Abstract
Methods for parallelizing Ising model simulations are presented. A parallel single-spin Metropolis algorithm [J. Chem. Phys. 21, 1087 (1953)] has been implemented with a speedup of 27 on 50 processors of the KSR-1 parallel computer. A parallel Swendsen-Wang algorithm [Phys. Rev. Lett. 58, 86 (1987)] obtains a speedup of 3.2 on nine processors of the same computer. Both of these simulations were carried out on 200 x 200 lattices. The parallel local cluster algorithm [Phys. Rev. Lett. 71, 2070 (1993)] has been implemented with an almost linear speedup. We also discuss ongoing research using the parallel local cluster algorithm.
| Original language | English |
|---|---|
| Pages (from-to) | 1623-1628 |
| Number of pages | 6 |
| Journal | Physical Review. E, Statistical Physics, Plasmas, Fluids, and Related Interdisciplinary Topics |
| Volume | 50 |
| Issue number | 2 |
| Publication status | Published - Aug 1994 |
Keywords
- MONTE-CARLO SIMULATION
- SPIN SYSTEMS