Adaptive fast multipole methods on the GPU

Abstract

We present a highly general implementation of fast multipole methods on graphics processing units (GPUs). Our two-dimensional double precision code features an asymmetric type of adaptive space discretization leading to a particularly elegant and flexible implementation. All steps of the multipole algorithm are efficiently performed on the GPU, including the initial phase which assembles the topological information of the input data. Through careful timing experiments we investigate the effects of the various peculiarities of the GPU architecture.

Software available at http://user.it.uu.se/~stefane/freeware.html

References in corpus (6)

• A Tuned and Scalable Fast Multipole Method as a Preeminent Algorithm for Exascale Systems
• PetFMM--A dynamically load-balancing parallel fast multipole library
• N-body simulation for self-gravitating collisional systems with a new SIMD instruction set extension to the x86 architecture, Advanced Vector eXtensions
• Enhanced molecular dynamics performance with a programmable graphics processor
• On well-separated sets and fast multipole methods
• Treecode and fast multipole method for N-body simulation with CUDA

Cited by in corpus (5)

• A hybrid Eulerian-Lagrangian flow solver
• Optimizing the adaptive fast multipole method for fractal sets
• Dynamic autotuning of adaptive fast multipole methods on hybrid multicore CPU & GPU systems
• A general high order two-dimensional panel method
• Distributed and Adaptive Fast Multipole Method In Three Dimensions