skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: An improved bi-level algorithm for partitioning dynamic structured grid hierarchies.

Abstract

Structured adaptive mesh refinement methods are being widely used for computer simulations of various physical phenomena. Parallel implementations potentially offer realistic simulations of complex three-dimensional applications. But achieving good scalability for large-scale applications is non-trivial. Performance is limited by the partitioner's ability to efficiently use the underlying parallel computer's resources. Designed on sound SAMR principles, Nature+Fable is a hybrid, dedicated SAMR partitioning tool that brings together the advantages of both domain-based and patch-based techniques while avoiding their drawbacks. But the original bi-level partitioning approach in Nature+Fable is insufficient as it for realistic applications regards frequently occurring bi-levels as 'impossible' and fails. This document describes an improved bi-level partitioning algorithm that successfully copes with all possible hi-levels. The improved algorithm uses the original approach side-by-side with a new, complementing approach. By using a new, customized classification method, the improved algorithm switches automatically between the two approaches. This document describes the algorithms, discusses implementation issues, and presents experimental results. The improved version of Nature+Fable was found to be able to handle realistic applications and also to generate less imbalances, similar box count, but more communication as compared to the native, domain-based partitioner in the SAMR framework AMROC.

Authors:
; ;
Publication Date:
Research Org.:
Sandia National Laboratories
Sponsoring Org.:
USDOE
OSTI Identifier:
947255
Report Number(s):
SAND2006-0966C
TRN: US200909%%140
DOE Contract Number:
AC04-94AL85000
Resource Type:
Conference
Resource Relation:
Conference: Proposed for presentation at the SIAM Parallel processing for scientific computing 2006 held February 22-24, 2006 in San Francisco, CA.
Country of Publication:
United States
Language:
English
Subject:
99 GENERAL AND MISCELLANEOUS//MATHEMATICS, COMPUTING, AND INFORMATION SCIENCE; ALGORITHMS; CLASSIFICATION; COMMUNICATIONS; COMPUTERIZED SIMULATION; IMPLEMENTATION; PARALLEL PROCESSING; PERFORMANCE; SWITCHES; COMPUTERS

Citation Formats

Deiterding, Ralf, Steensland, Johan, and Ray, Jaideep. An improved bi-level algorithm for partitioning dynamic structured grid hierarchies.. United States: N. p., 2006. Web.
Deiterding, Ralf, Steensland, Johan, & Ray, Jaideep. An improved bi-level algorithm for partitioning dynamic structured grid hierarchies.. United States.
Deiterding, Ralf, Steensland, Johan, and Ray, Jaideep. Wed . "An improved bi-level algorithm for partitioning dynamic structured grid hierarchies.". United States. doi:.
@article{osti_947255,
title = {An improved bi-level algorithm for partitioning dynamic structured grid hierarchies.},
author = {Deiterding, Ralf and Steensland, Johan and Ray, Jaideep},
abstractNote = {Structured adaptive mesh refinement methods are being widely used for computer simulations of various physical phenomena. Parallel implementations potentially offer realistic simulations of complex three-dimensional applications. But achieving good scalability for large-scale applications is non-trivial. Performance is limited by the partitioner's ability to efficiently use the underlying parallel computer's resources. Designed on sound SAMR principles, Nature+Fable is a hybrid, dedicated SAMR partitioning tool that brings together the advantages of both domain-based and patch-based techniques while avoiding their drawbacks. But the original bi-level partitioning approach in Nature+Fable is insufficient as it for realistic applications regards frequently occurring bi-levels as 'impossible' and fails. This document describes an improved bi-level partitioning algorithm that successfully copes with all possible hi-levels. The improved algorithm uses the original approach side-by-side with a new, complementing approach. By using a new, customized classification method, the improved algorithm switches automatically between the two approaches. This document describes the algorithms, discusses implementation issues, and presents experimental results. The improved version of Nature+Fable was found to be able to handle realistic applications and also to generate less imbalances, similar box count, but more communication as compared to the native, domain-based partitioner in the SAMR framework AMROC.},
doi = {},
journal = {},
number = ,
volume = ,
place = {United States},
year = {Wed Feb 01 00:00:00 EST 2006},
month = {Wed Feb 01 00:00:00 EST 2006}
}

Conference:
Other availability
Please see Document Availability for additional information on obtaining the full-text document. Library patrons may search WorldCat to identify libraries that hold this conference proceeding.

Save / Share:
  • Structured adaptive mesh refinement methods are being widely used for computer simulations of various physical phenomena. Parallel implementations potentially offer realistic simulations of complex three-dimensional applications. But achieving good scalability for large-scale applications is non-trivial. Performance is limited by the partitioner's ability to efficiently use the underlying parallel computer's resources. Designed on sound SAMR principles, Nature+Fable is a hybrid, dedicated SAMR partitioning tool that brings together the advantages of both domain-based and patch-based techniques while avoiding their drawbacks. But the original bi-level partitioning approach in Nature+Fable is insufficient as it for realistic applications regards frequently occurring bi-levels as ''impossible'' andmore » fails. This document describes an improved bi-level partitioning algorithm that successfully copes with all possible bi-levels. The improved algorithm uses the original approach side-by-side with a new, complementing approach. By using a new, customized classification method, the improved algorithm switches automatically between the two approaches. This document describes the algorithms, discusses implementation issues, and presents experimental results. The improved version of Nature+Fable was found to be able to handle realistic applications and also to generate less imbalances, similar box count, but more communication as compared to the native, domain-based partitioner in the SAMR framework AMROC.« less
  • This paper presents an efficient block detection algorithm that can convert a complex wire frame geometry into a set of oriented multi-block structured faces or blocks. The algorithm uses a recursive depth first search in conjunction with the Chid Topology Model (GTM) data structure. The GTM data structure is a Boundary Representation (B-Rep) radial edge non-manifold solid modeling data structure. The block detection algorithm abstracts the user from having to manually specify orientation and connectivity information traditionally needed in other grid generation systems.
  • This study presents a bi-directional multi-level power electronic interface for the grid interactions of plug-in hybrid electric vehicles (PHEVs) as well as a novel bi-directional power electronic converter for the combined operation of battery/ultracapacitor hybrid energy storage systems (ESS). The grid interface converter enables beneficial vehicle-to-grid (V2G) interactions in a high power quality and grid friendly manner; i.e, the grid interface converter ensures that all power delivered to/from grid has unity power factor and almost zero current harmonics. The power electronic converter that provides the combined operation of battery/ultra-capacitor system reduces the size and cost of the conventional ESS hybridizationmore » topologies while reducing the stress on the battery, prolonging the battery lifetime, and increasing the overall vehicle performance and efficiency. The combination of hybrid ESS is provided through an integrated magnetic structure that reduces the size and cost of the inductors of the ESS converters. Simulation and experimental results are included as prove of the concept presenting the different operation modes of the proposed converters.« less
  • Rhodium self-powered neutron detectors are utilized in many pressurized water reactors to determine the neutronic behaviour within the core. In order to compensate for the inherent time delay associated with the response of these detectors, a dynamic compensation algorithm is currently used in Combustion Engineering plants to reconstruct the dynamic flux signal which is being sensed by the rhodium detectors. This paper describes a new dynamic compensation algorithm, based on Kalman filtering, which improves on the noise gain and response time characteristics of the algorithm currently used, and offers the possibility of significantly expanding the capabilities of the present in-coremore » detector system without a major hardware investment.« less
  • Abstract not provided.