scholarly journals Non-Stationary Acceleration Strategies for PageRank Computing

Mathematics ◽  
2019 ◽  
Vol 7 (10) ◽  
pp. 911
Author(s):  
Héctor Migallón ◽  
Violeta Migallón ◽  
José Penadés
Keyword(s):  
Parallel Algorithms ◽  
Parallel Computation ◽  
Parallel Implementation ◽  
Power Method ◽  
Extrapolation Methods ◽  
Global Synchronization ◽  
Two Stage ◽  
Stationary Model ◽  
Stationary Approach ◽  
Parallel Techniques

In this work, a non-stationary technique based on the Power method for accelerating the parallel computation of the PageRank vector is proposed and its theoretical convergence analyzed. This iterative non-stationary model, which uses the eigenvector formulation of the PageRank problem, reduces the needed computations for obtaining the PageRank vector by eliminating synchronization points among processes, in such a way that, at each iteration of the Power method, the block of iterate vector assigned to each process can be locally updated more than once, before performing a global synchronization. The parallel implementation of several strategies combining this novel non-stationary approach and the extrapolation methods has been developed using hybrid MPI/OpenMP programming. The experiments have been carried out on a cluster made up of 12 nodes, each one equipped with two Intel Xeon hexacore processors. The behaviour of the proposed parallel algorithms has been studied with realistic datasets, highlighting their performance compared with other parallel techniques for solving the PageRank problem. Concretely, the experimental results show a time reduction of up to 58.4 % in relation to the parallel Power method, when a small number of local updates is performed before each global synchronization, outperforming both the two-stage algorithms and the extrapolation algorithms, more sharply as the number of processes increases.

scholarly journals Large-eddy simulation of wall-bounded turbulent flow with high-order discrete unified gas-kinetic scheme

2020 ◽  
Vol 2 (1) ◽  
Author(s):  
Rui Zhang ◽  
Chengwen Zhong ◽  
Sha Liu ◽  
Congshan Zhuo
Keyword(s):  
Turbulent Flow ◽  
Turbulent Flows ◽  
Parallel Implementation ◽  
Kinetic Scheme ◽  
Cavity Flow ◽  
Equilibrium Distribution Function ◽  
Third Order ◽  
Two Stage ◽  
Large Eddy ◽  
Unified Gas Kinetic Scheme

AbstractIn this paper, we introduce the discrete Maxwellian equilibrium distribution function for incompressible flow and force term into the two-stage third-order Discrete Unified Gas-Kinetic Scheme (DUGKS) for simulating low-speed turbulent flows. The Wall-Adapting Local Eddy-viscosity (WALE) and Vreman sub-grid models for Large-Eddy Simulations (LES) of turbulent flows are coupled within the present framework. Meanwhile, the implicit LES are also presented to verify the effect of LES models. A parallel implementation strategy for the present framework is developed, and three canonical wall-bounded turbulent flow cases are investigated, including the fully developed turbulent channel flow at a friction Reynolds number (Re) about 180, the turbulent plane Couette flow at a friction Re number about 93 and lid-driven cubical cavity flow at a Re number of 12000. The turbulence statistics, including mean velocity, the r.m.s. fluctuations velocity, Reynolds stress, etc. are computed by the present approach. Their predictions match precisely with each other, and they are both in reasonable agreement with the benchmark data of DNS. Especially, the predicted flow physics of three-dimensional lid-driven cavity flow are consistent with the description from abundant literature. The present numerical results verify that the present two-stage third-order DUGKS-based LES method is capable for simulating inhomogeneous wall-bounded turbulent flows and getting reliable results with relatively coarse grids.

scholarly journals A Parallel Implementation of Unscheduled Flow Control in Interconnected Power Systems

2012 ◽  
Vol 2012 ◽  
pp. 1-19
Author(s):  
G. Ozdemir Dag ◽  
Mustafa Bagriyanik
Keyword(s):  
Power Systems ◽  
Power System ◽  
Parallel Computation ◽  
High Performance ◽  
Power Flow ◽  
Parallel Implementation ◽  
Computation Time ◽  
Test System ◽  
Fuzzy Decision Making ◽  
Optimization Approach

The unscheduled power flow problem needs to be minimized or controlled as soon as possible in a deregulated power system since the transmission systems are mostly operated at their power-carrying limits or very close to it. The time spent for simulations to determine the current states of all the system and control variables of the interconnected power system is important. Taking necessary action in case of any failure of equipment or any other occurrence of an undesired situation could be critical. Using supercomputing facilities and parallel computing techniques together decreases the computation time greatly. In this study, a parallel implementation of a multiobjective optimization approach based on both genetic algorithms and fuzzy decision making to manage unscheduled flows is presented. Parallel computation techniques are applied using supercomputers (high-performance computers). The proposed method is applied to the IEEE 300 bus test system. Two different cases for some parameters of GA are considered to see the power of parallel computation technique. Then the simulation results are presented.

scholarly journals SPATIOTEMPORAL DOMAIN DECOMPOSITION FOR MASSIVE PARALLEL COMPUTATION OF SPACE-TIME KERNEL DENSITY

2015 ◽  
Vol II-4/W2 ◽  
pp. 7-11 ◽  
Author(s):  
A. Hohl ◽  
E. M. Delmelle ◽  
W. Tang
Keyword(s):  
Domain Decomposition ◽  
Parallel Computation ◽  
High Performance ◽  
Parallel Implementation ◽  
Kernel Density ◽  
Great Accuracy ◽  
Space Time ◽  
Size Diversity ◽  
Data Points ◽  
Tropical Climates

Accelerated processing capabilities are deemed critical when conducting analysis on spatiotemporal datasets of increasing size, diversity and availability. High-performance parallel computing offers the capacity to solve computationally demanding problems in a limited timeframe, but likewise poses the challenge of preventing processing inefficiency due to workload imbalance between computing resources. Therefore, when designing new algorithms capable of implementing parallel strategies, careful spatiotemporal domain decomposition is necessary to account for heterogeneity in the data. In this study, we perform octtree-based adaptive decomposition of the spatiotemporal domain for parallel computation of space-time kernel density. In order to avoid edge effects near subdomain boundaries, we establish spatiotemporal buffers to include adjacent data-points that are within the spatial and temporal kernel bandwidths. Then, we quantify computational intensity of each subdomain to balance workloads among processors. We illustrate the benefits of our methodology using a space-time epidemiological dataset of Dengue fever, an infectious vector-borne disease that poses a severe threat to communities in tropical climates. Our parallel implementation of kernel density reaches substantial speedup compared to sequential processing, and achieves high levels of workload balance among processors due to great accuracy in quantifying computational intensity. Our approach is portable of other space-time analytical tests.

scholarly journals Large-eddy Simulation of Wall-bounded Turbulent Flow with High-order Discrete Unified Gas-Kinetic Scheme

2020 ◽  
Author(s):  
Rui Zhang ◽  
Chengwen Zhong ◽  
Sha Liu ◽  
Congshan Zhuo
Keyword(s):  
Turbulent Flows ◽  
Parallel Implementation ◽  
Computational Cost ◽  
Three Dimensional ◽  
Kinetic Scheme ◽  
Equilibrium Distribution Function ◽  
Third Order ◽  
Two Stage ◽  
Large Eddy ◽  
Unified Gas Kinetic Scheme

Abstract In this paper, we introduce the incompressible discrete Maxwellian equilibrium distribution function and external forces into the two-stage third-order Discrete Unified Gas-Kinetic Scheme (DUGKS) for simulating low-speed incompressible turbulent flows with forcing term. The Wall-Adapting Local Eddy-viscosity (WALE) and Vreman sub-grid models for Large-Eddy Simulations (LES) of wall-bounded turbulent flows are coupled within the present framework. In order to simulate the three-dimensional turbulent flows associated with great computational cost, a parallel implementation strategy for the present framework is developed, and is validated by three canonical wall-bounded turbulent flows, viz., the fully developed turbulent channel flow at a friction Reynolds number (Re) about 180, the turbulent plane Couette flow at a friction Re number about 93 and three-dimensional lid-driven cubical cavity flow at a Re number of 12000. The turbulence statistics are computed by the present approach with both WALE and Vreman models, and their predictions match precisely with each other. Especially, the predicted flow physics of three-dimensional lid-driven cavity are consistent with the description from abundant literatures. While, they have small discrepancies in comparison to the Direct Numerical Simulation (DNS) due to the relatively low grid resolution. The present numerical results verify that the present two-stage third-order DUGKS-based LES method is capable for simulating inhomogeneous wall-bounded turbulent flows and getting reliable results with relatively coarse grids.

PARALLEL COMPUTATION OF INTERNAL AND EXTERNAL FARTHEST NEIGHBORS IN SIMPLE POLYGONS

1992 ◽  
Vol 02 (02) ◽  
pp. 175-190 ◽  
Author(s):  
SUMANTA GUHA
Keyword(s):  
Parallel Algorithms ◽  
Parallel Computation ◽  
Shortest Path ◽  
Simple Polygon ◽  
Divide And Conquer ◽  
Simple Polygons ◽  
Crew Pram ◽  
Serial Algorithm

We present efficient parallel algorithms for two problems in simple polygons: the all-farthest neighbors problem and the external all-farthest neighbors problem. The all-farthest neighbors problem is that of computing, for each vertex p of a simple polygon P, a point ψ(p) in P farthest from p when the distance between p and ψ(p) is measured by the shortest path between them constrained to lie inside P. The external all-farthest neighbors problem is that of computing, for each vertex p of a simple polygon P, a point ϕ(p) on (the boundary of) P farthest from p when the distance between p and ϕ(p) is measured by the shortest path between them constrained to lie outside (the interior of) P. Both our algorithms run in O( log 2 n) time on a CREW PRAM with O(n) processors. Our divide-and-conquer method for the external all-farthest neighbors problem, in fact, leads to a new O(n log n) time serial algorithm that matches the currently best serial algorithm for this problem, but is simpler.

ON DESIGNING COMMUNICATION-INTENSIVE ALGORITHMS FOR A SPANNING OPTICAL BUS BASED ARRAY

1995 ◽  
Vol 05 (03) ◽  
pp. 499-511 ◽  
Author(s):  
CHUNMING QIAO
Keyword(s):  
Parallel Algorithms ◽  
Parallel Implementation ◽  
Reconfigurable Array ◽  
The Matrix ◽  
Matrix Transposition ◽  
Optical Buses ◽  
And Control

The Reconfigurable Array with Spanning Optical Buses (or RASOB) architecture provides flexible reconfiguration and strong connectivities with low hardware and control complexities. We use a parallel implementation of the matrix transposition as well as multiplication algorithms as an example to show how the architectural capabilities can be taken advantage of in designing efficient parallel algorithms.

scholarly journals Design and implementation of pipelined and parallel AES encryption systems using FPGA

2020 ◽  
Vol 20 (1) ◽  
pp. 287
Author(s):  
Mohamed Nabil ◽  
Ashraf A. M. Khalaf ◽  
Sara M. Hassan
Keyword(s):  
Big Data ◽  
Parallel Algorithms ◽  
Processing Time ◽  
Advanced Encryption Standard ◽  
Time Consumption ◽  
Design And Implementation ◽  
Aes Algorithm ◽  
Detailed Explanation ◽  
Encryption Algorithms ◽  
Parallel Techniques

<p><span>The information security is one of the most important issues in the design of any communication network.One of the most common encryption algorithms is the Advanced Encryption Standard (AES).The main problem facing the AES algorithm is the high time consumption due to the large number of rounds used for performing the encryption operation. The more time the encryption system consumes to encrypt the data, the more chances the hackers have to break the system.This paper presents two effective algorithms that can be used to solve the mentioned problem and to achieve an effective processing time reduction using pipelined and parallel techniques to perform the encryption steps. These algorithms are based on using certain techniques to make the system able to encrypt many different states (the data will be encrypted) in the same time with no necessity to wait for the previous encryption operation to be completed. These two algorithms are very effective especially for big data size. This paper describes in detail the AES encryption system algorithm and a detailed explanation for the proposed algorithms. Moreover, the research shows the implementation of the three algorithms: the traditional, the pipelined, and the parallel algorithms, and finally a comparison between them.</span></p>

Sign in / Sign up

Export Citation Format

Share Document