Abstract
In a common approach for parallel processing applied to simulations of manyparticle
systems with short-ranged interactions and uniform density, the simulation
cell is partitioned into domains of equal shape and size, each of which is assigned
to one processor. We compare the commonly used simple-cubic (SC) domain shape
to domain shapes chosen as the Voronoi cells of BCC and FCC lattices. The latter
two are found to result in superior partitionings with respect to communication
overhead. Other domain shapes, relevant for a small number of processors, are also
discussed. The higher efficiency with BCC and FCC partitionings is demonstrated
in simulations of the sillium model for amorphous silicon.
systems with short-ranged interactions and uniform density, the simulation
cell is partitioned into domains of equal shape and size, each of which is assigned
to one processor. We compare the commonly used simple-cubic (SC) domain shape
to domain shapes chosen as the Voronoi cells of BCC and FCC lattices. The latter
two are found to result in superior partitionings with respect to communication
overhead. Other domain shapes, relevant for a small number of processors, are also
discussed. The higher efficiency with BCC and FCC partitionings is demonstrated
in simulations of the sillium model for amorphous silicon.
| Original language | English |
|---|---|
| Pages (from-to) | 121-134 |
| Number of pages | 14 |
| Journal | Computer Physics Communications |
| Volume | 149 |
| Issue number | 3 |
| DOIs | |
| Publication status | Published - Jan 2003 |
Bibliographical note
stijnman03Keywords
- Mathematics
- Wiskunde en computerwetenschappen
- Landbouwwetenschappen
- Wiskunde: algemeen
- parallel computing
- particle simulations
- space partitioning