RTNN: Accelerating Neighbor Search Using Hardware Ray Tracing

TL;DR

This paper introduces RTNN, a novel approach that formulates neighbor search as a ray tracing problem and leverages GPU hardware to significantly accelerate the process, achieving up to 65x speedups.

Contribution

It proposes a new formulation of neighbor search as a ray tracing problem and introduces two optimization techniques to improve GPU-based performance.

Findings

Achieves 2.2x to 65x speedup over existing libraries

Demonstrates effective utilization of GPU ray tracing hardware

Provides open-source implementation for community use

Abstract

Neighbor search is of fundamental important to many engineering and science fields such as physics simulation and computer graphics. This paper proposes to formulate neighbor search as a ray tracing problem and leverage the dedicated ray tracing hardware in recent GPUs for acceleration. We show that a naive mapping under-exploits the ray tracing hardware. We propose two performance optimizations, query scheduling and query partitioning, to tame the inefficiencies. Experimental results show 2.2X -- 65.0X speedups over existing neighbor search libraries on GPUs. The code is available at https://github.com/horizon-research/rtnn.