Ray Tracing in Makie: From Research Data to Photorealistic Renders
Hacker News
February 19, 2026
AI-Generated Deep Dive Summary
RayMakie and Hikari bring photorealistic ray tracing directly to the Makie framework, enabling researchers and developers to render complex 3D data with stunning realism. This integration allows any Makie scene to be rendered using global illumination, volumetric media, spectral rendering, and physically-based materials, all powered by GPU acceleration. By eliminating the need for separate tools or export steps, RayMakie simplifies the workflow for researchers who generate intricate 3D datasets, allowing them to visualize their work interactively and share it effectively.
The development of RayMakie and Hikari is particularly significant for fields like climate science, structural biology, and fluid dynamics, where complex simulations often require clear and compelling visualizations. Traditionally, exporting meshes and learning new tools have been barriers to effective communication of research data. By building ray tracing directly into Makie, researchers can now explore scenes interactively with GLMakie and switch to photorealistic rendering seamlessly using RayMakie. This approach preserves the familiar Makie API for scene construction and lighting while leveraging the power of GPU-based ray tracing.
Hikari, a Julia port of pbrt-v4, implements advanced features like wavefront volumetric path tracing and spectral rendering, supporting participating media such as NanoVDB volumes and physically-based materials. Raycore.jl, the standalone intersection engine behind Hikari, utilizes AMD's Radeon Rays SDK for efficient GPU acceleration. RayMakie connects these components to Makie's scene graph, allowing users to define materials, lights, and integrators through standard Makie calls. The system is designed to be hackable, enabling experimentation and customization in a high-level language.
Ray Tracing in Julia offers several advantages: performance on par with C++ ray tracers, cross-vendor GPU support (including AMD, NVIDIA, and CPU via KernelAbstractions.jl), and seamless integration with Makie's
Verticals
techstartups
Originally published on Hacker News on 2/19/2026