C/C++ engineer focused on real-time graphics, performance, and systems-level work.
A real-time 3D cyclic cellular automata simulation implemented with WebGPU compute shaders
- Uses a (8, 4, 2) 3D workgroup size to maximize cache-coherent reads along the X axis
- Bitpacks cell states to signifcantly reduce overall memory bandwidth
- Assigns multiple cellular automata states to a single GPU thread, improving memory re-use and ALU usage
- Rendered by rasterizing a bounding box and performing per-pixel DDA voxel traversal along the view ray
A brute-force ray tracer, optimized with SIMD and multi-threading
Live Demo: https://simd-ray-tracer.netlify.app/
- Compiles to WASM and x64, supporting WebAssembly SIMD, SSE2, and AVX2
- Data is layed out in an array of struct of arrays, ensuring good cache locality and compatibility with SIMD
- Supports matte, specular, and fresnel materials