MESH-REFINED P3M - A FAST ADAPTIVE N-BODY ALGORITHM

被引:182
作者
COUCHMAN, HMP
机构
[1] Department of Astronomy, University of Toronto, McLennan Physical Laboratories, Toronto, Ont. M5S 1A1
关键词
Cosmology; Galaxies: Clustering; Numerical methods;
D O I
10.1086/185939
中图分类号
P1 [天文学];
学科分类号
0704 ;
摘要
Substantial improvements in the performance of "particle-particle-particle-mesh" (P3) algorithms may be achieved by allowing spatially adaptive mesh refinements in regions of high particle density. The cycle time under heavy clustering is approximately 5 times faster that for a uniform particle distribution: approximately 10-20 times faster than conventional P3M. The algorithm is roughly 50% faster than a "tree code" while requiring significantly less memory.
引用
收藏
页码:L23 / &
相关论文
共 6 条