | 39.1 | 63.8 | 17.2 | ||||||
| .NET 6.0 | 91.5 | 53.0 | 40.9 | ||||||
| .NET Core 2.0 | 86.1 | 53.0 | 23.6 | ||||||
| Mono --llvm | 35.1 | 22.0 | |||||||
| Mono | 23.6 | 3.6 | 6.1 |
More detailed specs of the machines above are:
Ryzen 5950: AMD Ryzen 5950X (3.4GHz, 16c/32t), Visual Studio 2022.AMD TR1950: AMD ThreadRipper 1950X (3.4GHz, SMT disabled - 16c/16t), Visual Studio 2017.MBP 2021: Apple MacBook Pro M1 Max (8+2 cores), Xcode 13.2.MBP 2018: Apple MacBook Pro mid-2018 (Core i9 2.9GHz, 6c/12t).MBA 2020: Apple MacBook Air 2020 (Core i7 1.2GHz, 4c/8t).iPhone 11: A13 chip.iPhone X: A11 chip.iPhone SE: A9 chip.
Software versions:
- Unity 2021.3.16. Burst 1.6.6 (safety checks off). C# testing in editor, Release mode.
- Mono 6.12.
And on the GPU, via a compute shader in D3D11 or Metal depending on the platform:
| GPU | Perf |
|---|---|
| D3D11 | |
| GeForce RTX 3080Ti | 3920 |
| GeForce GTX 1080Ti | 1854 |
| Metal | |
| MBP 2024 (M4 Max) | 1680 |
| MBP 2021 (M1 Max) | 1065 |
| MBP 2018 (Radeon Pro 560X) | 246 |
| MBA 2020 (Iris Plus) | 201 |
| iPhone 11 Pro (A13) | 80 |
| iPhone X (A11) | 46 |
| iPhone SE (A9) | 20 |
A lot of stuff in the implementation is totally suboptimal or using the tech in a "wrong" way. I know it's just a simple toy, ok :)
Building
- C++ projects:
- Windows (Visual Studio 2017) in
Cpp/Windows/ToyPathTracer.sln. DX11 Win32 app that displays result as a fullscreen CPU-updated or GPU-rendered texture. Pressing G toggles between GPU and CPU tracing, A toggles animation, P toggles progressive accumulation. - Mac/iOS (Xcode 10) in
Cpp/Apple/ToyPathTracer.xcodeproj. Metal app that displays result as a fullscreen CPU-updated or GPU-rendered texture. Pressing G toggles between GPU and CPU tracing, A toggles animation, P toggles progressive accumulation. Should work on both Mac (Test Mactarget) and iOS (Test iOStarget). - WebAssembly in
Cpp/Emscripten/build.sh. CPU, single threaded, no SIMD.
- Windows (Visual Studio 2017) in
- C# project in
Cs/TestCs.sln. A command line app that renders some frames and dumps out final TGA screenshot at the end. - Unity project in
Unity. I used Unity 2021.3.16.
Explorar proyectos relacionados
Cargando…