Fast Simulations of Gravitational Many-body Problem on RV770 GPU

TL;DR

This paper demonstrates how GPU acceleration can significantly speed up gravitational many-body simulations, achieving peak performance of about 1 Tflops, thus offering a more efficient computational approach.

Contribution

The paper introduces GPU-based optimization techniques for gravitational many-body simulations, significantly improving computational speed over traditional CPU methods.

Findings

Achieved peak performance of about 1 Tflops on GPU

Demonstrated substantial speed-up over CPU implementations

Optimized GPU algorithms for gravitational calculations

Abstract

The gravitational many-body problem is a problem concerning the movement of bodies, which are interacting through gravity. However, solving the gravitational many-body problem with a CPU takes a lot of time due to O(N^2) computational complexity. In this paper, we show how to speed-up the gravitational many-body problem by using GPU. After extensive optimizations, the peak performance obtained so far is about 1 Tflops.