Ray Tracing in Makie: From Research Data to Photorealistic Renders
We're excited to announce RayMakie and Hikari, a physically-based GPU ray tracing pipeline integrated directly into Makie. Any Makie scene can now be rendered with photorealistic path tracing: just swap out the backend and get global illumination, volumetric media, spectral rendering, and physically-based materials, all running on the GPU.
All showcase scripts and demo scenes from this post are available at github.com/SimonDanisch/RayDemo.
Note: RayMakie, Hikari, and Raycore are not fully released yet — we plan to publish official releases in the coming weeks. In the meantime, the Project.toml in RayDemo will be kept up to date so you can already try things out.
Why Ray Tracing in Julia?
Research groups across many fields (climate science, structural biology, fluid dynamics, particle physics) produce complex 3D data that needs to be communicated clearly and compellingly. Photorealistic rendering can transform dense simulation output into images that reveal structure and tell a story. But getting research data into a traditional ray tracer usually means exporting meshes, learning new tools, and losing the interactive workflow.
By building ray tracing directly into Makie, we eliminate that gap. The same scene you explore interactively with GLMakie can be rendered photorealistically with RayMakie, no export step, no new API to learn.
Writing the implementation in Julia gives us:
Performance : Julia compiles to efficient GPU code, competitive with C++ ray tracers
Cross-vendor GPU support : A single codebase runs on AMD, NVIDIA, and CPU via KernelAbstractions.jl
... continue reading