By Topic

Frontiers of Massively Parallel Computation, 1988. Proceedings., 2nd Symposium on the Frontiers of

Date 10-12 Oct. 1988

Filter Results

Displaying Results 1 - 25 of 125
  • Proceedings. The 2nd Symposium on the Frontiers of Massively Parallel Computation (Cat. No.88CH2649-2)

    Publication Year: 1988
    Request permission for commercial reuse | PDF file iconPDF (46 KB)
    Freely Available from IEEE
  • Artificial neural network on a SIMD architecture

    Publication Year: 1988, Page(s):43 - 47
    Cited by:  Papers (8)  |  Patents (1)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (544 KB)

    An implementation of a fully connected artificial neural network using the multilayered perceptron model is described. The neural network is implemented on a systolic array processor based on the Geometric Arithmetic Parallel Processor (GAPP) chip. Arrays of GAPP chips make up a single-instruction multiple-data (SIMD) class machine which has fine-grained connections and is fully programma... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Modeling neural networks on the MPP

    Publication Year: 1988, Page(s):39 - 42
    Cited by:  Papers (3)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (220 KB)

    A network of fixed-connection-weight neuronlike elements is simulated on the massively parallel processor (MPP) in two ways. First, the square connectivity matrix of a 128-neuron network is mapped onto the square MPP processor array. This allows a highly parallel simulation in which 128 MPP processors were active at all times. Second, a 128-by-128 array of neurons is mapped onto the 16.384 MPP pro... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Characterizing the advantages of massively parallel computing

    Publication Year: 1988, Page(s):651 - 652
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (148 KB)

    The authors discuss the advantages of massively parallel processing (MPP) in general. Algorithmic solutions for difficult problems are stressed. The cost of programming and training is also discussed View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • The Ynet: an interconnect structure for a highly concurrent data base computer system

    Publication Year: 1988, Page(s):429 - 435
    Cited by:  Papers (4)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (572 KB)

    The author describes the design considerations for the Ynet, an interconnect structure for a highly concurrent database system which also requires high availability and real-time operation. The Ynet is the basis of a multiple-instruction, multiple-data (MIMD) concurrent processing architecture which can connect up to 1024 high-performance microprocessors to form a very-high performance system. Mos... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Massively parallel data optimization

    Publication Year: 1988, Page(s):551 - 558
    Cited by:  Papers (6)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (636 KB)

    Techniques for the automatic layout of arrays in a Fortran compiler supporting Fortran 8× array features and targeted to the Connection Machine computer system are discussed. The goal is primarily to minimize the costs of moving data between processors and secondarily to minimize memory usage. Improved array layout may allow communications operations to be eliminated or to be replaced by mor... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • How to cluster in parallel with neural networks

    Publication Year: 1988, Page(s):31 - 38
    Cited by:  Papers (1)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (624 KB)

    It is shown that the problem of partitioning a net of N patterns in a d-dimensional metric space into K clusters can, in spite of its exponential complexity, be formulated as an optimization problem for which very good, but not necessary optimal, solutions can be found by using a neural network. The network starts from many randomly selected initial states. The network i... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Object-oriented prototypes of parallel architectures for the performance evaluation of algorithms

    Publication Year: 1988, Page(s):647 - 649
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (256 KB)

    The author describes a methodology for constructing prototypes of parallel architectures. The prototypes model the functionality of the components of the architecture to any specified level of detail. The prototypes are programmable; thus benchmarks of algorithms can be executed. The prototypes are also flexible, allowing the fine tuning of the architecture to enhance the performance. A specific w... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Massively parallel computing applied to the one-dimensional bin packing problem

    Publication Year: 1988, Page(s):317 - 319
    Cited by:  Papers (1)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (168 KB)

    The author presents a parallel algorithm for packing Nn2 pieces in a massively parallel computing environment of n2 processors operating in single-instruction multiple-data (SIMD) mode. The heuristic approach that is developed utilizes a grid topology that structures the input to increase packing efficiency without sorting the data set. The res... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Dynamically allocating sets of fine-grained processors to running computations

    Publication Year: 1988, Page(s):191 - 194
    Cited by:  Patents (2)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (1044 KB)

    An approach to using general-purpose parallel computers is explored which involves mapping hardware resources onto computations instead of mapping computations onto hardware. Problems such as processor allocation, task scheduling and load balancing change significantly under this approach and may become amenable to new attacks. The implementation of this approach used by the FFP machine, whose com... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Optimal tilings for iterative PDE solvers

    Publication Year: 1988, Page(s):227 - 229
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (192 KB)

    A simple, universal procedure for deriving the optimal partition for a given iterative partial differential equation (PDE) algorithm is presented. A previous analysis is extended to include geometrical considerations-in particular, the relative orientation of the update stencil and the partition tile. On the basis of this procedure, two new partitionings are derived for specific PDE algorithms whi... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Minimum spanning tree on the HMESH architecture

    Publication Year: 1988, Page(s):121 - 124
    Cited by:  Papers (2)  |  Patents (1)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (332 KB)

    A fast algorithm to compute the minimum spanning tree of a given undirected graph on a hierarchical mesh-connected computer (HMESH) is presented. The time complexity of the algorithm is O(log2 n), where n is the number of nodes in the graph. HMESH is a broadcast bus VLSI architecture which consists of n× n processing elements (PEs) in a mesh-con... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Generalization of orthogonal multiprocessor for massively parallel computation

    Publication Year: 1988, Page(s):391 - 398
    Cited by:  Papers (7)  |  Patents (2)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (456 KB)

    The authors present a multidimensional orthogonal multiprocessor (OMP) conceived for massively parallel computation. The OMP compares favorably with the generalized hypercube and spanning-bus hypercube. All three architectures extend the original concept of binary n-cube. Among the three, the OMP requires equal or less hardware and yet presents some interesting application potentials in m... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Overview and extensions of a system for routing directed graphs on SIMD architectures

    Publication Year: 1988, Page(s):63 - 67
    Cited by:  Papers (6)  |  Patents (1)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (436 KB)

    A method is given for parallelizing the problems of directed graphs on a single-instruction multiple-data (SIMD) machine model that uses only nearest-neighbor connections for communication and has no facility for local indirect addressing. Each vertex of the graph is assigned to a processor in the machine. Rules for labeling are introduced that support the use of a simple algorithm for movement of... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Compiling collection-oriented languages onto massively parallel computers

    Publication Year: 1988, Page(s):575 - 585
    Cited by:  Papers (3)  |  Patents (9)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (776 KB)

    Some issues and techniques are discussed for compiling collection-oriented languages for use on parallel and vector machines. Techniques for translating nested parallelism into flat parallelism are described. A compiler that translates a subset of Paralation Lisp into the instruction set of a virtual machine is discussed. The instructions of the virtual machine are simple vector instructions which... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Controlling and programming the SPHINX multi-SIMD pyramid machine

    Publication Year: 1988, Page(s):423 - 428
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (380 KB)

    A control scheme with a very limited overhead for SPHINX, a cellular pyramidal machine primarily designed for image-processing applications, is presented. It is organized as a set of stacked layers of decreasing size interconnected according to a dual network: a mesh-based interneighbor interconnection network within a layer and a binary tree between adjacent layers. The SPHINX processing element ... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Application of massively parallel machines to molecular dynamics simulation of free clusters

    Publication Year: 1988, Page(s):275 - 277
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (228 KB)

    Molecular dynamics is used to examine the melting of a free cluster of up to 1000 sodium and fluoride ions. An algorithm designed originally for the distributed array processor (DAP) is implemented on the Connection Machine (CM), and their performance for this problem is compared with that of the Cray 1. The CM would be twice as fast as the DAP using twice the number of processors as the DAP's 409... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Neurocomputing with optical pipeline networks

    Publication Year: 1988, Page(s):545 - 548
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (224 KB)

    The authors present techniques for realizing connectionist-style neurocomputations on Opcom, an optical computer architecture based on the pipeline networking concept. The primary operation mode of Opcom is massively parallel and pipeline processing at the gate level. Special attention is paid to the implementation of the two most important functions of any neural network: learning and search (ret... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Performance of the ASP on the DARPA architecture benchmark

    Publication Year: 1988, Page(s):483 - 486
    Cited by:  Papers (5)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (284 KB)

    The associative string processor (ASP) is a homogeneous, reconfigurable, programmable massively parallel processor which offers step-function advantages in cost performance and application flexibility due to its unique architecture and its use of state-of-the-art microelectronics. The authors briefly describe the ASP architecture and its implementation and report the results of an evaluation of it... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • The function of a connection network between hosts and processing elements in massively parallel computer systems

    Publication Year: 1988, Page(s):455 - 458
    Cited by:  Patents (1)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (280 KB)

    The authors examine the function of a connection network between massive numbers of processing elements (PEs) and a single host by comparing the data structure machine (DSM), whose major connection network is a computationally powerful binary tree with the host connected at the root, and the Connection Machine (CM), which provides a very rich and general PE-to-PE connection network, but whose conn... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Estimating water flow through a hillslope using the massively parallel processor

    Publication Year: 1988, Page(s):249 - 256
    Cited by:  Patents (2)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (600 KB)

    A two-dimensional model of water flow in a hillslope which has been implemented on the massively parallel processor is described. Flow in the soil both in the saturated and unsaturated zones, evaporation and overland flow are modeled, and the rainfall rates are allowed to vary spatially. Previous models of this type have always been very limited computationally. This model takes less than a minute... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Parallel hashed key access on the Connection Machine

    Publication Year: 1988, Page(s):643 - 645
    Cited by:  Papers (1)  |  Patents (1)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (144 KB)

    The author discusses retrieval of a number of keys in parallel from a single hash table. The algorithm is a data parallel adaptation of a familiar sequential method. Changes made to suit the single-instruction stream, multiple-data stream Connection Machine are described. Running time is proportional to the maximum run of collisions rather than the average number. The algorithm was developed as a ... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Simulations of Conposit, a supra-connectionist architecture for commonsense reasoning

    Publication Year: 1988, Page(s):311 - 315
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (424 KB)

    A computation architecture called `Conposit' is outlined. Composit manipulates very-short-term complex symbolic data structures of types that are useful in high-level cognitive tasks such as commonsense reasoning, planning, and natural language understanding. Conposit's data structures are, essentially, temporary configurations of symbolic occurrences in a two-dimensional array of registers. Each ... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Parallel implementation considerations for a class of signal processing algorithms

    Publication Year: 1988, Page(s):353 - 356
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (224 KB)

    The Burg signal-processing algorithm, which is widely used in such areas as geophysical data analysis, speech processing, and spectral analysis, is investigated in order to determine its suitability for implementation on parallel-processing architectures. The algorithm is implemented on three different machines, representing a variety of parallel architectures: Denelcor HEP, the Intel iPSC/2 hyper... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Almost linear speed-up of distributed discrete event simulations

    Publication Year: 1988, Page(s):187 - 190
    Cited by:  Papers (1)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (372 KB)

    A distributed simulation algorithm is presented which explores the topology of the simulated system using precomputed minimum propagation delays between subsystems. The algorithm also uses opaque periods which are the delays caused by the nonpreemptive states these subsystems can enter. The algorithm achieves Θ(N/log N) speed-up when run on an appropriate physically realiza... View full abstract»

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