Proceedings of the Fifth SIAM Conference on Parallel Processing for Scientific ComputingJ. J. Dongarra SIAM, 1992 M01 1 - 648 páginas This text gives the proceedings for the fifth conference on parallel processing for scientific computing. |
Contenido
Parallel Implementation of a Nonsymmetric Tridiagonal | 11 |
A Comparison of Algorithms for Cholesky Factorization on | 22 |
An Efficient Parallel Homotopy Algorithm for Nonsymmetric | 29 |
Solution of Structurally Identical Sparse Linear Systems Based | 41 |
A Parallel Multifrontal Method for Sparse Symmetric Definite | 47 |
Contents | 61 |
A Unified Computational Model for the Parallel Solution | 63 |
Study of an Adaptive Blocking for a Parallel Nested Dissection | 71 |
Simulation of a Cancer Progression in the Colon on a Massively | 345 |
The Parallelization of an Oil Reservoir Simulator Invited | 352 |
Multitasking of a Reservoir Simulator on the Cray XMP | 360 |
New RealTime Robot Motion Algorithms Using Parallel VLSI | 369 |
Third Generation Cellular Automaton for Modeling Excitable | 376 |
Cellular Automata Modeling Isotropic Growth of Clusters | 382 |
ThreeDimensional MOS Device Simulation on a Connection | 388 |
Data Parallel Algorithms for the Numerical Modeling | 394 |
Parallel ICCG Algorithm on Distributed Memory Architecture | 78 |
An Implementation of the GMRES Method Using QR Factorization | 84 |
Projection Methods on a Distributed Memory MIMD | 92 |
Techniques for Accelerating the Block Cimmino Method | 98 |
A Greatly Simplified Theory for Parallel Multigrid or Projection | 105 |
Mapping Implicit Spectral Methods to Distributed Memory | 111 |
Solution of Generalized Eigenvalue Problems Using a Parallel | 117 |
Iterative Methods for Nonsymmetric Systems on MIMD Machines | 123 |
Vector Preconditioned sStep Methods on the IBM 3090600S6VF | 130 |
Parallelizable Restarted Iterative Methods for Nonsymmetric | 138 |
An Iterative LanczosBased Condition Estimator | 144 |
Approaches to Physical Optimization Invited Presentation | 153 |
A Parallel Implementation of an Interior Point Method for Linear | 163 |
An Asynchronous Relaxation Method for the Parallel Simulation | 174 |
Stochastic Global Optimization Applied to Reaction Network | 180 |
Performance of a Benchmark Implementation of the Van Slyke | 186 |
An Investigation of the Effect of Problem Structure on Stochastic | 193 |
Parallel and Symbolic Computation in Finance | 199 |
A Parallelized Algorithm for the AllRow Preconditioned Interval | 205 |
Decomposition Principles and their Applications in Scientific | 213 |
Parallelization and Convergence of a 3D Implicit Unsteady | 238 |
Parallelizing Across Time When Solving TimeDependent Partial | 246 |
A MIMD Implementation of a Parallel Euler Solver | 253 |
A Collocation Based Parallel Algorithm to Solve Two Phase Flow | 263 |
Parallel Solution of the FewGroup Neutron Diffusion Equations | 269 |
Fluctuation Simulations for Stochastic Reaction Diffusion Systems | 275 |
Asynchronous Nonlinear Iteration and Domain Decomposition | 281 |
A Distributed Solution and Visualization for 3D Flow Simulation | 287 |
CFD Equations Solutions on Massively Parallel Computers | 295 |
Parallel Implementation of a Control Volume Method for Solving | 301 |
Parallel Scalability of the Spectral Transform Method | 307 |
Computational Problems and the Human Genome Project | 315 |
On Parallel Search of DNA Sequence Databases | 332 |
Parallel Molecular Dynamics | 338 |
Parallel Ocean Circulation Modeling on Cedar | 401 |
The Monte Carlo Simulation of the Random ThreeDimensional | 411 |
Performance Comparison of Several SIMD Machines | 419 |
Performance Effects of Load Imbalance in Parallel | 425 |
RandomAccess Bandwidth Requirements of Point Parallelism | 435 |
A Comparison of Load Balancing Strategies for Particle Methods | 442 |
Parallelizing Dynamic Processes on Message Passing | 451 |
Optimal Data Placement for Distributed Memory Architectures | 462 |
Parallel GraphEmbedding Heuristics | 473 |
Parallel Distributed Supercomputing in a Heterogeneous | 484 |
A Comparison of Communication Costs for Three Parallel | 491 |
Performance Issues for MessagePassing MIMD Machines | 501 |
LoadBalancing and Performance of a Gridless Particle Simulation | 507 |
Examining Locality | 513 |
FineGrained Measurements of Loop Performance on the CRAY YMP | 519 |
Implementing Efficient and Portable Dense Matrix Factorizations | 525 |
Automatic Blocking by a Compiler | 537 |
A Linear Algebraic View of Loop Transformations and Their | 543 |
Maximum Performance Code Restructuring for Hierarchical | 549 |
Translating Control Parallelism to Data Parallelism | 555 |
Cross Machine Comparisons | 564 |
Unstructured Mesh Problems PARTI Primitives and the | 570 |
An Experimental Study of the Effectiveness of High Level Parallel | 581 |
Parallelizing ScatterandGather Applications Using | 588 |
Solving Computational Grand Challenges Using a Network | 596 |
Examples in Phred | 602 |
A Modular and Scalable Course Grain Systolic Architecture | 610 |
Design of Compact High Performance Processing Elements | 616 |
Progress Report | 625 |
A Package for the Iterative Solution of Large Sparse | 631 |
Vectorization and Parallelization of FISHPAK | 637 |
A Case Study in Parallel Mathematical | 643 |
Términos y frases comunes
applied approach approximate array asynchronous block boundary calculated cells column communication complexity condition number conjugate gradient Connection Machine convergence cost Cray Cray Y-MP defined diagonal discretization distributed memory domain decomposition Domain Decomposition Methods dynamic efficiency eigenvalues evaluation execution Figure finite element Fortran global GLOWINSKI GMRES graph grid points hypercube Intel Intel iPSC/860 interval iterative methods Lanczos linear systems load balancing loop mapping Mathematics megaflops mesh Mflops minimal multigrid multiprocessor neural networks Newton's method nodes nonlinear nonsymmetric number of iterations number of processors obtained operations optimization overhead parallel algorithm parallel computers parallel implementation parameter Partial Differential Equations partitioning PDEs performance phase pivoting preconditioned preconditioner problem programming protein reduced Research residual s-step scheme sequence SIAM simulation solution solver solving sparse matrix spectral speedup step structure subdomain subregions subspace Supercomputing symmetric system of equations Table techniques triangular update values variables vector
Pasajes populares
Página 595 - HeNCE is a graphical interface which the programmer uses to perform these tasks. This tool supports visual representations of many of its functions. In HeNCE, the programmer is responsible for explicitly specifying parallelism by drawing graphs that express the parallelism. The user directly inputs the graph using tools which are part of the HeNCE environment. Each node in a HeNCE graph represents a procedure written in either Fortran or C. There are three types of arcs in the HeNCE graph. One represents...