Real Time Cluster Path Tracing
Feng Xie, Petro Mishchuk, Warren Hunt

TL;DR
This paper introduces the first real-time cluster path tracing system capable of rendering film-quality animations with deforming characters, leveraging distributed GPU clusters for scalable performance and global illumination effects.
Contribution
It presents a novel architecture and implementation of a real-time, production-quality cluster path tracer supporting animated, deforming scenes using open source Blender and Cycles.
Findings
Achieves 15-30 fps rendering of complex animated scenes.
Scales linearly with the number of GPU cluster nodes.
Supports global illumination for deforming characters.
Abstract
Photorealistic rendering effects are common in films, but most real time graphics today still rely on scan-line based multi-pass rendering to deliver rich visual experiences. While there have been prior works in distributed path tracing for static scene and objects under rigid motion, real time path tracing of deforming characters has to support per-frame dynamic BVH changes. We present the architecture and implementation of the first real-time production quality cluster path tracing renderer that supports film quality animation and deformation. We build our cluster path tracing system using the open source Blender and its GPU accelerated production quality renderer Cycles. Our system's rendering performance and quality scales linearly with the number of RTX cluster nodes used. It is able to generate and deliver path traced images with global illumination effects to remote light-weight…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
