NBSymple, a double parallel, symplectic N-body code running on Graphic Processing Units
arXiv:1003.3896 · doi:10.1016/j.newast.2010.11.004
Abstract
We present and discuss the characteristics and performances, both in term of computational speed and precision, of a numerical code which numerically integrates the equation of motions of N 'particles' interacting via Newtonian gravitation and move in an external galactic smooth field. The force evaluation on every particle is done by mean of direct summation of the contribution of all the other system's particle, avoiding truncation error. The time integration is done with second-order and sixth-order symplectic schemes. The code, NBSymple, has been parallelized twice, by mean of the Computer Unified Device Architecture to make the all-pair force evaluation as fast as possible on high-performance Graphic Processing Units NVIDIA TESLA C 1060, while the O(N) computations are distributed on various CPUs by mean of OpenMP Application Program. The code works both in single precision floating point arithmetics or in double precision. The use of single precision allows the use at best of the GPU performances but, of course, limits the precision of simulation in some critical situations. We find a good compromise in using a software reconstruction of double precision for those variables that are most critical for the overall precision of the code. The code is available on the web site astrowww.phys.uniroma1.it/dolcetta/nbsymple.html
Paper composed by 29 pages, including 9 figures. Submitted to New Astronomy.
References in corpus (6)
- High Performance Direct Gravitational N-body Simulations on Graphics Processing Units -- II: An implementation in CUDA
- SAPPORO: A way to turn your graphics cards into a GRAPE-6
- High Performance Direct Gravitational N-body Simulations on Graphics Processing Unit I: An implementation in Cg
- High Performance Direct Gravitational N-body Simulations on Graphics Processing Units
- Formation and evolution of clumpy tidal tails around globular clusters
- The Chamomile Scheme: An Optimized Algorithm for N-body simulations on Programmable Graphics Processing Units
Cited by in corpus (12)
- A fully parallel, high precision, N-body code running on hybrid computing platforms
- Swarm-NG: a CUDA Library for Parallel n-body Integrations with focus on Simulations of Planetary Systems
- Stellar dynamics in gas: The role of gas damping
- Clumpy streams in a smooth dark halo: the case of Palomar 5
- A Performance Comparison of Different Graphics Processing Units Running Direct N-Body Simulations
- Mergers, tidal interactions, and mass exchange in a population of disc globular clusters: II. Long-term evolution
- QYMSYM: A GPU-Accelerated Hybrid Symplectic Integrator That Permits Close Encounters
- The secular evolution of the Kuiper belt after a close stellar encounter
- A Monte Carlo analysis of the velocity dispersion of the globular cluster Palomar 14
- Multiple stellar population mass loss in massive Galactic globular clusters
- A study of low-energy transfer orbits to the Moon: towards an operational optimization technique
- How well do STARLAB and NBODY compare? II: Hardware and accuracy