Earticle

Along with quad-core PC coming into the market and invention of 80-core processor in the laboratory, Software development will go through a foundational innovation due to multi-core technology. The technique of parallel programming on multi-core computers is explored in this thesis. First, the OpenMP standard which is an application programming interface (API) on parallel programming model of shared-memory is introduced and an overview of a set of compiler directives and a library of support functions are given. The OpenMP programs requires an OpenMP-compatible compiler and thread-safe libraries，therefore, both Intel C++ compiler 9.1 and Microsoft Visual Studio 2005 are perfect choices. Then, two-dimensional discrete fast Fourier transform (FFT) is studied by focusing on parallel program design，realization and optimization technology. Finally，the thesis predicts that high performance parallel computing software component library must be a perfect exploitation field in the further future.

The requirement of information security on network has become more and more important. Cryptography is a method to provide information confidentiality, authenticity and integrity. There are so many challenges to implement cryptography algorithm such as execution time, memory requirement, and computation power. Parallel computation is a promising technique to improve the performance of cryptography algorithm. Mainly divide-and-conquer strategy is used in parallel computation to solve the algorithms in parallel by partitioning and allocating, number of given subtask to available processing units. Parallel computation can be performed using multicore processors by parallelizing the execution of algorithm in multiple cores. In this paper we explore the implementation of AES (Advanced Encryption Algorithm) cryptography algorithm on dual core processor by using OpenMP API to reduce the execution time.

Secure communication is a challenging issue in modern industries and critical infrastructures. The core technology used for securing the communication is cryptography. Modular multiplication is an important operation in cryptosystems. This paper investigates a novel modular multiplication algorithm and architecture. In the proposed algorithm and architecture, the parallel architecture and compact SD technique are utilized to improve the performance of modular multiplication operation and cryptosystems. The proposed architecture is implemented on Xilinx Virtex 5 FPGA. The complexity analysis results and FPGA implementation results show that the proposed modular multiplication algorithm and architecture provide improvement on the total computation time and area×time complexity compared to other modified modular multiplication algorithms and architectures.

Query processing technology has recently received a lot of attention in the business intelligence and information service communities. However, the existing approaches can not efficiently optimize the query performance in the uncertain big data environment. In this paper, we propose QPPUBG, a novel and efficient query processing platform for uncertain big data. QPPUBG mainly includes four modules: (i) query equivalence reconstructing for uncertain big data; (ii) multiple query optimization over probability relation components; (iii) query execution plan constructing over probability relation components, and (iv) physical implementation solution of query for uncertain big data. Specially, QPPUBG can support the possible world instance semantics and efficiently handle arbitrary decision spaces. Moreover, QPPUBG can seamlessly integrate the above four modules into the modern parallel computation frameworks. We present the extensive experiments that demonstrate QPPUBG is both efficient and effective.

With more and more consumer electronics apply multi-core chips, the traditional serialized JavaScript execution engine, which is optimized by just-in-time (JIT) compilation technology, fails to utilize multi-core advantages. This paper proposes a mathematical model to detect the dependency of serial JavaScript tasks and a parallelism execution algorithm for serial JavaScript execution engines. Moreover, the parallel JavaScript execution engine with thread-level speculation technology is implemented based on the SquirrelFish Extreme engine of WebKit. As the experiment were conducted respectively on the general test platform platform Sunspider in the industry and world top 15 websites at traffic volume, the results indicate that both in the real Web application and Sunspider platform, the parallel JavaScript execution engines with 2 to 16 threads can raise the performance dramatically compared with a SquirrelFish execution engine with or without JIT acceleration, respectively.

This paper presents a parallel implementation of stiff-ness matrix calculation based on the processor farm model on a net-work of workstations running PVM programming environment. As the computational characteristics of stiffnes matrix exhibits good po-tentials for effective prallel computation the performance improve-ment is show to be almost linear with the number of sorkstations involved in the computation.

In this paper, we propose efficient algorithms for parallel prefix computation and sorting on a recursive dual-net. The recursive dual-net $RDN^k$(B) for k > 0 has $(2n_o)^{2K}/2$ nodes and $d_0$ + k links per node, where $n_0$ and $d_0$ are the number of nod es and the node-degree of the base-network B, respectively. Assume that each node holds one data item, the communication and computation time complexities of the algorithm for parallel prefix computation on $RDN^k$(B), k > 0, are $2^{k+1}-2+2^kT_{comm}(0)$ and $2^{k+1}-2+2^kT_{comp}(0)$, respectively, where $T_{comm}(0)$ and $T_{comp}(0)$ are the communication and computation time complexities of the algorithm for parallel prefix computation on the base-network B, respectively. The algorithm for parallel sorting on $RDN^k$(B) is restricted on B = $Q_m$ where $Q_m$ is an m-cube. Assume that each node holds a single data item, the sorting algorithm runs in $O((m2^k)^2)$ computation steps and $O((km2^k)^2)$ communication steps, respectively.

Recently, an iterative algorithm for finding the interior eigenvalues of a definite matrix by CG-type method has been proposed. This method compares to the inverse power method. The given matrices A, and B are assumed to be large and sparse, and SPD( Symmetric Positive Definite) The CG scheme for the optimization of the Rayleigh quotient has been proven a very attractive and promising technique for large sparse eigenproblems for smallest eigenvalue. Also, it is very amenable to parallel computations, like the CG method for the linear systems. A proper choice of the preconditioner significantly improves the convergence of the CG scheme. But for parallel computations we need to find an efficient parallel preconditioner. Our candidates we ILU(0) in the wave-front order, ILU(0) in the multi-coloring order, Point-SSOR(Symmetric Successive Overrelaxation), and Multi-Color Block SSOR preconditioner. Wavefront order is a simple way to increase parallelism in the natural order, and Multi-coloring realizes a parallelism of order(N), where N is the order of the matrix. Another choice is the Multi-Color Block SSOR(Symmetric Successive OverRelaxation) preconditioning. Block SSOR is a symmetric preconditioner which is expected to minimize the interprocessor communication due to the blocking. We implemented the results on the CRAY-T3E with 128 nodes. The MPI (Message Passing Interface) library was adopted for the interprocessor communications. The test problem was drawn from the discretizations of partial differential equations by finite difference methods. The results show that for small number of processors Multi-Color ILU(0) has the best performance, while for large number of processors Multi-Color Block SSOR performs the best.

The method of moments is applied to numerically compute the electrostatic capacitance of a parallel-plate rectangular capacitor of finite area. Each plate is discretized into 900 patches per unit area to ensure a high accuracy of computation. To further enhance computational results, the impedance matrix elements are additionally evaluated in the case that the observation patch is located above or below the source patch in the vertical direction. To examine the fringing effect at the edges of the capacitor, the normalized capacitances are computed as a function of separation distance. After these results have been verified by Palmer's formula, this method is extended to the computation of capacitances between two different size plates.

본 논문에서는 3차원 전자기장의 병렬 해석 기법을 제안하였다. 시간 조화 벡터 파동 방정식 및 유한요소 기법에 기반한 전자기장 산란 해석이 수행되었으며, 모서리 기반 요소 및 2차 흡수 경계 조건이 도입되었다. 개발한 알고리즘은 유한요소망을 분할한 뒤 각 프로세서에 할당함으로써 요소별 수치적분 및 행렬 조립 과정의 병렬화를 달성하였다. 이때 부영역 생성을 위해 그래프 분할 라이브러리인 METIS가 도입되었다. 대형 희박행렬 방정식의 계산은 다중 프론탈 기법 기반 병렬 연산 라이브러리인 MUMPS를 통해 수행되었다. 개발된 프로그램의 정확도는 Mie 이론해 및 ANSYS HFSS 결과와의 비교를 통해 검증되었다. 또한 사용된 프로세서 수에 따른 가속 지표를 측정하여 확장성을 확인하였다. 완전 전기 도체 구, 등·이방성 유전체 구 및 유도탄 예제 형상에 대한 전자기장 산란 해석이 수행되었다. 개발된 프로그램의 알고리즘은 추후 유한요소 분할 및 합성법에 활용될 예정이며, 더욱 확장된 병렬 연산 성능을 목표하고자 한다.

In this paper, parallel computing method on the three-dimensional electromagnetic field is proposed. The present electromagnetic scattering analysis is conducted based on the time-harmonic vector wave equation and the finite element method. The edge-based element and 2^nd -order absorbing boundary condition are used. Parallelization of the elemental numerical integration and the matrix assemblage is accomplished by allocating the partitioned finite element subdomain for each processor. The graph partitioning library, METIS, is employed for the subdomain generation. The large sparse matrix computation is conducted by MUMPS, which is the parallel computing library based on the multi-frontal method. The accuracy of the present program is validated by the comparison against the Mie-series analytical solution and the results by ANSYS HFSS. In addition, the scalability is verified by measuring the speed-up in terms of the number of processors used. The present electromagnetic scattering analysis is performed for a perfect electric conductor sphere, isotropic/anisotropic dielectric sphere, and the missile configuration. The algorithm of the present program will be applied to the finite element and tearing method, aiming for the further extended parallel computing performance.

알파벳 ${\Sigma}$로 구성된 길이가 각각 m, n인 두 문자열 X, Y가 주어졌을 때, X, Y의 확장편집거리는 동적프로그래밍을 이용하여 O(mn) 시간과 공간을 계산할 수 있다. 최근 m개의 쓰레드를 이용하여 O(m+n) 시간과 O(mn) 공간을 사용하여 X, Y의 확장편집거리를 계산하는 병렬알고리즘이 제시되었다. 본 논문에서는 GPU의 공유메모리를 활용하여 수행시간을 개선한 병렬알고리즘을 제시한다. 실험 결과, 개선된 병렬알고리즘이 기존의 병렬알고리즘보다 약 19~25배 이상 빠른 수행시간을 보였다.

Given two strings X and Y (|X|=m, |Y|=n) over an alphabet ${\Sigma}$, the extended edit distance between X and Y can be computed using dynamic programming in O(mn) time and space. Recently, a parallel algorithm that takes O(m+n) time and O(mn) space using m threads to compute the extended edit distance between X and Y was presented. In this paper, we present an improved parallel algorithm using the shared memory on GPU. The experimental results show that our parallel algorithm runs about 19~25 times faster than the previous parallel algorithm.

Parallel finite element code has been recently developed for the analysis of the incompressible Wavier-Stokes equations using domain decomposition method. Metis and MPI libraries are used for the domain partitioning of an unstructured mesh and the data communication between sub-domains, respectively. For unsteady computation of the incompressible Navier-Stokes equations, 4-step splitting method is combined with P1P1 finite element formulation. Smagorinsky and dynamic model are implemented for the simulation of turbulent flows. For the validation performance-estimation of the developed parallel code, three-dimensional Laplace equation has been solved. It has been found that the speed-up of 40 has been obtained from the present parallel code fir the bench mark problem. Lastly, the turbulent flows around the MIRA model and Tiburon model have been solved using 32 processors on IBM SMP cluster and unstructured mesh. The computed drag coefficient agrees better with the existing experiment as the mesh resolution of the region increases, where the variation of pressure is severe.

This paper describes an parallell computing algorithm in Gauss elimination of Jacobian matrix to large-scale power system. The structure of Jacobian matrix becomes different according to ordering method of buses. In sequential computation buses are ordered to minimize the number of fill-in in the triangulation of the Jacobian matrix. The proposed method using ND(nested dissection) ordering develops the parallelism in the Gauss elimination to have balance of computing load among processes and each processor uses the sequential computation method to preserve the sparsity of matrix.