GRay: Ray Tracing 3D Gaussians Near the Speed of Splats

2026-06-29

AI Summary

While 3D Gaussian rasterization is highly efficient, existing ray tracing methods struggle to balance rendering quality with optimization and rendering speed. This work proposes GRay—the first fast ray tracer tailored for 3D Gaussians—leveraging the key insight that rays interact only with intersecting Gaussians. It further reveals, for the first time, the critical role of dense, small-Gaussian initialization in accelerating ray tracing. Building on these observations, GRay integrates spatial acceleration structures with a customized architecture to achieve high-quality rendering at speeds approaching those of rasterization. Experiments demonstrate that GRay achieves nearly 4× faster rendering and almost 10× faster optimization compared to current 3D Gaussian ray tracing approaches, substantially narrowing the performance gap between ray tracing and rasterization.

Abstract

3D Gaussian Splatting (3DGS) is a popular representation for radiance field reconstruction, distinguished by the rendering speed of its rasterization-based renderer. While 3D Gaussians can also be ray traced, this approach has so far been slower, with 3D Gaussian Ray Tracing (3DGRT) taking nearly one order of magnitude longer to optimize. To address this, we present GRay, a fast ray tracer for 3D Gaussians designed to close this performance gap and match 3DGS's speed. Our method leverages the algorithmic difference between both approaches: unlike rasterization, ray tracing evaluates only Gaussians that are actually intersected by a ray, leading to potentially logarithmic--rather than linear--scaling in the number of primitives. This property allows ray tracing to better exploit dense scenes composed of numerous tiny Gaussians, a configuration which has largely been overlooked. Notably, we show that dense initialization--which creates many small Gaussians--slows down rasterization, but instead speeds up ray tracing. Designed to leverage this effect, GRay renders nearly 4x faster and optimizes nearly 10x faster than 3DGRT while maintaining similar quality, and has competitive speed with 3DGS albeit at somewhat lower quality. Code is available at https://repo-sam.inria.fr/nerphys/gray.

Problem

• 3D Gaussian Splatting
• Ray Tracing
• Rendering Speed
• Radiance Field
• Performance Gap

Innovation

• 3D Gaussian Splatting
• Ray Tracing
• Dense Initialization
• Real-time Rendering
• Radiance Fields