Parallel Computational Technologies: 15th International Conference, PCT 2021, Volgograd, Russia, March 30 – April 1, 2021, Revised Selected Papers (Communications in Computer and Information Science)

This book constitutes refereed proceedings of the 15th International Conference on Parallel Computational Technologies, PCT 2021, held in March-April 2021. Due to the COVID-19 pandemic the conference was held online. The 22 revised full papers presented were carefully reviewed and selected from 89 submissions. The papers are organized in topical sections on high performance architectures, tools and technologies; parallel numerical algorithms; supercomputer simulation. Preface Organization Contents High Performance Architectures, Tools and Technologies Equivalent Transformations of Some Kinds of Computing Structures of Non-linear Recurrent Expressions for Reconfigurable Computing Systems 1 Introduction 2 Transformations of Quadratic Computing Structures 3 Transformations of Conditional Computing Structures 4 Equivalent Transformations of Direct and Inverse Operational Nodes 5 Transformations of Fractional Computing Structures 6 Conclusions References Development and Practical Application of Methods for Detecting Similar Supercomputer Jobs 1 Introduction 2 Background and Related Work 3 Static Analysis Method for Package Detection 3.1 Description of the Static Method 3.2 Solving the Problem of Package Detection 4 Dynamic Analysis Method for Package Detection 4.1 Description of the Dynamic Method 4.2 Solving the Problem of Package Detection 5 Conclusions and Future Work References Detecting Changes in Communication Properties of Parallel Programs by InfiniBand Traffic Analysis 1 Introduction 2 Related Work 3 Basics of InfiniBand Internals and Analysis Difficulties of MPI over InfiniBand 4 InfiniBand Traffic Collection 5 The Proposed Approach 5.1 Architecture Overview 5.2 Analysis Features 5.3 Communication Matrices 5.4 Communication Graphs 5.5 Time Step 5.6 Analysis of the Communication Matrices 5.7 Detecting Behavior Changes 6 Experiments 6.1 Individual MPI Operations 6.2 NPB Tests 7 Conclusions and Future Work References Transformation of Graphs with Associative Operations in Terms of the Set@l Programming Language 1 Introduction 2 Transformation of Graphs with Associative Operations for Efficient Performance Reduction 3 Description of ``Head/Tail'' and ``Half-Splitting'' Attributes in the Set@l Programming Language 4 Development of Resource-Independent Programs in the Set@l Language 5 Conclusions References Parallel Numerical Algorithms GPU-Accelerated Fuzzy Inference Based on Fuzzy Truth Values 1 Introduction 2 Fuzzy Inference Based on Fuzzy Truth Values 3 Implementation of Individual Operations of Fuzzy Inference 3.1 Overview of the Inference Process 3.2 Global Memory Usage 3.3 Computing Fuzzy Truth Values 3.4 Reducing Fuzzy Truth Values 3.5 Defuzzification 4 Multi-block Implementation 4.1 General Remarks 4.2 Recursive Parallel Reduction and CUDA Streams 4.3 Breadth-First Reduction 5 Conclusions References Deflation of Periodic Orbits in Large-Scale Systems: Algorithm and Parallel Implementation 1 Introduction 2 Poincaré Shooting Method and Preconditioning 3 Deflation 4 The Parallelization 5 Results 6 Conclusions References Parallel Computations for Solving Multicriteria Mixed-Integer Optimization Problems 1 Introduction 2 Problems of Multicriteria Mixed-Integer Optimization 3 The Approach: Unrolling Mixed-Integer Optimization Problems and Dimensionality Reduction 3.1 Simultaneous Solution of Mixed-Integer Optimization Problems 3.2 Dimensionality Reduction of Mixed-Integer Optimization Problems 4 Parallel Computation for Solving Mixed-Integer Optimization Problems 5 Results of Numerical Experiments 6 Conclusions References Parallel Intelligent Computing in Algebraic Problems 1 Introduction 2 The Problem of Classification of Algebraic Tasks 3 Optimization of the Selection of Algorithms 4 Intelligent Mathematical Technologies 5 Conclusions References CUDA Implementation of an Algorithm for Batch Mode Detection of Collisions 1 Introduction 2 Problem Formulation 2.1 Terminology 2.2 Request Sources and Batch Mode Discussion 3 The CUDA Implementation 4 The Test Setup 5 Acceleration Results 5.1 Secondary Parameters Dependency 5.2 Primary Parameters Dependency 6 Conclusions References Functionally Arranged Data for Algorithms with Space-Time Wavefront 1 Introduction 2 Methods 2.1 LRnLA Algorithms 2.2 ConeTorre 2.3 Data Exchange Between Tasks 3 Implementation Example 3.1 Lattice Boltzmann Method 3.2 CUDA GPU Implementation 3.3 TorreFold 4 Results 5 Conclusions References Exploring the Limits of Problem-Specific Adaptations of SAT Solvers in SAT-Based Cryptanalysis 1 Introduction 2 Preliminaries 2.1 On CDCL SAT Solvers 2.2 On SAT-Based Cryptanalysis 3 Adapting SAT Solvers to Specific Instances 3.1 Choosing the Right Solver 3.2 Choosing the Best Encoding 3.3 Parameterization 3.4 Preprocessing 3.5 Adapting the rokk Solver to Grain_v1 Cryptanalysis 4 Improving the SAT-Based Guess-and-Determine Attack on Bivium 5 Improving the SAT-Based Guess-and-Determine Attack on Grain_v1 6 Discussion and Related Work 7 Conclusions References FRaGenLP: A Generator of Random Linear Programming Problems for Cluster Computing Systems 1 Introduction 2 Method for Generating Random LP Problems 3 Parallel Algorithm for Generating Random LP Problems 4 Software Implementation and the Computational Experiments 5 Conclusions References Application of the AmgX Library to the Discontinuous Galerkin Methods for Elliptic Problems 1 Introduction 2 Problem Formulation 3 Properties of the AmgX Library 4 Results 5 Discussion References Conjugate Direction Methods for Parallel Deflation 1 Introduction 2 Two-Level Methods of Conjugate and Semiconjugate Directions 3 Deflation and Augmentation Algorithms in Krylov Subspaces 4 Estimates of Parallelization Efficiency 5 Examples of Numerical Experiments 6 Conclusions References Supercomputer Simulation Research of the Mechanism of External Hormonal Regulation of the Development of Phyto- and Zooplankton Populations Using Supercomputer Technologies 1 Introduction 2 The Problem Statement 3 The Method of Solution of the Problem 4 Parallel Implementation of the Biological Kinetics Problem 5 Results of Numerical Experiments on the Research of Allelopathy of Phytoplankton Populations 6 Conclusions References Luthien: A Parallel PIC Code for Modeling the Interaction of Focused Electron Beams with Plasma 1 Introduction 2 The Mathematical Model and Its Implementation 2.1 Features of the Axial-Cylindrical Geometry 2.2 Focusing the Beams 2.3 Boundary Conditions 3 Software Implementation and Code Structure 3.1 Simulation Space 3.2 Performance of the Code 4 Conclusions References Simulation of Heat and Mass Transfer in Open Geothermal Systems: A Parallel Implementation 1 Introduction 2 Problem Statement and Mathematical Model 3 Simulation Results 4 Parallel Implementation for Graphics Processors 5 Conclusions References Supercomputer-Based Simulation of the Hydrodynamics of River Mouth Areas 1 Introduction 2 The Problem Statement 3 Physical Processes Splitting Schemes for Solving Hydrodynamic Problems 4 Method for the Solution of Grid Equations 5 The Software Implementation 6 Results of the Numerical Experiments 7 Conclusions References System for the Visualization of Meshes of Big Size Obtained from Gas-Dynamic Simulations 1 Introduction 2 Visualization System Architecture 3 Parallel Rendering Implementation 4 Tests and Discussion 5 Conclusions References Finite Element Simulation of the Influence of the Bending Roll Profile on the Stress-Strain State of Billets for Longitudinal-Welded Rolled Pipes 1 Production Schemes for Longitudinal-Welded Pipes 2 A Finite Element Model of Bending Roll Profile 3 The Parameters of the Finite Element Model 4 Experimental Results of the Finite Element Simulation 5 Conclusions References Digital Twins of Geological Objects: Development and Use 1 Introduction and Motivation 2 Construction of a Digital Twin for a Geological Object 3 3D Multiscale Numerical Simulation of Seismic Wavefields 3.1 The Mathematical Statement 3.2 Model Parameters and Acquisition 3.3 The Numerical Method 4 The Parallel Implementation 4.1 Scalability and Efficiency 5 Conclusions References A Parallel Algorithm to Compute the Transport of Suspended Particles Based on 3D Models 1 Introduction 2 Model of Transport of Suspended Particles 3 Approximation of the Problem of Transport of Suspended Particles 4 Parallel Algorithm for the Problem of Transport of Suspended Particles 5 Results of Numerical Experiments with the Model of Transport of Suspended Particles 6 Conclusions References Author Index