SAPPORO: A way to turn your graphics cards into a GRAPE-6
arXiv:0902.4463 · doi:10.1016/j.newast.2009.03.002
Abstract
We present Sapporo, a library for performing high-precision gravitational N-body simulations on NVIDIA Graphical Processing Units (GPUs). Our library mimics the GRAPE-6 library, and N-body codes currently running on GRAPE-6 can switch to Sapporo by a simple relinking of the library. The precision of our library is comparable to that of GRAPE-6, even though internally the GPU hardware is limited to single precision arithmetics. This limitation is effectively overcome by emulating double precision for calculating the distance between particles. The performance loss of this operation is small (< 20%) compared to the advantage of being able to run at high precision. We tested the library using several GRAPE-6-enabled N-body codes, in particular with Starlab and phiGRAPE. We measured peak performance of 800 Gflop/s for running with 10^6 particles on a PC with four commercial G92 architecture GPUs (two GeForce 9800GX2). As a production test, we simulated a 32k Plummer model with equal mass stars well beyond core collapse. The simulation took 41 days, during which the mean performance was 113 Gflop/s. The GPU did not show any problems from running in a production environment for such an extended period of time.
13 pages, 9 figures, accepted to New Astronomy
References in corpus (9)
- High Performance Direct Gravitational N-body Simulations on Graphics Processing Units -- II: An implementation in CUDA
- A multiphysics and multiscale software environment for modeling astrophysical systems
- Performance Analysis of Direct N-Body Algorithms on Special-Purpose Supercomputers
- Harvesting graphics power for MD simulations
- 6th and 8th Order Hermite Integrator for N-body Simulations
- High Performance Direct Gravitational N-body Simulations on Graphics Processing Unit I: An implementation in Cg
- A Hybrid N-Body Code Incorporating Algorithmic Regularization and Post-Newtonian Forces
- Parallel Algorithm for Solving Kepler's Equation on Graphics Processing Units: Application to Analysis of Doppler Exoplanet Searches
- The Chamomile Scheme: An Optimized Algorithm for N-body simulations on Programmable Graphics Processing Units
Cited by in corpus (89)
- Data Mining and Machine Learning in Astronomy
- Massive black hole binaries from runaway collisions: the impact of metallicity
- Dynamics of stellar black holes in young star clusters with different metallicities - II. Black hole-black hole binaries
- The DRAGON simulations: globular cluster evolution with a million stars
- Accelerating NBODY6 with Graphics Processing Units
- NBODY6++GPU: Ready for the gravitational million-body problem
- Efficient Merger of Binary Supermassive Black Holes in Merging Galaxies
- Relativistic Dynamics and Extreme Mass Ratio Inspirals
- Dissipationless Formation and Evolution of the Milky Way Nuclear Star Cluster
- A sparse octree gravitational N-body code that runs entirely on the GPU processor
- Supermassive Black Hole Binary Evolution in Axisymmetric Galaxies: the final parsec problem is not a problem
- The final-parsec problem in the collisionless limit
- Dynamical Evolution and Spatial Mixing of Multiple Population Globular Clusters
- PeTar: a high-performance N-body code for modeling massive collisional stellar systems
- N-body simulations of gravitational dynamics
- Implementation and performance of FDPS: A Framework Developing Parallel Particle Simulation Codes
- Structure in phase space associated with spiral and bar density waves in an N-body galactic disk
- GAMER: a GPU-Accelerated Adaptive Mesh Refinement Code for Astrophysics
- The Role of the Kozai-Lidov Mechanism in Black Hole Binary Mergers in Galactic Centers
- Dynamical Friction around Supermassive Black Holes
- Dynamical friction and the evolution of Supermassive Black hole Binaries: the final hundred-parsec problem
- NOTE: Explaining why the Uranian satellites have equatorial prograde orbits despite the large planetary obliquity
- Mergers of Unequal Mass Galaxies: Supermassive Black Hole Binary Evolution and Structure of Merger Remnants
- Long-term evolution of massive black hole binaries. IV. Mergers of galaxies with collisionally relaxed nuclei
- The GENGA Code: Gravitational Encounters in N-body simulations with GPU Acceleration
- Dynamics of stellar black holes in young star clusters with different metallicities - I. Implications for X-ray binaries
- Collisionless loss-cone refilling: there is no final parsec problem
- On the distribution of stellar remnants around massive black holes: slow mass segregation, star cluster inspirals and correlated orbits
- The final-parsec problem in non-spherical galaxies revisited
- A fully parallel, high precision, N-body code running on hybrid computing platforms
- A new Fokker-Planck approach for relaxation-driven evolution of galactic nuclei
- Dynamical friction of massive objects in galactic centres
- The loss cone problem in axisymmetric nuclei
- Evolution of second generation stars in stellar disks of globular and nuclear clusters: Omega Centauri as a test case
- Phantom-GRAPE: numerical software library to accelerate collisionless -body simulation with SIMD instruction set on x86 architecture
- Kinematical fingerprints of star cluster early dynamical evolution
- Effects of Intermediate Mass Black Holes on Nuclear Star Clusters
- N-body simulation for self-gravitating collisional systems with a new SIMD instruction set extension to the x86 architecture, Advanced Vector eXtensions
- A new Monte Carlo method for dynamical evolution of non-spherical stellar systems
- A critical look at the merger scenario to explain multiple populations and rotation in iron-complex globular clusters
- body dynamics of Intermediate mass-ratio inspirals in globular clusters
- Gravitational tree-code on graphics processing units: implementation in CUDA
- Hierarchical N-body simulations with auto-tuning for heterogeneous systems
- The Evolution of Protoplanetary Disks in the Arches Cluster
- Reconstructing the Arches I: Constraining the Initial Conditions
- Accelerating Dust Temperature Calculations with Graphics Processing Units
- Swarm-NG: a CUDA Library for Parallel n-body Integrations with focus on Simulations of Planetary Systems
- NBSymple, a double parallel, symplectic N-body code running on Graphic Processing Units
- Relativistic dynamics of stars near a supermassive black hole
- A deep view into the nucleus of the Sagittarius Dwarf Spheroidal Galaxy with MUSE. II. Kinematic characterization of the stellar populations
- Astrophysical Supercomputing with GPUs: Critical Decisions for Early Adopters
- Do open star clusters evolve toward energy equipartition?
- Roche-lobe overflow systems powered by black holes in young star clusters: the importance of dynamical exchanges
- Impact of metallicity on the evolution of young star clusters
- PyCOOL - a Cosmological Object-Oriented Lattice code written in Python
- Accelerated FDPS --- Algorithms to Use Accelerators with FDPS
- Defeating stochasticity: coalescence timescales of massive black holes in galaxy mergers
- Rates of capture of stars by supermassive black holes in non-spherical galactic nuclei
- Chaos in self-gravitating many-body systems: Lyapunov time dependence of and the influence of general relativity
- FROST: a momentum-conserving CUDA implementation of a hierarchical fourth-order forward symplectic integrator
- A Performance Comparison of Different Graphics Processing Units Running Direct N-Body Simulations
- The impact of metallicity-dependent mass loss versus dynamical heating on the early evolution of star clusters
- Roche volume filling and the dissolution of open star clusters
- Star formation at the Galactic Centre: coevolution of multiple young stellar discs
- QYMSYM: A GPU-Accelerated Hybrid Symplectic Integrator That Permits Close Encounters
- Schrödinger Evolution of Self-Gravitating Disks
- Dynamical evolution of massive black holes in galactic-scale N-body simulations - introducing the regularized tree code "rVINE"
- A prescription and fast code for the long-term evolution of star clusters - II. Unbalanced and core evolution
- MYRIAD: A new N-body code for simulations of Star Clusters
- A New Hybrid Technique for Modeling Dense Star Clusters
- A Monte Carlo study of early gas expulsion and evolution of star clusters: new simulations with the MOCCA code in the AMUSE framework
- A connected component-based method for efficiently integrating multiscale -body systems
- Simulating Star Clusters with the AMUSE Software Framework: I. Dependence of Cluster Lifetimes on Model Assumptions and Cluster Dissolution Modes
- Star cluster evolution in barred disc galaxies. I. Planar periodic orbits
- Near-Field Effects of Cherenkov Radiation Induced by Ultra High Energy Cosmic Neutrinos
- The influence of dense gas rings on the dynamics of a stellar disk in the Galactic center
- Finite element modelling of perturbed stellar systems
- High Performance Gravitational N-body Simulations on a Planet-wide Distributed Supercomputer
- Jeans Instability in a Tidally Disrupted Halo Satellite Galaxy
- A pilgrimage to gravity on GPUs
- On the Effects of Subvirial Initial Conditions and the Birth Temperature of R136
- GraviDy, a GPU modular, parallel direct-summation body integrator: Dynamics with softening
- Nuclear star clusters as probes of dark matter halos: the case of the Sagittarius Dwarf Spheroidal Galaxy
- How well do STARLAB and NBODY compare? II: Hardware and accuracy
- The 800pc long tidal tails of the Hyades star cluster: Possible discovery of candidate epicyclic overdensities from an open star cluster
- Dynamical Processes in Globular Clusters
- Efficient magnetohydrodynamic simulations on graphics processing units with CUDA
- Sapporo2: A versatile direct -body library
- GPU-Enabled Particle-Particle Particle-Tree Scheme for Simulating Dense Stellar Cluster System