High Performance Computing in Science and Engineering: 4th International Conference, HPCSE 2019, Karolinka, Czech Republic, May 20–23, 2019, Revised ... (Lecture Notes in Computer Science, 12456)

Preface
Organization
Contents
Thermal Characterization of a Tier0 Datacenter Room in Normal and Thermal Emergency Conditions
	1 Introduction and Related Works
	2 Methodology
	3 Experimental Results
		3.1 Spatial Study
		3.2 Temporal Study
		3.3 Thermal Emergency
	4 Conclusion
	References
Towards Local-Failure Local-Recovery in PDE Frameworks: The Case of Linear Solvers
	1 Introduction
		1.1 Motivation
		1.2 Contribution
		1.3 Related Work
	2 Preliminaries
		2.1 DUNE – The Distributed Unified Numerics Environment
		2.2 Restoring MPI Communicators with ULFM and Distributed Exceptions
		2.3 Data-Driven Compression with the SZ Library
	3 Preparing Iterative Linear Solvers for LFLR
		3.1 Backup Creation and Compression
		3.2 Remote Storing of Backups
		3.3 Recovery Strategies
		3.4 Recovery of Iterative Solvers
	4 Implementation for Iterative Solvers in DUNE
		4.1 Framework Extensions
		4.2 Modifications for CG and GMRES(m)
	5 Numerical Examples
		5.1 Small-Scale Viability Tests
		5.2 Overhead Quantification at Scale
	6 Conclusions
	References
Complexity Analysis of a Fast Directional Matrix-Vector Multiplication
	1 Introduction
	2 Derivation of the Fast Directional Method
		2.1 Box Cluster Trees
		2.2 A Directional Kernel Approximation
		2.3 Partitioning of the Matrix
		2.4 Choice of Directions
		2.5 Transfer Operations
		2.6 Main Algorithm
		2.7 Implementation Details
	3 Complexity Analysis
	4 Numerical Examples
	References
Fast Large-Scale Boundary Element Algorithms
	1 Introduction
	2 H2-Matrices
	3 Hybrid Cross Approximation
	4 Green Cross Approximation
	5 Numerical Experiments
	References
Solving Large-Scale Interior Eigenvalue Problems to Investigate the Vibrational Properties of the Boson Peak Regime in Amorphous Materials
	1 Introduction
	2 Numerical Solution Procedures
		2.1 Spectral Projector
		2.2 Chebyshev Polynomial Expansions
		2.3 Dealing with the Gibbs Phenomenon
		2.4 Counting the Eigenvalues in an Interval
		2.5 Computing a Basis of R(p[,])
	3 Implementation
	4 Numerical Experiments
	5 Physics Results
	6 Conclusions
	References
Performance Evaluation of Pseudospectral Ultrasound Simulations on a Cluster of Xeon Phi Accelerators
	1 Introduction
	2 Local Fourier Basis Domain Decomposition
	3 Target Architecture
	4 Implementation
		4.1 Execution Mode
		4.2 Fast Fourier Transforms
		4.3 Overlap Exchanges
		4.4 Parallel Input and Output
	5 Scaling Results
		5.1 Overview
		5.2 Performance Scaling on Large Domains
		5.3 Performance Scaling on Small Domains
		5.4 Simulation Time Breakdown
	6 Platform Investigation
		6.1 Overview
		6.2 Performance of 3D FFTs on Intel Xeon Phi
		6.3 Performance of Intel MPI on Intel Xeon Phi
	7 Conclusion
	References
Estimation of Execution Parameters for k-Wave Simulations
	1 Introduction
	2 Automatic Offloading of k-Wave Workflows
	3 Optimization of Workflow Execution Parameters
		3.1 Single-Pass Optimization
		3.2 Interpolation Heuristics
		3.3 k-Wave Workflow Properties
		3.4 Typical Problems of Performance Data Interpolations
	4 Experimental Results
		4.1 Comparison of Interpolation Techniques for Known Simulation
		4.2 Comparison of Interpolation Techniques for Unknown Simulations
	5 Conclusions
		5.1 Future Work
	References
Analysis and Visualization of the Dynamic Behavior of HPC Applications
	1 Introduction
	2 Understanding Application Behavior
		2.1 Charts, Plots and Heatmaps
		2.2 Tables
	3 Conclusion
	References
A Convenient Graph Connectedness for Digital Imagery
	1 Introduction
	2 Preliminaries
	3 8-Adjacency Graph with a Set of Paths of Length 2
	4 Conclusions
	References
Author Index