TraverseWork with us
Menu

Traverse Research

Neural radiance caching on mobile GPUs

Samsung and Traverse researchers co-authored Neural Radiance Cache Implementation on Mobile GPU for SIGGRAPH Asia 2025.

Collin Allen, Luca Quartesan, Jasper de Winther and Jasper Bekkers1 min read
The Evolve benchmark scene from Figure 1, with neural radiance caching visible in the arch-shaped mirror on the left. Captured with the desktop-quality preset.
The Evolve benchmark scene from Figure 1, with neural radiance caching visible in the arch-shaped mirror on the left. Captured with the desktop-quality preset.

A published mobile implementation

Neural Radiance Cache Implementation on Mobile GPU appears in the Proceedings of the SIGGRAPH Asia 2025 Technical Communications. It examines neural radiance caching on mobile GPUs, connecting implementation with hardware/software performance modelling.

Authors and affiliations

The work builds on the broader neural-radiance-caching literature and is implemented in the context of Traverse’s rendering research. It follows the team’s earlier NeuBTF material-representation publication at SIGGRAPH Asia 2022.

Read the evidence

The four-page publication details the method, experimental conditions and results.

Read the ACM publication.

Read the publication

More from the team, soon.

Newsletter signup opens when delivery and double opt-in are connected.