Cluster2002 Contributed Papers

For the detailed schedule, please see the full program.

Managing Tasks I

ZENTURIO: An Experiment Management System for Cluster and Grid Computing

An Agent-Based Infrastructure for Parallel Java on Heterogeneous Clusters

Shell over a Cluster (SHOC): Towards Achieving Single System Image via the Shell

Operating and Using Clusters

Supermon: A High-Speed Cluster Monitoring System

Leveraging Standard Core Technologies to Programmatically Build Linux Cluster Appliances

Clusters as Large-Scale Development Facilities

Network Hardware I

Experience in Offloading Protocol Processing to a Programmable NIC

Supporting MPI Collective Communication on Network Processors

Efficient Barrier Using Remote Memory Operations on VIA-based Clusters

Managing Tasks II

I-Cluster: The Execution Sandbox

BioOpera: Cluster-aware Computing

Selective Buddy Allocation for Scheduling Parallel Jobs on Clusters

Applications

I/O Analysis and Optimization for an AMR Cosmology Application

Parallel Pattern Identification in Biological Sequences on Clusters

Performance and Implementation of Distributed Data CPHF and SCF

Network Hardware II

Design and Implementation of CC-NUMA Card II for SCI-Based PC Clustering

Impact of On-Demand Connection Management in MPI over VIA

MyVIA: A Design and Implementation of the High-Performance Virtual Interface Architecture

Programming Clusters I

High Performance User-Level Sockets over Gigabit Ethernet

Protocol-Dependent Message-Passing Performance on Linux Clusters

Mixed Mode Matrix Multiplication

Scalable Clusters I

Cplant Runtime System Support for Multi-Processor and Heterogeneous Compute Nodes

Scalable Cluster Administration - Chiba City Approach and Lessons Learned

Disaster Tolerant Wolfpack Geo-Clusters

Novel and Innovative Hardware

Memory Mapped Networks: A New Deal for Distributed Shared Memories?

Adaptive Message Management Using a Hybrid Channel Model in a Parallel File System

The Bladed Beowulf: A Cost-Effective Alternative to Traditional Beowulfs

Programming Clusters II

Goals Driving Design: PVM and MPI

Design and Validation of Portable Communication Infrastructure for Fault-Toleran t Cluster Middleware

A Parallelization Technique that Improves Performance and Cluster Utilization Efficiency for Heterogeneous Clusters of Workstations

Scalable Clusters II

An Extensible, Portable, Scalable Cluster Management Software Architecture

Processor Allocation on Cplant: Achieving General Processor Locality Using One-Dimensional Allocation Scalable Resource Management in High Performance Computers

Network Software

Reliable Blast UDP: Predictable High Performance Bulk Data Transfer

Integrated Admission and Congestion Control for QoS Support in Clusters

Socket Cloning for Cluster-Based Web Servers