Euro-Par 2009 - Parallel Processing: 15th International Euro-Par Conference, Delft, The Netherlands, August 25-28, 2009, Proceedings (Lecture Notes in Computer Science (5704))

This book constitutes the refereed proceedings of the 15th International Conference on Parallel Computing, Euro-Par 2009, held in Delft, The Netherlands, in August 2009. The 85 revised papers presented were carefully reviewed and selected from 256 submissions. The papers are organized in topical sections on support tools and environments; performance prediction and evaluation; scheduling and load balancing; high performance architectures and compilers; parallel and distributed databases; grid, cluster, and cloud computing; peer-to-peer computing; distributed systems and algorithms; parallel and distributed programming; parallel numerical algorithms; multicore and manycore programming; theory and algorithms for parallel computation; high performance networks; and mobile and ubiquitous computing. Front Matter....Pages - Multicore Programming Challenges....Pages 1-2 Ibis: A Programming System for Real-World Distributed Computing....Pages 3-3 What Is in a Namespace?....Pages 4-4 Front Matter....Pages 5-5 Introduction....Pages 7-8 Atune-IL: An Instrumentation Language for Auto-tuning Parallel Applications....Pages 9-20 Assigning Blame: Mapping Performance to High Level Parallel Programming Abstractions....Pages 21-32 A Holistic Approach towards Automated Performance Analysis and Tuning....Pages 33-44 Pattern Matching and I/O Replay for POSIX I/O in Parallel Programs....Pages 45-56 An Extensible I/O Performance Analysis Framework for Distributed Environments....Pages 57-68 Grouping MPI Processes for Partial Checkpoint and Co-migration....Pages 69-80 Process Mapping for MPI Collective Communications....Pages 81-92 Front Matter....Pages 93-93 Introduction....Pages 95-96 Stochastic Analysis of Hierarchical Publish/Subscribe Systems....Pages 97-109 Characterizing and Understanding the Bandwidth Behavior of Workloads on Multi-core Processors....Pages 110-121 Hybrid Techniques for Fast Multicore Simulation....Pages 122-134 PSINS: An Open Source Event Tracer and Execution Simulator for MPI Applications....Pages 135-148 A Methodology to Characterize Critical Section Bottlenecks in DSM Multiprocessors....Pages 149-161 Front Matter....Pages 163-163 Introduction....Pages 165-165 Dynamic Load Balancing of Matrix-Vector Multiplications on Roadrunner Compute Nodes....Pages 166-177 A Unified Framework for Load Distribution and Fault-Tolerance of Application Servers....Pages 178-190 Front Matter....Pages 163-163 On the Feasibility of Dynamically Scheduling DAG Applications on Shared Heterogeneous Systems....Pages 191-202 Steady-State for Batches of Identical Task Trees....Pages 203-215 A Buffer Space Optimal Solution for Re-establishing the Packet Order in a MPSoC Network Processor....Pages 216-227 Using Multicast Transfers in the Replica Migration Problem: Formulation and Scheduling Heuristics....Pages 228-240 A New Genetic Algorithm for Scheduling for Large Communication Delays....Pages 241-252 Comparison of Access Policies for Replica Placement in Tree Networks....Pages 253-264 Scheduling Recurrent Precedence-Constrained Task Graphs on a Symmetric Shared-Memory Multiprocessor....Pages 265-280 Energy-Aware Scheduling of Flow Applications on Master-Worker Platforms....Pages 281-292 Front Matter....Pages 293-293 Introduction....Pages 295-296 Last Bank: Dealing with Address Reuse in Non-Uniform Cache Architecture for CMPs....Pages 297-308 Paired ROBs: A Cost-Effective Reorder Buffer Sharing Strategy for SMT Processors....Pages 309-320 REPAS: Reliable Execution for Parallel ApplicationS in Tiled-CMPs....Pages 321-333 Impact of Quad-Core Cray XT4 System and Software Stack on Scientific Computation....Pages 334-344 Front Matter....Pages 345-345 Introduction....Pages 347-348 Unifying Memory and Database Transactions....Pages 349-360 A DHT Key-Value Storage System with Carrier Grade Performance....Pages 361-374 Selective Replicated Declustering for Arbitrary Queries....Pages 375-386 Front Matter....Pages 387-387 Introduction....Pages 389-389 POGGI: Puzzle-Based Online Games on Grid Infrastructures....Pages 390-403 Enabling High Data Throughput in Desktop Grids through Decentralized Data and Metadata Management: The BlobSeer Approach....Pages 404-416 Front Matter....Pages 387-387 MapReduce Programming Model for .NET-Based Cloud Computing....Pages 417-428 The Architecture of the XtreemOS Grid Checkpointing Service....Pages 429-441 Scalable Transactions for Web Applications in the Cloud....Pages 442-453 Provider-Independent Use of the Cloud....Pages 454-465 MPI Applications on Grids: A Topology Aware Approach....Pages 466-477 Front Matter....Pages 479-479 Introduction....Pages 481-482 A Least-Resistance Path in Reasoning about Unstructured Overlay Networks....Pages 483-497 SiMPSON: Efficient Similarity Search in Metric Spaces over P2P Structured Overlay Networks....Pages 498-510 Uniform Sampling for Directed P2P Networks....Pages 511-522 Adaptive Peer Sampling with Newscast....Pages 523-534 Exploring the Feasibility of Reputation Models for Improving P2P Routing under Churn....Pages 535-547 Selfish Neighbor Selection in Peer-to-Peer Backup and Storage Applications....Pages 548-560 Zero-Day Reconciliation of BitTorrent Users with Their ISPs....Pages 561-573 Surfing Peer-to-Peer IPTV: Distributed Channel Switching....Pages 574-586 Front Matter....Pages 587-587 Introduction....Pages 589-589 Distributed Individual-Based Simulation....Pages 590-601 A Self-stabilizing K-Clustering Algorithm Using an Arbitrary Metric....Pages 602-614 Active Optimistic Message Logging for Reliable Execution of MPI Applications....Pages 615-626 Front Matter....Pages 627-627 Introduction....Pages 629-629 A Parallel Numerical Library for UPC....Pages 630-641 Front Matter....Pages 627-627 A Multilevel Parallelization Framework for High-Order Stencil Computations....Pages 642-653 Using OpenMP vs. Threading Building Blocks for Medical Imaging on Multi-cores....Pages 654-665 Parallel Skeletons for Variable-Length Lists in SkeTo Skeleton Library....Pages 666-677 Stkm on Sca : A Unified Framework with Components, Workflows and Algorithmic Skeletons....Pages 678-690 Grid-Enabling SPMD Applications through Hierarchical Partitioning and a Component-Based Runtime....Pages 691-703 Reducing Rollbacks of Transactional Memory Using Ordered Shared Locks....Pages 704-715 Front Matter....Pages 717-717 Introduction....Pages 719-720 Wavelet-Based Adaptive Solvers on Multi-core Architectures for the Simulation of Complex Systems....Pages 721-734 Localized Parallel Algorithm for Bubble Coalescence in Free Surface Lattice-Boltzmann Method....Pages 735-746 Fast Implicit Simulation of Oscillatory Flow in Human Abdominal Bifurcation Using a Schur Complement Preconditioner....Pages 747-759 A Parallel Rigid Body Dynamics Algorithm....Pages 760-771 Optimized Stencil Computation Using In-Place Calculation on Modern Multicore Systems....Pages 772-784 Parallel Implementation of Runge–Kutta Integrators with Low Storage Requirements....Pages 785-796 PSPIKE: A Parallel Hybrid Sparse Linear System Solver....Pages 797-808 Out-of-Core Computation of the QR Factorization on Multi-core Processors....Pages 809-820 Adaptive Parallel Householder Bidiagonalization....Pages 821-833 Front Matter....Pages 835-835 Introduction....Pages 837-838 Tile Percolation: An OpenMP Tile Aware Parallelization Technique for the Cyclops-64 Multicore Processor....Pages 839-850 An Extension of the StarSs Programming Model for Platforms with Multiple GPUs....Pages 851-862 StarPU : A Unified Platform for Task Scheduling on Heterogeneous Multicore Architectures....Pages 863-874 Front Matter....Pages 835-835 XJava: Exploiting Parallelism with Object-Oriented Stream Programming....Pages 875-886 JCUDA: A Programmer-Friendly Interface for Accelerating Java Programs with CUDA....Pages 887-899 Fast and Efficient Synchronization and Communication Collective Primitives for Dual Cell-Based Blades....Pages 900-911 Searching for Concurrent Design Patterns in Video Games....Pages 912-923 Parallelization of a Video Segmentation Algorithm on CUDA–Enabled Graphics Processing Units....Pages 924-935 A Parallel Point Matching Algorithm for Landmark Based Image Registration Using Multicore Platform....Pages 936-947 High Performance Matrix Multiplication on Many Cores....Pages 948-959 Parallel Lattice Basis Reduction Using a Multi-threaded Schnorr-Euchner LLL Algorithm....Pages 960-973 Efficient Parallel Implementation of Evolutionary Algorithms on GPGPU Cards....Pages 974-985 Front Matter....Pages 987-987 Introduction....Pages 989-989 Implementing Parallel Google Map-Reduce in Eden....Pages 990-1002 A Lower Bound for Oblivious Dimensional Routing....Pages 1003-1010 Front Matter....Pages 1011-1011 Introduction....Pages 1013-1014 A Case Study of Communication Optimizations on 3D Mesh Interconnects....Pages 1015-1028 Implementing a Change Assimilation Mechanism for Source Routing Interconnects....Pages 1029-1039 Dependability Analysis of a Fault-Tolerant Network Reconfiguring Strategy....Pages 1040-1051 RecTOR: A New and Efficient Method for Dynamic Network Reconfiguration....Pages 1052-1064 NIC-Assisted Cache-Efficient Receive Stack for Message Passing over Ethernet....Pages 1065-1077 A Multipath Fault-Tolerant Routing Method for High-Speed Interconnection Networks....Pages 1078-1088 Hardware Implementation Study of the SCFQ-CA and DRR-CA Scheduling Algorithms....Pages 1089-1100 Front Matter....Pages 1101-1101 Introduction....Pages 1103-1103 Optimal and Near-Optimal Energy-Efficient Broadcasting in Wireless Networks....Pages 1104-1115 Back Matter....Pages - Euro-Par is an annual series of international conferences dedicated to the p- motion and the advancement of all aspects of parallel computing. In Euro-Par, the ?eld of parallel computing is divided into the four broad categories of t- ory, high performance, cluster and grid, and distributed and mobile computing. These categories are further subdivided into 14 topics that focus on particular areas in parallel computing. The objective of Euro-Par is to provide a forum for promoting the development of parallel computing both as an industrial te- nique and as an academic discipline, extending the frontier of both the state of the art and the state of the practice. The target audience of Euro-Par c- sists of researchers in parallel computing in academic departments, government laboratories, and industrial organizations. Euro-Par 2009 was the 15th conference in the Euro-Par series, and was - ganized by the Parallel and Distributed Systems Group of Delft University of Technology in Delft, The Netherlands. The previous Euro-Par conferences took placeinStockholm,Lyon,Passau,Southampton,Toulouse,Munich,Manchester, Paderborn,Klagenfurt,Pisa,Lisbon, Dresden, Rennes, and Las Palmasde Gran Canaria. Next year, the conference will be held in Sorrento, Italy. More inf- mation on the Euro-Par conference series and organization is available on its website athttp://www.europar.org.