High-Performance Computing

This book constitutes the refereed joint post-conference proceedings of the 6th International Symposium on High-Performance Computing, ISHPC 2005, held in, Japan, in 2005. It also includes the refereed post-proceedings of the First International Workshop on Advanced Low Power Systems 2006, ALPS2006, and some from the Workshop on Applications for PetaFLOPS Computing, APC 2005. A total of 42 papers were carefully selected from 76 submissions, covering a huge range of topics.

High Performance Computing.- Multiple Stream Prediction.- Enhanced Loop Coalescing: A Compiler Technique for Transforming Non-uniform Iteration Spaces.- Folding Active List for High Performance and Low Power.- Reducing Misspeculation Penalty in Trace-Level Speculative Multithreaded Architectures.- Exploiting Execution Locality with a Decoupled Kilo-Instruction Processor.- Decoupled State-Execute Architecture.- A Scalable Methodology for Computing Fault-Free Paths in InfiniBand Torus Networks.- Using a Way Cache to Improve Performance of Set-Associative Caches.- Design of Fast Collective Communication Functions on Clustered Workstations with Ethernet and Myrinet.- Dynamic Load Balancing in MPI Jobs.- Workload Characterization of Stateful Networking Applications.- Using Recursion to Boost ATLAS’s Performance.- Towards Generic Solver of Combinatorial Optimization Problems with Autonomous Agents in P2P Networks.- New Evaluation Index of Incomplete Cholesky Preconditioning Effect.- T-Map: A Topological Approach to Visual Exploration of Time-Varying Volume Data.- Cross-Line — A Globally Adaptive Control Method of Interconnection Network.- The Bandwidth Expansion Effectiveness of Cache Levels Block Prefetch.- Implementation and Evaluation of the Mechanisms for Low Latency Communication on DIMMnet-2.- Computationally Efficient Parallel Matrix-Matrix Multiplication on the Torus.- A New Dynamic Load Balancing Technique for Parallel Modified PrefixSpan with Distributed Worker Paradigm and Its Performance Evaluation.- Performance-Based Loop Scheduling on Grid Environments.- Reconfigurable Middleware for Grid Environment.- : An Enhanced Stream-Based Communication Mechanism.- Performance of Coupled Parallel Finite Element Analysis in Grid Computing Environment.- Photo-Realistic Visualization for the Blast Wave of TNT Explosion by Grid-Based Rendering.- Development of an Interactive Visual Data Mining System for Atmospheric Science.- A Calculus Effectively Performing Event Formation with Visualization.- A Similarity Evaluation Method for Volume Data Sets by Using Critical Point Graph.- Hybrid Parallelization and Flat Parallelization in HPF (High Performance Fortran).- Mapping Normalization Technique on the HPF Compiler fhpf.- Development of Electromagnetic Particle Simulation Code in an Open System.- Development of Three-Dimensional Neoclassical Transport Simulation Code with High Performance Fortran on a Vector-Parallel Computer.- Distributed Parallelization of Exact Charge Conservative Particle Simulation Code by High Performance Fortran.- Pipelined Parallelization in HPF Programs on the Earth Simulator.- Sampling of Protein Conformations with Computers to Predict the Native Structure.- Spacecraft Plasma Environment Analysis Via Large Scale 3D Plasma Particle Simulation.- PetaFLOPS Computing and Computational Nanotechnology on Industrial Issues.- 16.14 TFLOPS Eigenvalue Solver on the Earth Simulator: Exact Diagonalization for Ultra Largescale Hamiltonian Matrix.- Numerical Simulation of Combustion Dynamics at ISTA/JAXA.- Realization of a Computer Simulation Environment Based on ITBL and a Large Scale GW Calculation Performed on This Platform.- Computations of Global Seismic Wave Propagation in Three Dimensional Earth Model.- Lattice QCD Simulations as an HPC Challenge.- Energy-Efficient Embedded System Design at 90nm and Below – A System-Level Perspective –.- Empirical Study for Optimization of Power-Performance with On-Chip Memory.- Performance Evaluation of Compiler Controlled Power Saving Scheme.- Program Phase Detection Based Dynamic Control Mechanisms for Pipeline Stage Unification Adoption.- Reducing Energy in Instruction Caches by Using Multiple Line Buffers with Prediction.