By Topic

Eighth International Conference on High-Performance Computing in Asia-Pacific Region (HPCASIA'05)

Date Nov. 30 2005-Dec. 3 2005

Filter Results

Displaying Results 1 - 25 of 103
  • Proceedings. Eighth International Conference on High-Performance Computing in Asia-Pacific Region

    Publication Year: 2005, Page(s): c1
    Request permission for commercial reuse | PDF file iconPDF (2757 KB)
    Freely Available from IEEE
  • Eighth International Conference on High-Performance Computing in Asia-Pacific Region - Title Page

    Publication Year: 2005, Page(s):i - iii
    Request permission for commercial reuse | PDF file iconPDF (134 KB)
    Freely Available from IEEE
  • Eighth International Conference on High-Performance Computing in Asia-Pacific Region - Copyright

    Publication Year: 2005, Page(s): iv
    Request permission for commercial reuse | PDF file iconPDF (117 KB)
    Freely Available from IEEE
  • Eighth International Conference on High-Performance Computing in Asia-Pacific Region - Table of Contents

    Publication Year: 2005, Page(s):v - xi
    Request permission for commercial reuse | PDF file iconPDF (150 KB)
    Freely Available from IEEE
  • General Co-Chairs Message

    Publication Year: 2005, Page(s): xii
    Request permission for commercial reuse | PDF file iconPDF (101 KB) | HTML iconHTML
    Freely Available from IEEE
  • Program Co-Chairs Message

    Publication Year: 2005, Page(s): xiii
    Request permission for commercial reuse | PDF file iconPDF (101 KB)
    Freely Available from IEEE
  • Organizing Committee

    Publication Year: 2005, Page(s): xiv
    Request permission for commercial reuse | PDF file iconPDF (102 KB)
    Freely Available from IEEE
  • Acknowledgement to additional reviewers

    Publication Year: 2005, Page(s): xvi
    Request permission for commercial reuse | PDF file iconPDF (97 KB)
    Freely Available from IEEE
  • Towards a more realistic BSP cost model

    Publication Year: 2005, Page(s):10 pp. - 12
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (271 KB) | HTML iconHTML

    In this paper, we present a refinement of the BSP (bulk synchronous parallel) cost model, in order to allow a more exact prediction of the parallel algorithms communication cost. Our approach is based on two point: (I) a deepening of the benchmarks to take into account all influential factors on the word sending cost in a communication; and (II) a more elaborate manner of prediction which carefull... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Optimal algorithms for scheduling large-scale divisible load on heterogeneous systems in non-blocking mode of communication

    Publication Year: 2005, Page(s):8 pp. - 20
    Cited by:  Papers (1)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (299 KB) | HTML iconHTML

    This paper considers scheduling large-scale divisible load on heterogeneous single-level tree networks to minimize the whole processing time. A more realistic and general model, i.e., both processors and links may have different speeds and arbitrary start-up costs, and with nonblocking mode of communication, is introduced. Under such environment, results are obtained including: (1) closed-form exp... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • A scalable dynamic network memory service system

    Publication Year: 2005, Page(s):7 pp. - 27
    Cited by:  Papers (4)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (215 KB) | HTML iconHTML

    The performance of accessing network memory via high-speed network has much advantage compared with that via local disk. So, to expand local physical memory by using the network memory has gained broad attention. This paper presents a SDNMS (scalable dynamic network memory service) system. SDNMS provides standard service interface therefore can be applied to any client system which needs to expand... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Optimizing strided remote memory access operations on the Quadrics QsNetII network interconnect

    Publication Year: 2005, Page(s):8 pp. - 35
    Cited by:  Papers (4)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (484 KB) | HTML iconHTML

    This paper describes and evaluates protocols for optimizing strided noncontiguous communication on the Quadrics QsNetII high-performance network interconnect. Most of previous related studies focused primarily on NIC-based or host-based protocols. This paper discusses merits for using both approaches and tries to determine types and data sizes in the communication operations for which these protoc... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Achieving high performance to support multicast traffic in a parallel packet switch with space division multiplexing expansion

    Publication Year: 2005, Page(s):6 pp. - 44
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (248 KB) | HTML iconHTML

    A parallel packet switch (PPS) is a switch in which the memories run slower than the line rate. Till now, most of theoretical researches on PPS simply rely on increasing the speedup of switch-layers to emulate a "First-come First-serve" output queued (FCFS-OQ) switch. However, when the input traffic is multicast pattern, the speedup requirement of a N-port PPS will increase to 2radicN + 1. Clearly... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Adaptive Strassen and ATLAS's DGEMM: a fast square-matrix multiply for modern high-performance systems

    Publication Year: 2005, Page(s):8 pp. - 52
    Cited by:  Patents (2)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (948 KB) | HTML iconHTML

    Strassen's algorithm has practical performance benefits for architectures with simple memory hierarchies, because it trades computationally expensive matrix multiplications (MM) with cheaper matrix additions (MA). However, it presents no advantages for high-performance architectures with deep memory hierarchies, because MAs exploit limited data reuse. We present an easy-to-use adaptive algorithm c... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Efficient subtorus processor allocation in a multi-dimensional torus

    Publication Year: 2005, Page(s):8 pp. - 60
    Cited by:  Papers (3)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (293 KB) | HTML iconHTML

    Processor allocation in a mesh or torus connected multicomputer system with up to three dimensions is a hard problem that has received some research attention in the past decade. With the recent deployment of multicomputer systems with a torus topology of dimensions higher than three, which are used to solve complex problems arising in scientific computing, it becomes imminent to study the problem... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • The linear layout of the incomplete hypercube

    Publication Year: 2005, Page(s):5 pp. - 66
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (372 KB) | HTML iconHTML

    In this paper, we study the linear layout problem of an incomplete hypercube by the embedding-in-book technique. An incomplete hypercube is a generalization of the hypercube in the sense that the number of nodes can be an arbitrary number. Embedding a graph in a book is to place nodes on the spine of a book and to draw the edges such that edges residing in a page do not cross. In this paper, we pr... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Parallel branch-and-bound algorithm for constructing evolutionary trees from distance matrix

    Publication Year: 2005, Page(s):7 pp. - 72
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (208 KB) | HTML iconHTML

    An ultrametric tree is an evolutionary tree in which the distances from the root to all leaves in the tree are equal. The Minimum Ultrametric Tree construction problem is the problem of constructing an ultrametric tree from distance matrices with minimum cost. It is shown that to construct a minimum cost ultrametric tree is NP-hard. In this paper, we present an efficient parallel branch and bound ... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Discrete optimization problems - some new heuristic approaches

    Publication Year: 2005, Page(s):8 pp. - 82
    Cited by:  Papers (2)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (268 KB) | HTML iconHTML

    We consider in this paper some heuristic methods of decision-making in various discrete optimization problems. The object of each of these problems is programming anytime algorithms. Considered methods for solving these problems are constructed on the basis of special combination of some heuristics. We use some modifications of truncated branch-and-bound method; for the selecting immediate step, w... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Parallel programming models for finite-element method using preconditioned iterative solvers with multicolor ordering on various types of SMP cluster supercomputers

    Publication Year: 2005, Page(s):8 pp. - 90
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (360 KB) | HTML iconHTML

    Parallel iterative linear solvers for unstructured grids in FEM application, which were developed for the Earth Simulator (ES), have been ported to various types of SMP cluster supercomputers. Performance of flat-MPI and hybrid parallel programming model has been compared using more than 100 SMP nodes of ES, Hitachi SR8000 and IBM SP-3. Effect of coloring and method for storage of coefficient matr... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Parallel 3-dimensional DCT computation on k-ary n-cubes

    Publication Year: 2005, Page(s):7 pp. - 97
    Cited by:  Papers (1)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (234 KB) | HTML iconHTML

    The three dimensional discrete cosine transform (3D DCT) has been widely used in many applications such as video compression. On the other hand, the k-ary n-cube is one of the most popular interconnection networks used in many recent multicomputers. As direct calculation of 3D DCT is very time consuming, many researchers have been working on developing algorithms and special-purpose architectures ... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Fracture analysis using reconfigurable computing systems

    Publication Year: 2005, Page(s):7 pp. - 104
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (220 KB) | HTML iconHTML

    The fracture mechanics analysis determines if the small flaws in structural components will grow into large enough cracks to cause them to fail catastrophically. An accurate finite element analysis is needed so that the strain-energy-release components can be calculated from the local forces and displacements around the crack tip. Complex fracture mechanics problems demand very high computing powe... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Design and evaluation of dynamic key message algorithms for cluster computing

    Publication Year: 2005, Page(s):8 pp. - 112
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (407 KB) | HTML iconHTML

    The Key Message (KM) approach prioritizes communications along the critical path to speed up the execution of parallel applications in a cluster environment. In this article, we introduce a family of three dynamic KM algorithms that take into account of the dynamic nature of underlying networks to improve over their predecessor. These three dynamic KM algorithms are: Fixed Critical Path Dynamic KM... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Congestion management of self similar IP traffic - weighted probability normal and exponential RED

    Publication Year: 2005, Page(s):8 pp. - 122
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (320 KB) | HTML iconHTML

    Schemes described in the literature on network congestion management are in general based on queue management. Also it is widely accepted that Poisson model is not sufficient to characterize the traffic in current Internet. In this paper, we first discuss the details of alternate RED (Random Early Detection) algorithm for traffic congestion management in IP networks having self similar input. We p... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Parallel distributed structure of the operational data assimilation SSI system in the National Centers for Environmental Predictions (NCEP)

    Publication Year: 2005, Page(s):8 pp. - 130
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (419 KB) | HTML iconHTML

    The parallel-distributed structure of the current SSI operational data assimilation system in the National Centers for Environmental Predictions (NCEP) is presented. In all three SSI computational spaces, the spectral coefficient space, the spectral transform space and the Gaussian grid space, different special designed parallel-distributed structures are used to enhance its performance to meet th... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Applying Web service and Windows clustering for high volume risk analysis

    Publication Year: 2005, Page(s):8 pp. - 138
    Cited by:  Patents (1)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (571 KB) | HTML iconHTML

    We present the development of a distributed system to calculate the Value at Risk (VaR) measure when a large number of users are presented. A scalable architecture based on Windows clustering and Web services is proposed. In addition, we develop a load balancing algorithm to distribute the workload among the compute nodes in the Windows cluster. The experimental results show that our system can su... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.