The parallelization of the diagonalization step of the COLUMBUS MRSDCI program system is reported. A coarse grain algorithm has been developed by means of a segmentation of the trial and resulting update vectors of th...
详细信息
The parallelization of the diagonalization step of the COLUMBUS MRSDCI program system is reported. A coarse grain algorithm has been developed by means of a segmentation of the trial and resulting update vectors of the iterative Davidson scheme. Message passing based on the TCGMSG toolkit and the global array (GA) tools are used. The latter program system allows an asynchronous access to data structures in the spirit of shared memory. The importance of portable facilities like GA going beyond message passing is stressed for quantum chemical methods and benchmark result for the Intel Touchstone Delta are given.
Concept-cognitive learning, as an interdisciplinary study of concept lattice and cognitive learning, has become a hot research direction among the communities of rough set, formal concept analysis and granular computi...
详细信息
Concept-cognitive learning, as an interdisciplinary study of concept lattice and cognitive learning, has become a hot research direction among the communities of rough set, formal concept analysis and granular computing in recent years. The main objective of concept-cognitive learning is to learn concepts from a give clue with the help of cognitive learning methods. Note that this kind of studies can provide concept lattice insight to cognitive learning. In order to deal with more complex data and improve learning efficiency, this paper investigates parallel computing techniques for concept-cognitive learning in terms of large data and multi-source data based on granular computing and information fusion. Specifically, for large data, a parallel computing framework is designed to extract global granular concepts by combining local granular concepts. For multi-source data, an effective information fusion strategy is adopted to obtain final concepts by integrating the concepts from all single-source data. Finally, we conduct some numerical experiments to evaluate the effectiveness of the proposed parallel computing algorithms.
作者：
PEKERGIN, MFUNIV PARIS 11
RECH INFORMAT LABSYNCHRONIZAT & PARALLELISM RES GRPCNRSURA 410F-91405 ORSAYFRANCE
In this paper we deal with the performance of parallel computing in a network of Apollo workstations where the processes use the Remote Procedure Call (RPC) mechanism for communication. The speedup in such systems can...
详细信息
In this paper we deal with the performance of parallel computing in a network of Apollo workstations where the processes use the Remote Procedure Call (RPC) mechanism for communication. The speedup in such systems cannot be accurately estimated without taking into account the relatively large communication overheads. Moreover, it decreases by increasing parallelism when the latter exceeds some certain limit. The main aim of this work is to estimate the speedup and optimum degree of parallelism from the characteristics of applications. Hence the speedup can be maximized by the appropriate choice of the degree of parallelism, or the parallelization efforts can be economized when the estimated speedup reveals to be unsatisfactory. In order to estimate the speedup and determine the optimum degree of parallelism, first we characterize the parallelization and the communication overheads in the system considered. Then parallel applications are modeled and their execution times are expressed for the general case of nonidentical tasks and workstations. The general case study allows to take into account the structural constraints of the applications by permitting their partitioning into heterogenous tasks. A simple expression of the optimum degree of parallelism is obtained for identical tasks where the inherent constraints are neglected. The fact that the theoretical maximum speedup is bounded by half of the optimum degree of parallelism shows the importance of this measure. Our approach is valid for all systems which use RPC as a communication mechanism.
It is shown that a relatively simple dynamical dc electric arc model shows complicated two-parameter (2-D) bifurcations with both periodic and chaotic responses. 2-D bifurcation diagrams for the arc model [a system of...
详细信息
It is shown that a relatively simple dynamical dc electric arc model shows complicated two-parameter (2-D) bifurcations with both periodic and chaotic responses. 2-D bifurcation diagrams for the arc model [a system of three ordinary differential equations (ODEs)] are obtained by using parallel computations because obtaining a single 2-D diagram requires solving the ODE system hundreds of thousands or even a few millions of times (depending on the intervals of parameters and assumed resolution). Several color 2-D bifurcation diagrams are presented, and the speedup factors of their parallel computations are provided. Numerical computations of periodic and chaotic responses of the 2-D bifurcation diagrams are confirmed by both the one-parameter (1-D) diagrams and the 0-1 test for chaos. Some further theoretical aspects of parallel computing of 2-D bifurcation diagrams are also considered.
This paper presents the application of parallel computing techniques to large-scale modeling of fluid flow in the unsaturated zone (UZ) at Yucca Mountain, Nevada. In this study, parallel computing techniques, as imple...
详细信息
This paper presents the application of parallel computing techniques to large-scale modeling of fluid flow in the unsaturated zone (UZ) at Yucca Mountain, Nevada. In this study, parallel computing techniques, as implemented into the TOUGH2 code, are applied in large-scale numerical simulations on a distributed-memory parallel computer. The modeling study has been conducted using an over-1-million-cell three-dimensional numerical model, which incorporates a wide variety of field data for the highly heterogeneous fractured formation at Yucca Mountain. The objective of this study is to analyze the impact of various surface infiltration scenarios (under current and possible future climates) on flow through the UZ system, using various hydrogeological conceptual models with refined grids. The results indicate that the 1-million-cell models produce better resolution results and reveal some flow patterns that cannot be obtained using coarse-grid modeling models. Published by Elsevier Science B.V.
Minimum weight formulations with stress constraints have been recently proposed as an alternative to the traditional maximum stiffness statements for the topology optimization of structures. These minimum weight appro...
详细信息
Minimum weight formulations with stress constraints have been recently proposed as an alternative to the traditional maximum stiffness statements for the topology optimization of structures. These minimum weight approaches offer some advantages and avoid unwanted phenomena associated to maximum stiffness approaches, e.g. mesh dependency, checkerboard layouts. In addition, minimum weight formulations analyze more usual statements from a practical point of view in engineering since they reduce the cost of the structure and impose stress constraints. The minimum weight formulation with stress constraints proposed in this paper guarantees the feasibility of the optimal solutions obtained while the cost is minimized. However, these formulations also require bigger computing effort than the traditional maximum stiffness statements since the number of highly non-linear stress constraints is drastically increased while the number of design variables is analogous. Thus, it is necessary to introduce numerical methods and computational techniques that allow to reduce the resources required. In this paper, we propose the use of parallelization techniques in order to reduce the computing time required to solve the topology optimization problem with stress constraints proposed. (C) 2013 Elsevier Ltd. All rights reserved.
A new algorithm for the parallel computing of the boundary-element and finite-element combination method is presented in this paper. By introducing the domain decomposition of an entire domain under consideration into...
详细信息
A new algorithm for the parallel computing of the boundary-element and finite-element combination method is presented in this paper. By introducing the domain decomposition of an entire domain under consideration into the boundary-element and finite-element subdomains, each analysis is performed independently and in parallel. A renewal iterative scheme for parallel computing is the Schwarz method which was previously adopted to the domain decomposition parallel scheme in boundary-element analysis. A cluster parallel computing system by workstations connected by the LAN is constructed and employed aiming at efficient analysis. Convergence and accuracy of solutions on the internal virtual boundaries are shown through some sample examples. (C) 1977 Elsevier Science Ltd.
The Markov chain random field(MCRF)model is a spatial statistical approach for modeling categorical spatial variables in multiple ***,this approach tends to be computationally costly when dealing with large data sets ...
详细信息
The Markov chain random field(MCRF)model is a spatial statistical approach for modeling categorical spatial variables in multiple ***,this approach tends to be computationally costly when dealing with large data sets because of its sequential simulation ***,improving its computational efficiency is necessary in order to run this model on larger sizes of spatial *** this study,we suggested four parallel computing solutions by using both central processing unit(CPU)and graphics processing unit(GPU)for executing the sequential simulation algorithm of the MCRF model,and compared them with the nonparallel computing solution on computation time spent for a land cover *** four parallel computing solutions are:(1)multicore processor parallel computing(MP),(2)parallel computing by GPU-accelerated nearest neighbor searching(GNNS),(3)MP with GPU-accelerated nearest neighbor searching(MPGNNS),and(4)parallel computing by GPU-accelerated approximation and GPU-accelerated nearest neighbor searching(GA-GNNS).Experimental results indicated that all of the four parallel computing solutions are at least 1.8×faster than the nonparallel ***,the GA-GNNS solution with 512 threads per block is around 83×faster than the nonparallel solution when conducting a land cover post-classification with a remotely sensed image of 1000×1000 pixels.
Available computing power for researchers has been increasing exponentially over the last decade. parallel computing is possibly the best way to harness computing power provided by multiple computing units. This paper...
详细信息
Available computing power for researchers has been increasing exponentially over the last decade. parallel computing is possibly the best way to harness computing power provided by multiple computing units. This paper reviews parallel computing applications in railway research as well as the enabling techniques used for the purpose. Nine enabling techniques were reviewed and Message Passing Interface, Domain Decomposition and Hadoop & Apache are the top three most widely used enabling techniques. Seven major application topics were reviewed and iterative optimisations, continuous dynamics and data & signal analysis are the most widely reported applications. The reasons why these applications are suitable for parallel computing were discussed as well as the suitability of various enabling techniques for different applications. computing time speed-ups that were reported from these applications were summarised. The challenges for applying parallel computing for railway research are discussed.
作者：
Macfarlane, ARobertson, SEMcCann, JA
(School of Informatics City University Northampton Square London EC1V 0HB)
(School of Informatics City University Northampton Square London EC1V 0HB)
(School of Informatics City University Northampton Square London EC1V 0HB)
The progress of parallel computing in Information Retrieval (IR) is reviewed. In particular we stress the importance of the motivation in using parallel computing for text retrieval. We analyse parallel IR systems usi...
详细信息
The progress of parallel computing in Information Retrieval (IR) is reviewed. In particular we stress the importance of the motivation in using parallel computing for text retrieval. We analyse parallel IR systems using a classification defined by Rasmussen [1] and describe some parallel IR systems. We give a description of the retrieval models used in parallel information processing. We describe areas of research which we believe are needed.
暂无评论