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

Title: AMRSim: an object-oriented performance simulator for parallel adaptive mesh refinement

Abstract

Adaptive mesh refinement is complicated by both the algorithms and the dynamic nature of the computations. In parallel the complexity of getting good performance is dependent upon the architecture and the application. Most attempts to address the complexity of AMR have lead to the development of library solutions, most have developed object-oriented libraries or frameworks. All attempts to date have made numerous and sometimes conflicting assumptions which make the evaluation of performance of AMR across different applications and architectures difficult or impracticable. The evaluation of different approaches can alternatively be accomplished through simulation of the different AMR processes. In this paper we outline our research work to simulate the processing of adaptive mesh refinement grids using a distributed array class library (P++). This paper presents a combined analytic and empirical approach, since details of the algorithms can be readily predicted (separated into specific phases), while the performance associated with the dynamic behavior must be studied empirically. The result, AMRSim, provides a simple way to develop bounds on the expected performance of AMR calculations subject to constraints given by the algorithms, frameworks, and architecture.

Authors:
; ; ;
Publication Date:
Research Org.:
Lawrence Livermore National Lab. (LLNL), Livermore, CA (United States)
Sponsoring Org.:
US Department of Energy (US)
OSTI Identifier:
15005479
Report Number(s):
UCRL-JC-141940
TRN: US200322%%463
DOE Contract Number:  
W-7405-ENG-48
Resource Type:
Conference
Resource Relation:
Conference: Joint Association for Computing Machinery Java Grande - International Scientific Computing in Object-Oriented Parallel Environments Conference, Palo Alto, CA (US), 06/02/2001--06/04/2001; Other Information: PBD: 8 Jan 2001
Country of Publication:
United States
Language:
English
Subject:
99 GENERAL AND MISCELLANEOUS//MATHEMATICS, COMPUTING, AND INFORMATION SCIENCE; ALGORITHMS; ARCHITECTURE; EVALUATION; JAVA; MACHINERY; PERFORMANCE; PROCESSING; SIMULATION; SIMULATORS

Citation Formats

Miller, B, Philip, B, Quinlan, D, and Wissink, A. AMRSim: an object-oriented performance simulator for parallel adaptive mesh refinement. United States: N. p., 2001. Web.
Miller, B, Philip, B, Quinlan, D, & Wissink, A. AMRSim: an object-oriented performance simulator for parallel adaptive mesh refinement. United States.
Miller, B, Philip, B, Quinlan, D, and Wissink, A. 2001. "AMRSim: an object-oriented performance simulator for parallel adaptive mesh refinement". United States. https://www.osti.gov/servlets/purl/15005479.
@article{osti_15005479,
title = {AMRSim: an object-oriented performance simulator for parallel adaptive mesh refinement},
author = {Miller, B and Philip, B and Quinlan, D and Wissink, A},
abstractNote = {Adaptive mesh refinement is complicated by both the algorithms and the dynamic nature of the computations. In parallel the complexity of getting good performance is dependent upon the architecture and the application. Most attempts to address the complexity of AMR have lead to the development of library solutions, most have developed object-oriented libraries or frameworks. All attempts to date have made numerous and sometimes conflicting assumptions which make the evaluation of performance of AMR across different applications and architectures difficult or impracticable. The evaluation of different approaches can alternatively be accomplished through simulation of the different AMR processes. In this paper we outline our research work to simulate the processing of adaptive mesh refinement grids using a distributed array class library (P++). This paper presents a combined analytic and empirical approach, since details of the algorithms can be readily predicted (separated into specific phases), while the performance associated with the dynamic behavior must be studied empirically. The result, AMRSim, provides a simple way to develop bounds on the expected performance of AMR calculations subject to constraints given by the algorithms, frameworks, and architecture.},
doi = {},
url = {https://www.osti.gov/biblio/15005479}, journal = {},
number = ,
volume = ,
place = {United States},
year = {Mon Jan 08 00:00:00 EST 2001},
month = {Mon Jan 08 00:00:00 EST 2001}
}

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: