A parallel gravitational N-body kernel

TL;DR

This paper introduces a parallelization method for the { t kira} N-body gravitational integrator, enabling efficient simulations of dense stellar systems on various parallel computer architectures with good scalability.

Contribution

It presents a novel 'j-parallelization' strategy for the { t kira} code, improving parallel performance for large N-body simulations across different hardware setups.

Findings

Code scales well from 1024 to 65536 stars on 1-128 processors.

Speedup is satisfactory on high-speed interconnects and NUMA architectures.

Grid computing becomes inefficient beyond 60 processors across three sites.

Abstract

We describe source code level parallelization for the {\tt kira} direct gravitational $N$-body integrator, the workhorse of the {\tt starlab} production environment for simulating dense stellar systems. The parallelization strategy, called ``j-parallelization'', involves the partition of the computational domain by distributing all particles in the system among the available processors. Partial forces on the particles to be advanced are calculated in parallel by their parent processors, and are then summed in a final global operation. Once total forces are obtained, the computing elements proceed to the computation of their particle trajectories. We report the results of timing measurements on four different parallel computers, and compare them with theoretical predictions. The computers employ either a high-speed interconnect, a NUMA architecture to minimize the communication overhead or are distributed in a grid. The code scales well in the domain tested, which ranges from 1024 - 65536 stars on 1 - 128 processors, providing satisfactory speedup. Running the production environment on a grid becomes inefficient for more than 60 processors distributed across three sites.