Introducing RayMakie and Hikari, a physically-based GPU ray tracer integrated into Julia's Makie visualization ecosystem. Researchers can now render any Makie scene with photorealistic path tracing using global illumination and physically-based materials. Julia compiles custom physics directly into GPU kernels, enabling extensions like a gravitational lensing black hole implementation in just ~200 lines of code.

The project bridges a critical gap between research data visualization and photorealistic rendering. Complex 3D simulations from climate science, structural biology, and fluid dynamics can transform into compelling images without export steps or new APIs. Hikari, a Julia port of pbrt-v4, supports spectral rendering and volumetric path tracing across AMD, NVIDIA, and CPU platforms via KernelAbstractions.jl.

Practical applications span multiple scientific domains. Breeze.jl renders photorealistic volumetric clouds from MIT's CliMA climate simulations. PlantGeom.jl creates biophysically accurate plant models for agricultural research. ProtPlot.jl visualizes protein structures with materials that reveal different structural aspects. The Makie ecosystem now covers the entire spectrum from interactive exploration to publication-ready renders.