Parallel Programming Methods Research Articles

A Comparison of Ryu and Pox Controllers: A Parallel Implementation

Software Defined Network (SDN) network controllers have limitations in handling large volumes of data generated by switches, which can slow down their performance. Using parallel programming methods such as threading, multiprocessing, and MPI aims to improve the performance of the controller in handling a large number of switches. By considering factors such as memory usage, CPU consumption, and execution time. The test results show that although RYU outperforms POX in terms of faster execution time and lower CPU utilization rate, POX shows its prowess by exhibiting less memory usage despite higher CPU utilization rate than RYU. The use of the parallel approach proves advantageous as both controllers exhibit enhanced efficiency levels. Ultimately, RYU's impressive speed and superior resource optimization capabilities may prove to be more strategic than POX over time. Taking into account the specific needs and prerequisites of a given system, this research provides insights in selecting the most suitable controller to handle large-scale switches with optimal efficiency.

Risk assessment of computer network information using a proposed approach: Fuzzy hierarchical reasoning model based on scientific inversion parallel programming

Abstract Background Computer networks are involved in many fields such as business, education, marketing, government, and tourism in several forms. Technologies related to security protection and improvement of information integrity are used and developed for computer networks intruded on by unauthorized people and help save their confidentiality. Methods To improve the risk identification of computer networks, this manuscript combined a fuzzy hierarchical reasoning model with the scientific inversion parallel programming method to study the risk of computer networks. Moreover, this article defined and analyzed a d-order neighborhood message propagation algorithm. A d-order neighborhood parallel message propagation algorithm using the Gaussian graph model was proposed. Results The risk of the computer network was analyzed using the proposed method resulting in better protection effectiveness. Conclusion The simulations showed that the proposed algorithm could effectively detect risks and improve the security of the computer network.

Syntax Tree Development for Automated Serial-to-Parallel Code Translator for Multicore Processors

The emergence of multicore architectures has extremely stimulated the area of parallel computing. However, developing a parallel program and manually paralleling inherited sequential program codes are time-consuming work. The programmer should have good skills in using parallel programming methods. This fact determines the relevance of the subject of the research – the development of a serial-to-parallel code translator. The article gives a review of existing solutions in the chosen direction of research and considers their advantages and disadvantages. The principle of formation of a syntactic tree which is based on JSON format (the text format of data exchange based on JavaScript) is offered and an example of formation of a syntactic tree on the basis of this principle is considered. The result of the work is an approach for building a program platform for translating sequential code into parallel code. The distinctive feature of the developed platform is the web-service, which potentially allows you to extend the translator with other programming languages. The interaction with the programming environment is realized by means of REST-requests (HTTP-requests designed to call remote procedures). The developed software platform consists of three modules: the query processing module, which provides interaction with external systems through REST-requests; the tree building module, which forms a syntax tree on the basis of the source program code; the code conversion module, which obtains parallel program code on the basis of the syntax tree.

Multi-phase field model simulation based on MPI+OpenMP parallel: Evolution of seaweed and dendritic structure in directional solidification

A multi-phase model was established to imitate the growth of algal and dendritic grains during directional solidification. We studied the effects of temperature on the growth of bi-crystals and quantitatively analyzed the influence of anisotropic strength, thermal gradient, and pulling velocity on the evolution of bi-crystals. The results show that both weaker anisotropy strength and smaller pulling velocity can maintain the formation of seaweed tissue. The increase in the pulling velocity can degenerate the seaweed grains into dendrites and improve the growth rate of the dendrites, which make grain B produce more spindles, thereby accelerating the elimination of grain A. The thermal gradient is inversely proportional to the average initial spacing of dendrites. When the thermal gradient is too small, dendritic dendrites produce developed secondary dendrite arms, which, in turn, develop into tertiary dendrite arms to occupy the grain boundary, accelerating the elimination of seaweed grains. In addition, the multi-phase field model is solved by using central processing unit serial computation, single MPI (message passing interface) parallel programming method calculation, and MPI+OpenMP hybrid parallel programming structure, and the relevant factors affecting the efficiency of program operation are analyzed and tested. By comparing the computational efficiency of the three methods, it can be seen that the MPI+OpenMP hybrid parallel programming technology can make full use of computing resources in the case of large computing scale, further optimize the MPI parallel model, and obtain a higher acceleration ratio.

The system of predictive analysis of Bank investments using technologies of supercomputer simulation

The article discusses an approach to building a system for generating a bank investment prediction with the technology of parallel computing. We develop a predictive model of the price of securities for the banking business on the stock market. For predictive analysis of investments, the choice of a mathematical model was justified using the Monte Carlo method. Computer simulation was carried out to form a forecast of securities prices based on historical data on securities of Goldman Sachs. Estimates of the execution time of the forecasting algorithm by the parallel programming method are obtained. The conclusion is made about the applicability of the proposed model when using the method of algorithmic trading in the stock market. The technical requirements for the development of a predictive analytics system for the Bank's investment portfolio are formulated.

Representación en tiempo real de señales de radar empleando Odroid XU4

El sistema de representación en los radares modernos se realiza a partir de la obtención de señales digitales. En el caso de los radares de seguimiento la actualización de la representación se realiza en el orden de las décimas de milisegundos. Para lograr la actualización del display en tiempo real el proceso de adquisición, procesamiento de la señal y representación tiene que cumplir con la exigencia de tiempo de exploración del radar. El objetivo del trabajo es lograr la visualización de la información para un radar en tiempo real empleando la placa Odroid XU4. Se utilizó el método de programación paralela, mediante la creación de hilos en el Entorno de Desarrollo Integrado Qt Creator y el empleo del patrón de programación paralela segmentación de cauce. Esto permitió el uso eficiente de los recursos de cómputo, obteniéndose una disminución del tiempo de ejecución y una mayor aceleración con respecto a la variante secuencial.

Parallel Operation of Self‐Limited Analog Programming for Fast Array‐Level Weight Programming and Update

Memristive neural networks perform vector matrix multiplication efficiently, which is used for the accelerator of neuromorphic computing. To train the memristor cells in a memristive neural network, the analog conductance state of the memristor should be programmed in parallel; otherwise, the resulting long training time can limit the size of the neural network. Herein, a novel parallel programming method using the self‐limited analog switching behavior of the memristor is proposed. A Pt/Ti:NbOx/NbOx/TiN charge trap memristor device for the programming demonstration is utilized, and a convolutional neural network is emulated to train the MNIST dataset, based on the device characteristics. In the simulation, the proposed programming method is able to reduce programming time to as low as 1/130, compared with the sequential programming method. The simulation suggests that the programming time required by the proposed method is not affected by array size, which makes it very promising in a high‐density neural network.

Parallel computing in linear mixed models

In this study, we propose a parallel programming method for linear mixed models (LMM) generated from big data. A commonly used algorithm, expectation maximization (EM), is preferred for its use of maximum likelihood estimations, as the estimations are stable and simple. However, EM has a high computation cost. In our proposed method, we use a divide and recombine to split the data into smaller subsets, running the algorithm steps in parallel on multiple local cores and combining the results. The proposed method is used to fit LMM with dense and sparse parameters and for large number of observations. It is faster than the classical approach and generalizes for big data. Supplementary sources for the proposed method are available in the R package lmmpar.

A tale of lock-free agents: towards Software Transactional Memory in parallel Agent-Based Simulation

With the decline of Moore’s law and the ever increasing availability of cheap massively parallel hardware, it becomes more and more important to embrace parallel programming methods to implement Agent-Based Simulations (ABS). This has been acknowledged in the field a while ago and numerous research on distributed parallel ABS exists, focusing primarily on Parallel Discrete Event Simulation as the underlying mechanism. However, these concepts and tools are inherently difficult to master and apply and often an excess in case implementers simply want to parallelise their own, custom agent-based model implementation. However, with the established programming languages in the field, Python, Java and C++, it is not easy to address the complexities of parallel programming due to unrestricted side effects and the intricacies of low-level locking semantics. Therefore, in this paper we propose the use of a lock-free approach to parallel ABS using Software Transactional Memory (STM) in conjunction with the pure functional programming language Haskell, which in combination, removes some of the problems and complexities of parallel implementations in imperative approaches. We present two case studies, in which we compare the performance of lock-based and lock-free STM implementations in two different well known Agent-Based Models, where we investigate both the scaling performance under increasing number of CPU cores and the scaling performance under increasing number of agents. We show that the lock-free STM implementations consistently outperform the lock-based ones and scale much better to increasing number of CPU cores both on local hardware and on Amazon EC. Further, by utilizing the pure functional language Haskell we gain the benefits of immutable data and lack of unrestricted side effects guaranteed at compile-time, making validation easier and leading to increased confidence in the correctness of an implementation, something of fundamental importance and benefit in parallel programming in general and scientific computing like ABS in particular.

An OpenMP-Based Algorithm for Multi-Nodes Computational of Super-Resolution

A new common OpenMP based parallel programming method MPMC (multi-node paralleling model base on multiprocessor devices) is proposed and implemented for data separation based to accelerate Super-Resolution (SR) task. PanguOS, a common parallel programming system designed with MPMC, is deployed with Secure Shell (SSH) to control devices and Secure Copy (SCP) to transmit the data stream on Ubuntu 16.04, and it has a good performance for SR task with remote sensing images. Experiments with images from geostationary-orbit earth observing satellite GaoFen(GF)-4, the method proposed can achieve almost 2.95 times acceleration at PanguOS, deployed with 3 Jetson TX2s, than a single Jetson TX2.

A study on PGEP to evolve heuristic rules for FJSSP considering the total cost of energy consumption and weighted tardiness

Performance indicators such as makespan, flow time and tardiness are considered to be optimisation objectives in the traditional flexible job shop scheduling problem (FJSSP). However, the cost of energy consumption or environmental problems should not be ignored. This paper addresses the FJSSP by minimising the sum of the cost of energy consumption and the weighted tardiness. First, a mathematical model of the problem and a heuristic algorithm for the problem are presented. Second, a parallel gene expression programming (PGEP) method with a migration scheme is put forward to evolve rules for the proposed heuristic algorithm to solve the problem. To speed up the system learning process, a parallel and distributed computing framework is also designed. Finally, the performance of the proposed PGEP approach is evaluated through extensive simulations. The time-of-use electricity pricing, due date tightness and tardiness penalty weight are considered when evaluating the effect of the heuristic rules. Experimental results show that the proposed PGEP approach can significantly improve the quality of the heuristic rules, and the PGEP-evolved rules can fast and effectively solve FJSSP.

PARALLEL PROGRAMMING MODELS FOR CALCULATING THE PHYSICAL CHARACTERISTICS OF COMPOSITES

This article reviews the performance of supercomputers. The increase in computational performance is also provided through parallel programming, the development of which is represented by a set of high-level programming languages, the emergence of new architectures and distributed and cluster systems. The article presents a numerical simulation of the calculation of the physical characteristics of the composites using the OpenMX package. At the same time, the use of parallel programming methods led to a significant increase in computing performance.

NoT: a high-level no-threading parallel programming method for heterogeneous systems

Multithreading is the core of mainstream heterogeneous programming methods such as CUDA and OpenCL. However, multithreaded parallel programming requires programmers to handle low-level runtime details, making the programming process complex and error prone. This paper presents no-threading (NoT), a high-level no-threading programming method. It introduces the association structure, a new language construct, to provide a declarative runtime-free expression of different data parallelisms and avoid the use of multithreading. The NoT method designs C-like syntax for the association structure and implements a compiler and runtime system using OpenCL as an intermediate language. We demonstrate the effectiveness of our techniques with multiple benchmarks. The size of the NoT code is comparable to that of the serial code and is far less than that of the benchmark OpenCL code. The compiler generates efficient OpenCL code, yielding a performance competitive with or equivalent to that of the manually optimized benchmark OpenCL code on both a GPU platform and an MIC platform.

Application of Parallel Programming Methods for Simulating Flow Diversion Technologies on Hybrid Architecture Computers

Application of parallel programming methods for simulating the impact of polymer dispersed systems on oil reservoirs on a hybrid computer system that uses the central processor cores along with the graphics processing unit is discussed. The efficiency of the proposed approach for solving practical problems of simulating waterflooding of oil reservoirs using polymer dispersed systems on computers with hybrid architecture is demonstrated.

Analysis and possibility of modification of available algorithms for finding the optimal path on a square grid using parallel programming methods

Analysis and possibility of modification of available algorithms for finding the optimal path on a square grid using parallel programming methods

MapReduce functions to remote sensing distributed data processing—Global vegetation drought monitoring as example

SummaryGlobal change models for different applications are developed, according to the principle of remote sensing technology. Data for these models are generally remote sensing image, which is multiplatform, multidimentional, multiband, and multisource. Moreover, such data may be in different parts of the world and perhaps up to terabyte or petabyte level. Therefore, a data‐intensive computing problem in the global change has emerged. Distributed computing infrastructures are suitable to store large‐scale datalike satellite images that have to be written only once and read frequently. The emergence of the cloud computing technology brings new information architecture, and global change models implemented in the cloud platform provide users with stable, effective, on‐demand cloud computing services. In this paper, the experiment is carried out on the cloud framework based on open cloud computing platform‐Hadoop. In addition, on this framework, it achieves a prototype example for monitoring global vegetation drought conditions. Oriented to a variety of remote sensing data, we propose an abstract data format to achieve the unified description of remote sensing data. The data abstraction is to discretize the multidimensional remote sensing data for easy‐distributed storage and computation. Using MapReduce paradigm, the complexity of remote sensing algorithms is resolved. Experimental results show that based on the parallel programming method, good scalability changing with the amount of processed data in the Hadoop distributed environment.

Drainage network extraction from a high-resolution DEM using parallel programming in the .NET Framework

High-resolution Digital Elevation Models (DEMs) can be used to extract high-accuracy prerequisite drainage networks. A higher resolution represents a larger number of grids. With an increase in the number of grids, the flow direction determination will require substantial computer resources and computing time. Parallel computing is a feasible method with which to resolve this problem. In this paper, we proposed a parallel programming method within the .NET Framework with a C# Compiler in a Windows environment. The basin is divided into sub-basins, and subsequently the different sub-basins operate on multiple threads concurrently to calculate flow directions. The method was applied to calculate the flow direction of the Yellow River basin from 3 arc-second resolution SRTM DEM. Drainage networks were extracted and compared with HydroSHEDS river network to assess their accuracy. The results demonstrate that this method can calculate the flow direction from high-resolution DEMs efficiently and extract high-precision continuous drainage networks.

Mathematical modeling of the process of fluid filtration through a multi-layer filtering element

The paper presents a mathematical model of the fluid filtration process in a porous medium. The mathematical model of fluid filtration in a porous medium is based on the equation of continuity and the Darcy law for a hollow porous cylinder. The porous cylinder consists of three materials having different mechanical and filtration properties, made of polypropylene fibers by extrusion, consisting of seven layers. Filtration is described by different types of research laws that establish the relationship between a velocity vector of the fluid filtration and the field of pressure. There are several ways to describe mathematically the process of filter colmatation. Present work considers a single-component model of the flow of suspension through a porous medium. An analysis of this model revealed that it can be implemented in the presence of empirical coefficients that take into account porosity of the sediment deposited on the walls of a porous space, linear functions that describe adsorption and desorption of precipitation on the pore walls, as well as functional dependences for the permeability of filter and the viscosity of suspension. An efficient technique to calculate complex models is to apply parallel programming methods, which make it possible to split calculations into streams performed on the basis of supercomputers, clusters, and other high-performance computing systems. The proposed method for organizing distributed calculations for a finite-element filter model could be used to construct a hydrodynamic model of the filtration process, as well as its program implementation, which is planned in the future.

Practical Algorithms for Finding Extremal Sets

The minimal sets within a collection of sets are defined as the ones that do not have a proper subset within the collection, and the maximal sets are the ones that do not have a proper superset within the collection. Identifying extremal sets is a fundamental problem with a wide range of applications in SAT solvers, data mining, and social network analysis. In this article, we present two novel improvements of the high-quality extremal set identification algorithm, AMS-Lex , described by Bayardo and Panda. The first technique uses memoization to improve the execution time of the single-threaded variant of the AMS-Lex, while our second improvement uses parallel programming methods. In a subset of the presented experiments, our memoized algorithm executes more than 400 times faster than the highly efficient publicly available implementation of AMS-Lex. Moreover, we show that our modified algorithm's speedup is not bounded above by a constant and that it increases as the length of the common prefixes in successive input itemsets increases. We provide experimental results using both real-world and synthetic datasets, and show our multithreaded variant algorithm outperforming AMS-Lex by 3 to 6 times. We find that on synthetic input datasets, when executed using 16 CPU cores of a 32-core machine, our multithreaded program executes about as fast as the state-of-the-art parallel GPU-based program using an NVIDIA GTX 580 graphics processing unit.

A parallel quadratic programming method for dynamic optimization problems

Quadratic programming problems (QPs) that arise from dynamic optimization problems typically exhibit a very particular structure. We address the ubiquitous case where these QPs are strictly convex and propose a dual Newton strategy that exploits the block-bandedness similarly to an interior-point method. Still, the proposed method features warmstarting capabilities of active-set methods. We give details for an efficient implementation, including tailored numerical linear algebra, step size computation, parallelization, and infeasibility handling. We prove convergence of the algorithm for the considered problem class. A numerical study based on the open-source implementation qpDUNES shows that the algorithm outperforms both well-established general purpose QP solvers as well as state-of-the-art tailored control QP solvers significantly on the considered benchmark problems.