Cluster 2004

This year's program contains 1 workshop, 3 half-day tutorials, 48 paper presentations in 16 sessions with 2 parallel tracks, and 3 outstanding keynote speakers, as seen below in the preliminary program.

Sunday, September 19
5:00pm - 7:00pm	Conference and Tutorial Registration (Brittany Room)

Monday, September 20
7:30am - 5:00pm	Conference and Tutorial Registration (Brittany Room)
8:30am - 10:00am	Morning Tutorials (Garden Salon 1 and 2)		RAIT Workshop (Windsor/Hampton Room)
	Parallel I/O: Lessons learnt in the last 20 years (1st half) T. Cortes	Building Highly Available HPC Clusters with HA-OSCAR (1st half) C. Leangsuksun, I. Haddad	The Remote Direct Memory Access Networks for Software DSMs Hongda Yin, Mingchang Hu, Gang Shi, Weiwu Hu, and Zhimin Tang, Institute of Computing Technology, Chinese Academy of Sciences, P. R. China. Exploiting Remote Memory Operations to Design Efficient Reconfiguration for Shared Data-Centers over InfiniBand P. Balaji, K. Vaidyanathan, S. Narravula, K. Savitha, H. - W. Jin, and D. K. Panda, Computer Science and Engineering, The Ohio State University, USA. Reducing Diff Overhead in Software DSM Systems Using RDMA Operations in InfiniBand R. Noronha and D. K. Panda, Computer Science and Engineering, The Ohio State University, USA.
10:00am - 10:30am	Break and Refreshments (Regency Courtyard)
10:30am - 12:00pm	Morning Tutorials(Garden Salon 1 and 2)		RAIT Workshop (Windsor/Hampton Room)
	Parallel I/O: Lessons learnt in the last 20 years (2nd half) T. Cortes	Building Highly Available HPC Clusters with HA-OSCAR (2nd half) C. Leangsuksun, I. Haddad	Software RDMA over TCP/IP on a General Purpose CPU Steven R. King and Frank L. Berry, Intel Corporation, USA. Sockets and RDMA Interface over 10-Gigabit Networks: An In-Depth Analysis of the Memory Traffic Bottleneck P. Balaji, D. K. Panda, Computer Science and Engineering, The Ohio State University, USA and Hemal V. Shah, Intel Corporation, USA. Open RDMA Project: Building an RDMA Ecosystem for Linux Venkata Jagana, Linux Technology Center, IBM, USA, Bernard Metzler, and Fredy Neeser, Zurich Research Laboratory, IBM.
12:00pm - 1:00pm	Lunch on your own
1:00pm - 2:30pm	Afternoon Tutorials (Garden Salon 1 and 2)		RAIT Workshop (Windsor/Hampton Room)
	State of InfiniBand in Designing HPC Clusters, Storage/File Systems, and Datacenters (1st half) D.K. Panda	MPI Tuning with Intel© Trace Analyzer and Intel© Trace Collector[TUTORIAL CANCELLED]	Protocol Off-load and RDMA Technology: Architecture and Design Considerations Michael Krause, Hewlett-Packard, USA. Mapping Socket Direct Protocol (SDP) Between iWARP and InfiniBand Jim Pinkerton and Stanley Feng, Microsoft Corporation, USA.
2:30pm - 3:00pm	Break and Refreshments (Regency Courtyard)
3:00pm - 4:30pm	Afternoon Tutorials(Garden Salon 1 and 2)
	State of InfiniBand in Designing HPC Clusters, Storage/File Systems, and Datacenters (2nd half) D.K. Panda	MPI Tuning with Intel© Trace Analyzer and Intel© Trace Collector[TUTORIAL CANCELLED]

Tuesday, September 21
7:30am - 5:00pm	Conference Registration (Brittany Room)
8:00am - 8:45pm	Breakfast (Charlies Patio)
8:45 - 9:10am	Welcome and Opening Remarks (Windsor/Hampton Room)
9:10am - 10:00am	Keynote (Windsor/Hampton Room) Using Clusters in Biomedical Image Analysis: Strategies for Neuroinformatics Research S. Pieper (Harvard University, USA) Chair: Daniel Katz
10:00am - 10:30am	Break and Refreshments (Garden Ballroom)
10:30am - 12:00pm	Session 1A (Windsor/Hampton Room) Optimizing with MPI Chair: Mark Baker	Session 1B (Sheffield Room) Scheduling I Chair: Cho Li Wang
	MPIIMGEN-A Code Transformer that parallelizes Image processing codes to run on a COW [CANCELLED] V. Varma, P. Baruah The Design and Implementation of an Asynchronous Communication Mechanism for the MPI Communication Model M. Matsuda, T. Kudoh, H. Tazuka, Y. Ishikawa Communicating Efficiently on Cluster Based Grids with MPICH-VMI A. Pant, H. Jafri	On Fairness in Distributed Job Scheduling Across Multiple Sites G. Sabin, V. Sahasrabudhe, P. Sadayappan Bandwidth-aware Co-allocating Meta-schedulers for Mini-grid Architectures W. Jones, L. Pang, D. Stanzione, W. Ligon A Comparison of Local and Gang Scheduling on a Beowulf Cluster P. Strazdins, J. Uhlmann
12:00pm - 1:30pm	Lunch sponsored by Intel (Charlies Patio)
1:30pm - 3:00pm	Session 2A (Windsor/Hampton Room) Parallel I/O and Efficient Communications Chair: Dhabeleswar (DK) Panda	Session 2B (Sheffield Room) Fault Tolerance Chair: Box Chokchai Leangsuksun
	A Parallel I/O Mechanism for Distributed Systems T. Baer, P. Wyckoff RFS: Implementing Efficient and Flexible Remote File Access for MPI-IO J. Lee, X. Ma, R. Ross, R. Thakur, M. Winslett An Efficient End-host Architecture for Cluster Communication Services X. Qi, G. Parmer, R. West	FTC-Charm++: An In-Memory Checkpoint-Based Fault Tolerant Runtime for Charm++ and MPI G. Zheng, L. Shi, L. V. Kale Fault-tolerant Grid Services Using Primary-Backup: Feasibility and Performance X. Zhang, D. Zagorodnov, M. Hiltunen, K. Marzullo, R. Schlichting Improved Message logging versus Improved coordinated checkpointing for fault tolerant MPI P. Lemarinier, A. Bouteiller, T. Herault, G. Krawezik, F. Cappello
3:00pm - 3:30pm	Break and Refreshments (Garden Ballroom)
3:30pm - 5:00pm	Session 3A (Windsor/Hampton Room) Collective Communication Optimizations Chair: Thomas Rauber	Session 3B (Sheffield Room) Novel Storage Architectures Chair: Toni Cortes
	Scalable and High Performance NIC-Based Allgather over Myrinet/GM W. Yu, D. Buntinas, D. K. Panda Efficient Barrier and Allreduce on IBA clusters using hardware multicast and adaptive algorithms A. Mamidala, J. Liu, D. K Panda On Optimizing Collective Communication E. Chan, M. Heimlich, A. Purkayastha, R. van de Geijn	RAAC: An Architecture for Scalable, Reliable Storage in Clusters M. Pillai, M. Lauria Hierarchical Bloom Filter Arrays (HBA): A Novel, Scalable Metadata Management System for Large Storage Y. Zhu, H. Jiang, J. Wang Simplifying Administration Through Dynamic Reconfiguration in a Cooperative Cluster Storage System R. Lachaize, J. S. Hansen

5:30pm - 7:30pm	Poster Session (Crescent Room) Poster List
	6:00pm - 7:30pm	Reception in Exhibitor Area (Garden Ballroom) and Poster Area (Crescent Room, Galleria One and Galleria Two)

7:00pm - 8:30pm	Steering Committee Meeting Tiffany's room, Trellises Restaurant.
8:30pm - 9:30pm	Steering Committee Dinner Tiffany's room, Trellises Restaurant.

Wednesday, September 22
8:00am - 5:00pm	Conference Registration (Brittany Room)
8:00am - 9:00pm	Breakfast (Charlies Patio)
9:00 - 9:10am	Welcome and Opening Remarks (Windsor/Hampton Room)
9:10am - 10:00am	Keynote (Windsor/Hampton Room) PC Cluster as Main Computing Resource at Supercomputer Center Dr. Ryutaro Himeno(Riken, Japan) Chair: Philip Papadopoulos
10:00am - 10:30am	Break and Refreshments (Garden Ballroom)
10:30am - 12:00pm	Session 4A (Windsor/Hampton Room) Networking Chair : Taisuke Boku	Session 4B (Sheffield Room) Distributed Shared Memory Chair : Walt Ligon
	GNET-1: Gigabit Ethernet Network Testbed Y. Kodama, T. Kudoh, T. Takano, H. Sato, O. Tatebe, S. Sekiguchi A Comparison of 4X InfiniBand and Quadrics Elan-4 Technologies R. Brightwell, D. Doerfler, K. D. Underwood NIC-Based Offload of Dynamic User-Defined Modules for Myrinet Clusters A. Wagner, H.-W. Jin, D.K. Panda, R. Riesen	A Novel Adaptive Home Migration Protocol in Home-based DSM W. Fang, C.-L. Wang, W. Zhu, F. Lau LOTS: A Software DSM Supporting Large Object Space B. Cheung, C.-L. Wang, F. Lau Cluster Computing Environment supporting Single System Image M. Choi, D. Lee, S. Maeng
12:00pm - 1:30pm	Lunch (Charlies Patio)
1:30pm - 3:00pm	Session 5A (Windsor/Hampton Room) Scheduling II Chair: Mason Katz	Session 5B (Sheffield Room) SSI and Location Aware Algorithms Chair: Philip Papadopoulos
	Towards Provision of Quality of Service Guarantees in Job Scheduling M. Islam, P. Balaji, P. Sadayappan, D. K. Panda Implementation and Design Analysis of a Network Messaging Module Using Virtual Interface Architecture G. Amerson, A. Apon A Distributed Data Management Middleware for Data-Driven Application Systems S. Langella, S. Hastings, S. Oster, T. Kurc, U. Catalyurek, J. Saltz	Kerrighed and Data Parallelism: Cluster Computing on Single System Image Operating Systems C. Morin, R. Lottiaux, G. Vallée, P. Gallard, D. Margery, J.-Y. Berthou, I. Scherson An Evaluation of the Close-to-Files Processor and Data Clo-Allocation Policy in Multiclusters H. Mohamed, D. Epema Location Management in Object-based Distributed Computing A. Fedorov, N. Chrisochoides
3:00pm - 3:30pm	Break and Refreshments (Garden Ballroom)
3:30pm - 5:00pm	Session 6A (Windsor/Hampton Room) Systems Analysis Chair: Ronald Minnich	Session 6B (Sheffield Room) Grid Systems Chair: Avneesh Pant
	Towwards Informatic Analysis of Syslogs J. Stearley Component-Based Systems Software Architecture: A Case Study N. Desai, R. Bradshaw, E. Lusk, R. Butler Predicting Memory-Access Cost Based on Data Access Patterns S. Byna, X.-H. Sun, W. Gropp, R. Thakur	Grid Systems Deployment & Management using Rocks F. Sacerdoti, S. Chandra, K. Bhatia Computation-at-Risk:Employing the Grid for Computational Risk Management S. Kleban, S. Clearwater Management of Grid Jobs and Data within SAMGrid A. Baranovski, G. Garzoglio, I. Terekhov, A. Roy, T. Tannenbaum

6:30pm - 9:30pm	Conference Banquet sponsored by AMD at Casa Guadalajara

Thursday, September 23
8:00am - 12:00pm	Conference Registration (Brittany Room)
8:00am - 9:00pm	Breakfast (Charlies Patio)
9:00am - 10:30pm	Session 7A (Windsor/Hampton Room) Performance Analysis Chair : Patricia Kovatch	Session 7B (Sheffield Room) Visualization and Simulation Chair : Greg Bruno
	Perfomance Analysis Tools for Large Scale Clusters Z. Cvetanovic Analysis of microbenchmarks for performance tuning of clusters M. Sottile, R. Minnich NWPerf: A System Wide Performance Monitoring Tool for Large Linux Clusters R. Mooney, R. Studham, K. Schmidt, J. Nieplocha	TeraVision : A Distributed, Scalable, High Resolution Graphics Streaming System R. Singh, B. Jeong, L. Renambot, A. Johnson, J. Leigh ClusterSim: A Java-Based Parallel Discrete-Event Simulation Tool for Cluster Computing L. Góes, L. Ramos, C. Martins JuxtaView -- A Tool for Interactive Visualization of Large Imagery on Scalable Tiled Displays N. Krishnaprasad, V. Vishwanath, S. Venkataraman, A. Rao, L. Renambot, J. Leigh, A. Johnson, B. Davis
10:30am - 11:00am	Break and Refreshments (Regency Courtyard)
11:00am - 12:30pm	Session 8A (Windsor/Hampton Room) Systems Management Chair : Nadya Williams	Session 8B (Sheffield Room) Applications Techniques Chair : Amy Apon
	Rolls: Adding a First-Class Method to a Standard System Installer to Reliably Deploy Cluster Software G. Bruno, M. Katz, F. Sacerdoti, P. Papadopoulos A Client-centric Grid Knowledgebase G. Kola, T. Kosar, M. Livny Give your bootstrap the boot: Using the Operating System to boot the Operating System R. Minnich	Parallel Competitive Learning Algorithm for Fast Codebook Design on Partitioned Space S. Momose, K. Sano, K. Suzuki, T. Nakamura Implementing Parallel Conjugate Gradient on the EARTH Multithreaded Architecture F. Chen, K. B. Theobald, G. R. Gao XChange: Coupling Parallel Applications in a Dynamic Environment H. Abbasi, M. Wolf, K. Schwan, G. Eisenhauer, A. Hilton
12:30pm - 1:30pm	Lunch (Regency Courtyard)
1:30pm - 2:30pm	Endnote (Windsor/Hampton Room) The Cluster Agenda: First Achieve World Domination, then Kick Ass Dr. Thomas Sterling (Caltech , JPL, USA) Chair: Henri Casanova
2:30pm - 3:00pm	Closing Remarks (Windsor/Hampton Room)
6:00pm - 10:00pm	Social Activity: San Diego Padres vs. Los Angeles Dodgers at Petco Park, sponsored by Sun Microsystems