Sustained Simulation Performance 2017

Name: Sustained Simulation Performance 2017 | Proceedings of the Joint Workshop on Sustained Simulation Performance, University of Stuttgart (HLRS) and Tohoku University, 2017
Brand: Springer
Price: 96.29 EUR
Availability: OnlineOnly

Proceedings of the Joint Workshop on Sustained Simulation Performance, University of Stuttgart (HLRS) and Tohoku University, 2017

Michael M. Resch Wolfgang Bez Erich Focht Michael Gienger Hiroaki Kobayashi(Editor)

Springer (Publisher)

Published on 24. October 2017

VIII, 192 pages

E-Book

PDF with digital watermarking

System requirements

978-3-319-66896-3 (ISBN)

€96.29incl. 7% vat

System requirements

for PDF with digital watermarking

E-Book Single Licence

Available for download

Description

More details

Other editions

Persons

Content

Intro
Preface
Contents
Part I System Management
Theory and Practice of Efficient Supercomputer Management
1 Introduction
2 Moscow State University HPC Toolkit for Efficiency Analysis
2.1 DiMMon Monitoring System
2.2 JobDigest System for Application Behavior Analysis
2.3 Statistics Analysis Using OctoStat
2.4 OctoTron: Autonomous Life of Supercomputers
2.5 OctoScreen Visualization System
2.6 OctoShell System for Work Organization
3 Conclusion
References
Towards A Software Defined Secure Data Staging Mechanism
1 Introduction
2 Challenges and Issues
3 Key Technologies
3.1 Software Defined Networking
3.2 ExpEther Technology
3.3 Job Management System
4 Proposal
5 Conclusion
References
Part II Mathematical Methods and Approaches
The Numerical Approximation of Koopman Modes of a Nonlinear Operator Along a Trajectory
1 Introduction
2 The Koopman Operator
3 Trajectories and Observables
4 The Relation to Time Series Analysis
5 Approximation of an ?-Eigenmode Along a Trajectory
5.1 Determining the Polynom Coefficient Vector c
5.2 Roots and Pseudo-Eigenvectors
5.3 Handling Intermediate Data Steps
6 The ?-Eigenmode Mapping Operator
6.1 Incompressible Navier-Stokes Equations as Example
7 Remarks
8 Computational Costs and Performance Aspects
9 Application Example
10 Conclusions
References
Part III Optimisation and Vectorisation
Code Modernization Tools for Assisting Users in Migrating to Future Generations of Supercomputers
1 Introduction
2 General Process for Manual Code Modernization and Migration
3 Using IPT for Code Modernization (Parallelization)
4 Overview of KNL Processors
4.1 Multiple Memory Modes
4.2 Multiple Cluster Modes
5 Using ICAT for Code Modernization and Migration (Porting Code to KNL Processors)
6 Using IPT and ICAT with a Sample Application
6.1 Using IPT to Parallelize the MD Application
6.2 Using ICAT to Adapt the MD Application for KNL Processors
7 Conclusion
References
Vectorization of High-Order DG in Ateles for the NEC SX-ACE
1 Introduction
2 High-Order Discontinuous Galerkin in Ateles
2.1 The Modal Basis
2.2 The Mesh Structure
3 Vectorization on the NEC SX-ACE
3.1 Porting of Ateles
4 Measurements and Observations
4.1 Linear Equations
4.2 Nonlinear Equations
5 Summary and Outlook
References
Vectorization of Cellular Automaton-Based Labeling of 3-D Binary Lattices
1 Introduction
2 Related Work
3 Algorithm
3.1 Initialization
3.2 Cellular Automaton Update
3.2.1 Maximum Operator
3.2.2 Branch-Based Maximum Operator
3.2.3 Closed-Form Maximum Operator
4 Implementation
4.1 Dense Data Representation
4.2 OpenMP-Code
4.3 CUDA Implementation
4.3.1 Termination Criterion
4.4 Multi-GPU Computation
4.5 Sparse Data Representation
4.5.1 OpenMP-Code
5 Simulation Results
5.1 Maximum Operator
6 Conclusions
References
Part IV Computational Fluid Dynamics
Turbulence in a Fluid Stratified by a High Prandtl-Number Scalar
1 Introduction
2 Direct Numerical Simulations
3 Results
4 Conclusions
References
Wavelet-Based Compression of Volumetric CFD Data Sets
1 Introduction
2 Technical Description
2.1 Fixed Point Number Format
2.2 The Discrete Wavelet Transform
2.3 Quantization
3 Experimental Results
3.1 Data Sets
3.2 Results
4 Conclusions
References
Validation of Particle-Laden Large-Eddy SimulationUsing HPC Systems
1 Introduction
2 Mathematical Models
2.1 Navier-Stokes Equations
2.2 Particle Dynamics
2.2.1 Direct Particle-Fluid Simulation
2.2.2 Euler-Lagrange Model
3 Numerical Methods
3.1 Direct Particle-Fluid Simulation
3.2 Implicit Large-Eddy Simulation
3.3 Euler-Lagrange Model
4 Results and Discussion
4.1 Large-Eddy Simulation of Isotropic Turbulence
4.2 Turbulence Modulation by Particles
5 Conclusion
References
Coupled Simulation with Two Coupling Approacheson Parallel Systems
1 Introduction
2 Data Mapping Methods
2.1 Interpolation
2.2 Data Mapping by Evaluation
3 Results
3.1 Configuration of the Simulation
3.2 Coupled Simulation Results
3.3 Performance of the Mapping Methods
4 Conclusion
References
MRI-Based Computational Hemodynamics in Patients
1 Introduction
2 Methodology
2.1 Geometries
2.2 Meshing
2.3 CFD Setup
2.4 Optimization Workflow
3 Results
3.1 Optimization Results
3.2 Global Target Values
3.3 Local Variations
3.4 Performance Issues
4 Conclusion
Appendix
References
Part V High Performance Data Analytics
A Data Analytics Pipeline for Smart Healthcare Applications
1 Introduction
2 Healthcare Data Analytics Pipeline
2.1 Data Curation Phase
2.1.1 Data Type
2.1.2 Data Cleansing
2.1.3 Data Annotation
2.1.4 Data Integration
2.2 Data Analytics Phase
2.2.1 Analytics Methods
3 Smart Orthodontic Treatment in Dentistry
3.1 Assessment of Treatment Need (App 1)
3.2 Morphological Landmarking (App 2)
3.3 Casenote Generation (App 3)
4 Conclusion
References

System requirements

Save as PDF Copy link into clipboard

Schweitzer Fachinformationen

Sustained Simulation Performance 2017

Description

More details

Other editions

Additional editions

Persons

Content

System requirements