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

Title: Performance of a direct numerical simulation solver for turbulent combustion on the Cray XT3/4

Abstract

Parallel performance of a flow solver, S3D, for direct numerical simulation of turbulent combustion is evaluated on the latest Cray XT system installed at the National Center for Computational Sciences (NCCS). S3D is a massively parallel code that uses MPI for message passing parallelism. It has been ported to various platforms at the Office of Science computational facilities and scales very well to a large fraction of the system. In the past, S3D showed near perfect parallel scaling in the performance studies conducted on the Cray XT3 system. However, S3D was found to exhibit higher parallel communication overheads on the Cray XT system at NCCS following a recent upgrade. This article presents the results of the scaling study and an analysis of the parallel and dual-core performance.

Authors:
 [1];  [1];  [2]
  1. ORNL
  2. Sandia National Laboratories (SNL)
Publication Date:
Research Org.:
Oak Ridge National Lab. (ORNL), Oak Ridge, TN (United States); Center for Computational Sciences
Sponsoring Org.:
USDOE Office of Science (SC)
OSTI Identifier:
966081
DOE Contract Number:
DE-AC05-00OR22725
Resource Type:
Conference
Resource Relation:
Conference: Cray User Group, Seattle, WA, USA, 20070507, 20070510
Country of Publication:
United States
Language:
English
Subject:
37 INORGANIC, ORGANIC, PHYSICAL AND ANALYTICAL CHEMISTRY; 97 MATHEMATICS AND COMPUTING; COMBUSTION KINETICS; PERFORMANCE; NUMERICAL SOLUTION; PARALLEL PROCESSING; TURBULENT FLOW; CRAY COMPUTERS

Citation Formats

Sankaran, Ramanan, Fahey, Mark R, and Chen, Jacqueline H. Performance of a direct numerical simulation solver for turbulent combustion on the Cray XT3/4. United States: N. p., 2007. Web.
Sankaran, Ramanan, Fahey, Mark R, & Chen, Jacqueline H. Performance of a direct numerical simulation solver for turbulent combustion on the Cray XT3/4. United States.
Sankaran, Ramanan, Fahey, Mark R, and Chen, Jacqueline H. Mon . "Performance of a direct numerical simulation solver for turbulent combustion on the Cray XT3/4". United States. doi:.
@article{osti_966081,
title = {Performance of a direct numerical simulation solver for turbulent combustion on the Cray XT3/4},
author = {Sankaran, Ramanan and Fahey, Mark R and Chen, Jacqueline H},
abstractNote = {Parallel performance of a flow solver, S3D, for direct numerical simulation of turbulent combustion is evaluated on the latest Cray XT system installed at the National Center for Computational Sciences (NCCS). S3D is a massively parallel code that uses MPI for message passing parallelism. It has been ported to various platforms at the Office of Science computational facilities and scales very well to a large fraction of the system. In the past, S3D showed near perfect parallel scaling in the performance studies conducted on the Cray XT3 system. However, S3D was found to exhibit higher parallel communication overheads on the Cray XT system at NCCS following a recent upgrade. This article presents the results of the scaling study and an analysis of the parallel and dual-core performance.},
doi = {},
journal = {},
number = ,
volume = ,
place = {United States},
year = {Mon Jan 01 00:00:00 EST 2007},
month = {Mon Jan 01 00:00:00 EST 2007}
}

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:
  • This paper will present an overview of the current status of I/O on the Cray XT line of supercomputers and provide guidance to application developers and users for achieving efficient I/O. A large amount of I/O benchmark results will be presented, motivated by projected I/O requirements for some widely-used scientific applications in the DOE. Finally, the authors will interpret and summarize these benchmark results to give forward-looking guidance for I/O in large-scale application runs on a Cray XT3/XT4.
  • The Community Atmosphere Model (CAM) is the atmospheric component of the Community Climate System Model (CCSM) and is the primary consumer of computer resources in typical CCSM simulations. We analyze the performance of CAM on the Cray XT3 and Cray X1E systems at Oak Ridge National Laboratory, describing the performance sensitivities of the two systems to the numerous tuning options available in CAM. We then compare performance on the Cray systems with that on a number of other systems, including the Earth Simulator, an IBM POWER4 cluster, and an IBM POWER5 cluster.
  • Graphical processing units (GPUs), characterized by significant computing performance, are nowadays very appealing for the solution of computationally demanding tasks in a wide variety of scientific applications. However, to run on GPUs, existing codes need to be ported and optimized, a procedure which is not yet standardized and may require non trivial efforts, even to high-performance computing specialists. In the present paper we accurately describe the porting to CUDA (Compute Unified Device Architecture) of a finite-difference compressible NavierÔÇôStokes solver, suitable for direct numerical simulation (DNS) of turbulent flows. Porting and validation processes are illustrated in detail, with emphasis on computationalmore » strategies and techniques that can be applied to overcome typical bottlenecks arising from the porting of common computational fluid dynamics solvers. We demonstrate that a careful optimization work is crucial to get the highest performance from GPU accelerators. The results show that the overall speedup of one NVIDIA Tesla S2070 GPU is approximately 22 compared with one AMD Opteron 2352 Barcelona chip and 11 compared with one Intel Xeon X5650 Westmere core. The potential of GPU devices in the simulation of unsteady three-dimensional turbulent flows is proved by performing a DNS of a spatially evolving compressible mixing layer.« less
  • In recent years multidimensional numerical models have been developed to predict reacting flows in a wide variety of combustion systems. These models must be validated by comparing them to experimental results. This paper presents three numerical simulations using the KIVA and CONCHAS-SPRAY hydrodynamics programs. The KIVA program produces numerical solutions of the full nonlinear, transient, compressible flow, Navier-Stokes equations in two and three dimensions, including moving boundaries, chemical reactions, and turbulence models. The numerical algorithm is based on the arbitrary Lagrangian-Eulerian (ALE) technique. CONCHAS-SPRAY is the two-dimensional ICED-ALE program that was extended to produce KIVA.