Implementation and performance of FDPS: a framework for developing parallel particle simulation codes

Abstract

We present the basic idea, implementation, measured performance and performance model of FDPS (Framework for developing particle simulators). FDPS is an application-development framework which helps the researchers to develop simulation programs using particle methods for large-scale distributed-memory parallel supercomputers. A particle-based simulation program for distributed-memory parallel computers needs to perform domain decomposition, exchange of particles which are not in the domain of each computing node, and gathering of the particle information in other nodes which are necessary for interaction calculation. Also, even if distributed-memory parallel computers are not used, in order to reduce the amount of computation, algorithms such as Barnes-Hut tree algorithm or Fast Multipole Method should be used in the case of long-range interactions. For short-range interactions, some methods to limit the calculation to neighbor particles are necessary. FDPS provides all of these functions which are necessary for efficient parallel execution of particle-based simulations as "templates", which are independent of the actual data structure of particles and the functional form of the particle-particle interaction. By using FDPS, researchers can write their programs with the amount of work necessary to write a simple, sequential and unoptimized program of O(N 2 ) calculation cost, and yet the program, once compiled with FDPS, will run efficiently on largescale parallel supercomputers. A simple gravitational N -body program can be written in around 120 lines. We report the actual performance of these programs and the performance model. The weak scaling performance is very good, and almost linear speedup was obtained for up to the full system of K computer. The minimum calculation time per timestep is in the range of 30 ms (N = 10 7 ) to 300 ms (N = 10 9 ). These are currently limited by the time for the calculation of the domain decomposition and communication necessary for the interaction calculation. We discuss how we can overcome these bottlenecks.

Does Explosive Nuclear Burning Occur in Tidal Disruption Events of White Dwarfs by Intermediate-mass Black Holes?

“…We calculate self gravity among SPH particles with adaptive gravitational softening (Price & Monaghan 2007). We optimize the SPH and self-gravity calculations on distributedmemory systems, using FDPS (Iwasawa et al 2015;Iwasawa et al 2016) and explicit AVX instructions (e.g. Tanikawa et al 2012Tanikawa et al , 2013.…”

High-resolution simulations of catastrophic disruptions: Resultant shape distributions

“…We use the Tillotson equation of state (Tillotson 1962). The simulation code is parallelized using Framework for Developing Particle Simulator (Iwasawa et al 2015;Iwasawa et al 2016).…”

Collisional elongation: Possible origin of extremely elongated shape of 1I/‘Oumuamua

“…To investigate the formation of irregularly shaped objects through planetesimal collisions and gravitational reaccumulation of fragments, we use a Smoothed Particle Hydrodynamics (SPH) code for elastic dynamics (Libersky & Petschek 1991) that includes self-gravity, a fracture model for rocky material (Grady-Kipp fragmentation model: Benz & Asphaug 1995), and a friction model for granular material (Drucker-Prager yield criterion with zero cohesion: Jutzi 2015). We parallelize our code utilizing Framework for Developing Particle Simulator (Iwasawa et al 2015(Iwasawa et al , 2016. We use basaltic spheres with zero rotation as colliding planetesimals.…”