Show HN: A physically-based GPU ray tracer written in Julia
6 days ago
- #visualization
- #julia
- #ray-tracing
- RayMakie and Hikari introduce a physically-based GPU ray tracing pipeline integrated into Makie, enabling photorealistic path tracing for any Makie scene.
- Features include global illumination, volumetric media, spectral rendering, and physically-based materials, all running on the GPU.
- The integration eliminates the need for exporting meshes or learning new tools, allowing seamless transition from interactive exploration to photorealistic rendering.
- Julia's performance and cross-vendor GPU support enable efficient ray tracing, competitive with C++ implementations.
- Hikari is a Julia port of pbrt-v4, supporting wavefront volumetric path tracing with spectral rendering and physically-based materials.
- Applications span climate science, structural biology, fluid dynamics, and particle physics, showcasing the versatility of RayMakie.
- Custom materials and media can be defined, leveraging Julia's multiple dispatch for novel visualization and simulation use cases.
- Interactive exploration is supported with progressive refinement of path-traced images.
- Future optimizations include memory management, performance improvements, and broader GPU backend testing.
- The project is supported by the Sovereign Tech Agency and Muon Space, with contributions from the Julia community.