full
search.png

Search

close.png

Cancel

arrow
Volume 30, Issue 4
Distributed and Adaptive Fast Multipole Method in Three Dimensions

Jonathan Bull & Stefan Engblom

DOI: 10.4208/cicp.OA-2020-0072

Commun. Comput. Phys., 30 (2021), pp. 959-984.

Published online: 2021-08

Preview Purchase PDF 2348 49149

Cited by

google scholar semantic scholar
Export citation
  • Abstract

We develop a general distributed implementation of an adaptive fast multipole method in three space dimensions. We rely on a balanced type of adaptive space discretization which supports a highly transparent and fully distributed implementation. A complexity analysis indicates favorable scaling properties and numerical experiments on up to 512 cores and 1 billion source points verify them. The parameters controlling the algorithm are subject to in-depth experiments and the performance response to the input parameters implies that the overall implementation is well-suited to automated tuning.

  • Keywords

Adaptive fast multipole method, distributed parallelization, Message Passing Interface (MPI), multipole acceptance criterion, balanced tree.

  • AMS Subject Headings

65Y05, 68W10, 65Y10, 65Y20, 68W15

  • Copyright

COPYRIGHT: © Global Science Press

  • Email address
  • BibTex
  • RIS
  • TXT
@Article{CiCP-30-959, author = {Bull , Jonathan and Engblom , Stefan}, title = {Distributed and Adaptive Fast Multipole Method in Three Dimensions}, journal = {Communications in Computational Physics}, year = {2021}, volume = {30}, number = {4}, pages = {959--984}, abstract = {

We develop a general distributed implementation of an adaptive fast multipole method in three space dimensions. We rely on a balanced type of adaptive space discretization which supports a highly transparent and fully distributed implementation. A complexity analysis indicates favorable scaling properties and numerical experiments on up to 512 cores and 1 billion source points verify them. The parameters controlling the algorithm are subject to in-depth experiments and the performance response to the input parameters implies that the overall implementation is well-suited to automated tuning.

}, issn = {1991-7120}, doi = {https://doi.org/10.4208/cicp.OA-2020-0072}, url = {http://global-sci.org/intro/article_detail/cicp/19390.html} }
TY - JOUR T1 - Distributed and Adaptive Fast Multipole Method in Three Dimensions AU - Bull , Jonathan AU - Engblom , Stefan JO - Communications in Computational Physics VL - 4 SP - 959 EP - 984 PY - 2021 DA - 2021/08 SN - 30 DO - http://doi.org/10.4208/cicp.OA-2020-0072 UR - https://global-sci.org/intro/article_detail/cicp/19390.html KW - Adaptive fast multipole method, distributed parallelization, Message Passing Interface (MPI), multipole acceptance criterion, balanced tree. AB -

We develop a general distributed implementation of an adaptive fast multipole method in three space dimensions. We rely on a balanced type of adaptive space discretization which supports a highly transparent and fully distributed implementation. A complexity analysis indicates favorable scaling properties and numerical experiments on up to 512 cores and 1 billion source points verify them. The parameters controlling the algorithm are subject to in-depth experiments and the performance response to the input parameters implies that the overall implementation is well-suited to automated tuning.

Bull , Jonathan and Engblom , Stefan. (2021). Distributed and Adaptive Fast Multipole Method in Three Dimensions. Communications in Computational Physics. 30 (4). 959-984. doi:10.4208/cicp.OA-2020-0072
Copy to clipboard
BibteX RIS TXT
The citation has been copied to your clipboard