This is an archived page of the 2004 conference

program info

LCI's 5th International Conference, Linux Clusters: The HPC Revolution 2004, was held in Austin, Texas. The 4-day event, which included the conference and tutorials, featured a broad range of presentations and papers from HPC and large-scale cluster computing professionals in industry, academia, and government. Speakers addressed efforts to integrate and develop science and engineering applications for large-scale clusters, to achieve maximum performance and scalability. A full day of tutorials will be offered as part of the overall program.

The technical program included both peer-reviewed paper presentations and discipline- and industry-specific sessions that offered researchers and industry leaders the opportunity to present their experiences with applications. tools, user environments, and administration of large-scale clusters. Presentations from industry were specifically invited from the petroleum and geophysics, bioinformatics, aerospace and automotive, and finance sectors.

monday | tuesday | wednesday | thursday


Monday, May 17, 2004

 
8:30am Tutorial A-I:

Application Performance Analysis Tools for Linux Clusters
Felix Wolf, University of Tennessee, USA; Rick Kufrin, NCSA/UIUC, USA; Philip J. Mucci, Royal Institute of Technology, Sweden
Tutorial S-I:  

Clustermatic: An Innovative Approach to Simplified Cluster Computing
Gregory R. Watson, Ronald Minnich, Erik A. Hendriks, Matthew J. Sottile, Los Alamos National Lab, USA
10:00am Break 
10:30am Tutorial A-I: (Continued)

Application Performance Analysis Tools for Linux Clusters
Felix Wolf, University of Tennessee, USA; Rick Kufrin, NCSA/UIUC, USA; Philip J. Mucci, Royal Institute of Technology, Sweden
Tutorial S-I: (Continued)  

Clustermatic: An Innovative Approach to Simplified Cluster Computing
Gregory R. Watson, Ronald Minnich, Erik A. Hendriks, Matthew J. Sottile, Los Alamos National Lab, USA
12:00pm Lunch 
1:30pm Tutorial A-II:  

LCI Applications Tutorial: IA32, IA64 and Opteron Architecture and Programming Techniques for the Performance Programmer
John Towns, NCSA/University of Illinois, USA, and David Klepacki, IBM, USA
Tutorial S-I: (Continued)  

Clustermatic: An Innovative Approach to Simplified Cluster Computing
Gregory R. Watson, Ronald Minnich, Erik A. Hendriks, Matthew J. Sottile, Los Alamos National Lab, USA
3:00pm Break
3:30pm Tutorial A-II: (Continued)

LCI Applications Tutorial: IA32, IA64 and Opteron Architecture and Programming Techniques for the Performance Programmer
John Towns, NCSA/University of Illinois, USA, and David Klepacki, IBM, USA
Tutorial S-I: (Continued)  

Clustermatic: An Innovative Approach to Simplified Cluster Computing
Gregory R. Watson, Ronald Minnich, Erik A. Hendriks, Matthew J. Sottile, Los Alamos National Lab, USA
5:00pm End of Tutorials
   

Tuesday, May 18, 2004

8:30am Welcome to the Conference
8:45am Plenary Session I:
Brian Ropers-Huilman
Center for Computation and Technology, Louisiana State University
10:00am Break 
10:30am Cluster Health

A Failure Predictive and Policy-Based High Availability Strategy for Linux High Performance Computing Cluster
Chokchai Leangsuksun, Louisiana Tech University, USA

Listening to Your Cluster with LoGS
Jim Prewett, HPC@UNM, University of New Mexico, USA
12:00pm Lunch 
1:30pm Applications Papers I:  

Dynamic Load-Balancing Algorithm Porting on MIMD Machines
Francisco Muniz, CDTN/CNEN, Brazil

Optimizing Linux Cluster Performance by Exploring the Correlation between Application Characteristics and Gigabit Ethernet Device Parameters
Onur Celebioglu, Dell Inc., USA

Performance Analysis of a Hybrid Parallel Linear Algebra Kernel
Sue Goudy, New Mexico Institute of Mining and Technology, USA
Systems Papers I:  

Unified Heterogeneous HPCC Hardware Management Framework
Ying-Chin Fang, Dell Inc., USA

Cluster Security as a Unique Problem with Emergent Properties: Issues and Techniques
Joseph Greenseid, NCSA/University of Illinois, USA

Batch System Deployment on a Production Terascale Cluster
Karl W. Schulz, TACC/University of Texas, USA
3:00pm Break
3:30pm Vendor Presentations I:

The File System Challenge in HPC
Ben Rosen, Dell Inc., USA

Smart Interconnect: Recent Developments in Myricom Hardware and Software
Patrick Geoffray, Myricom, Inc.
5:00pm End of Presentations
6:00-8:00pm Conference Reception at the Renaissance Hotel
       

Wednesday, May 19, 2004

8:45am Plenary Session II:

Transforming the Sensing and Prediction of Intense Local Weather Through Dynamic Adaptation: People and Technologies Interacting with the Atmosphere
Kelvin Droegemeier
University of Oklahoma, USA
10:00am Break 
10:30am Applications Papers III: Performance Measurement

PerfSuite: An Accessible, Open Source Performance Analysis Environment for Linux Development and Performance
Rick Kufrin, NCSA/University of Illinois, USA

Development and Performance Analysis of a Simulation-Optimization Framework on TeraGrid Linux Clusters
Baha Y. Mirghani, North Carolina State University, USA

Optimizing Performance on Linux Clusters Using Advanced Communication Protocols: Achieving Over 10 Teraflops on an 8.6 Teraflops Linpack-Rated Linux Cluster
Manojkumar Krishnan, Pacific Northwest National Laboratory, USA
Systems Papers III: Cluster Management Systems

Concept and Implementation of CLUSTERIX: National Cluster of Linux Systems
Roman Wyrzykowski, Czestochowa University of Technology, Poland

HA-Rocks: A Cost-Effective High-Availability System for Rocks-Based Linux HPC Cluster
Tong Liu, Dell, USA

A Specialized Approach for HPC System Software
Ron Brightwell, Sandia National Laboratories, USA
12:00pm Lunch 
1:00pm Vendor Presentations VI:
Patrick Geoffray, Myricom
1:30pm Vendor Presentations VII:
Cooling for Ultra-High Density Racks and Blade Servers
Richard Sawyer, American Power Conversion (APC)
2:00pm Applications Papers IV: Applications, Visualization

Scalable Visualization Clusters Using Sepia Technology
Jim Kapadia, Hewlett Packard Corporation, USA

The Case for an MPI ABI
Greg Lindahl, Pathscale, USA
Systems Papers IV: Monitoring and Detection

Deploying LoGS to Analyze Console Logs on an IBM JS20
James E. Prewett, HPC@UNM, USA

Detection of Privilege Escalation for Linux Cluster Security
Michael Treaster, NCSA/University of Illinois, USA
3:00pm Break
3:30pm Vendor Presentations VIII:
TBD, Cray
4:00pm Vendor Presentations IX:
AMD's Road to Dual Core
Doug O'Flaherty, AMD
4:30pm Vendor Presentations X:
Doug Miles, Portland Group (PGI)
5:00pm Posters
6:00-9:30pm Tour of Local Facilities

Thursday, May 20, 2004

9:00am Plenary Session III:  
Dr. Reza Rooholamini
Director, Enterprise Solutions, Dell Product Group, USA
10:15am Break 
10:30am Applications Papers II:  Education and Training

In Search of Clusters for High-Performance Computing Education
Paul Gray, University of Northern Iowa, USA

Classroom Exercises for Grid Services
Amy Apon, University of Arkansas, USA

Automating the Large-Scale Collection and Analysis of Performance Data on Linux Clusters
Rick Kufrin, NCSA/University of Illinois, USA
Systems Papers II:  Processor and File System Performance

Performance Characteristics of Dual-Processor HPC Systems Based on 64-bit Commodity Processors
Chona S. Guiang, TACC/University of Texas, USA

An Analysis of State-of-the-Art Parallel File System for Linux
Martin W. Margo, SDSC/University of California - San Diego, USA

Comparing Linux Clusters for the Community Climate System Model
Matthew Woitaszek, University of Colorado - Boulder, USA
12:00am Lunch 
1:30pm Panel Discussion:

The Future of Linux on HPC Clusters
Moderator: Robert Ballance, Sandia National Laboratories
Panelists: Pete Beckman, Argonne National Laboratories; Dilma Da Silva, IBM Research; Barney Maccabe, University of New Mexico; and Ron Minnich, LANL
3:00pm Adjourn Conference