Search arXiv⌕ Search
Home/Archive/arXiv

arXiv · physics/9709003

A Parallel Tree code for large Nbody simulation: dynamic load balance and data distribution on CRAY T3D system

U. Becciani·R. Ansaloni·V. Antonuccio-Delogu·G. Erbacci·M. Gambera·A. Pagliaro·-
View original source ↗Read source document ↗Export metadata ↓

Abstract

N-body algorithms for long-range unscreened interactions like gravity belong to a class of highly irregular problems whose optimal solution is a challenging task for present-day massively parallel computers. In this paper we describe a strategy for optimal memory and work distribution which we have applied to our parallel implementation of the Barnes & Hut (1986) recursive tree scheme on a Cray T3D using the CRAFT programming environment. We have performed a series of tests to find an " optimal data distribution " in the T3D memory, and to identify a strategy for the " Dynamic Load Balance " in order to obtain good performances when running large simulations (more than 10 million particles). The results of tests show that the step duration depends on two main factors: the data locality and the T3D network contention. Increasing data locality we are able to minimize the step duration if the closest bodies (direct interaction) tend to be located in the same PE local memory (contiguous block subdivison, high granularity), whereas the tree properties have a fine grain distribution. In a very large simulation, due to network contention, an unbalanced load arises. To remedy this we have devised an automatic work redistribution mechanism which provided a good Dynamic Load Balance at the price of an insignificant overhead.

Explore related subjects

physics.comp-phastro-ph

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

U. Becciani, R. Ansaloni, V. Antonuccio-Delogu, G. Erbacci, M. Gambera, A. Pagliaro, -. 1997-09-02. A Parallel Tree code for large Nbody simulation: dynamic load balance and data distribution on CRAY T3D system. https://doi.org/10.1016/s0010-4655(97)00102-1

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

A componentwise spherical Dirichlet-to-Neumann closure for finite-volume MHD at finite magnetic Reynolds number

At finite magnetic Reynolds number, the field induced by a moving conducting fluid or solid extends into open space, so a finite mesh needs an outer boundary condition that represents the decaying exterior field. Magnetohydrodynamic (MHD) codes widely use pseudo-vacuum or truncation at a finite radius. Neither condition is physically correct for a vacuum exterior, but the exact alternatives are costly or limited to simple geometries. We present a Dirichlet-to-Neumann (DtN) closure on a fixed enclosing sphere and release it as part of the open-source multi-region finite-volume solver FreeMHD. The closure matches each Cartesian component of the vector potential to its decaying exterior solution, which retains all three decaying vector-Laplace families. A meshed vacuum buffer connects non-spherical conductors to the sphere. We verify the closure with manufactured modes, sphere free decay, sphere and oblate dynamos, and a freely braking rotor. Under refinement, oblate dynamo thresholds and sphere growth rates converge toward independent spectral references, and the rotor torque matches an exact transient reference. On the same sphere, pseudo-vacuum and truncation give the larger errors predicted by their exterior models.

physics.comp-ph↗
arXiv2026

Integration of Silica in G4CMP for Phonon Simulations: Framework and Tools for Material Integration

Superconducting detectors with sub-eV energy resolution have demonstrated success setting limits on Beyond the Standard Model (BSM) physics due to their unique sensitivity to low-energy events. G4CMP, a Geant4-based extension for condensed matter physics, provides a comprehensive toolkit for modeling phonon and charge dynamics in cryogenic materials. This paper introduces a technical formalism to support the superconducting qubit and low-threshold detector community in implementing phonon simulations in custom materials into the G4CMP. As a case study, we present the results of a detailed analysis of silica phonon transport properties relevant for simulating substrate backgrounds in Beryllium Electron capture in Superconducting Tunnel junctions (BeEST)-style experiments using G4CMP. Additionally, Python-based tools were developed to aid users in implementing their own materials and are available on the G4CMP repository.

physics.comp-ph↗
arXiv2026

Discovering like a scientist: Literature-informed dimensionless learning for physical closure of data-driven equations

Data-driven methods can recover compact equations, yet the physical dependence of their coefficients across operating conditions often remains unresolved. Dimensionless groups can explain this variation, but selecting relevant groups from the many dimensionally admissible combinations remains challenging. Here, we introduce a knowledge-guided framework for dimensionless learning, called π-former. It learns patterns of variable association and exponent assignment from a corpus of 1,288 established dimensionless groups. Under exact dimensional constraints, the model proposes interpretable basic groups whose combinations are optimized against observations. Across seven physical problems, the framework identifies predictive dimensionless coordinates within seconds. Integration with equation discovery further expresses condition-dependent coefficients through these coordinates, recovering established physical scalings in multi-condition simulations. Applied to experimental measurements of evaporation-driven wetting, the framework discovers dimensionless coordinates and uses them to construct a unified trajectory equation. The resulting model generalizes to held-out conditions across liquids and initial volumes, outperforming existing empirical models. These results establish dimensionless learning as a transferable bridge between accumulated scientific knowledge and data-driven equation discovery, enabling artificial intelligence to uncover physically meaningful laws from observations.

physics.comp-ph↗