Dan Reed, NCSA and UIUC
Clusters: Challenges and Opportunities
Communications HardwareEach presentation is for 15 minutes. The last 15 minutes of the session is for discussion.
Session Chair: Mitchell Gusat, IBM Zurich Research Laboratory
- Cost/Performance Tradeoffs in Network Interconnects for Clusters of Commodity PCs, Christian Kurmann, Felix Rauch and Thomas M. Stricker (ETH Zurich, Switzerland) Presentation Slides
- A Criteria for Cost Optimal Construction of Irregular Networks, Geir Horn, Olav Lysne and Tor Skeie (Simula Research Laboratory, Norway) Presentation Slides (in Microsoft Power Point Slide Show format)
- A Solution for Handling Hybrid Traffic in Clustered Environments: The MultiMedia Router MMR, Blanca Caminero, Carmen Carrion, Francisco J. Quiles (UCLM, Spain), Jose Duato (UPV, Spain) and Sudhakar Yalamanchili (Georgia Tech.) Presentation Slides
Network Interfaces and Collective CommunicationEach presentation is for 15 minutes. The last 15 minutes of the session is for discussion.
Session Chair: Olav Lysne, Univ. of Oslo, Norway
- A New DMA Registration Strategy for Pinning-Based High Performance Networks, Christian Bell and Dan Bonachea (Univ. of California at Berkeley) Presentation Slides
- Efficient Broadcasts and Simple Algorithms for Parallel Linear Algebra Computing in Clusters, Fernando G. Tinetti (Universidad Nacional de La Plata, Argentina) and Emilio Luque (Universidad Autonoma de Barcelona, Spain) Presentation Slides
- Optimizing Synchronization Operations for Remote Memory Communication Systems, Darius Buntinas (Ohio State Univ.), Amina Saify (Dell), Dhabaleswar K. Panda (Ohio State Univ.), and Jarek Nieplocha (Pacific Northwest National Lab)
Communication LibrariesEach presentation is for 15 minutes. The last 15 minutes of the session is for discussion.
Session Chair: Scott Pakin, Los Alamos National Laboratory
- A message Passing Interface Library for Inhomogeneous Coupled Clusters, Martin Poeppe, Silke Schuch, and Thomas Bemmerl (Aachen University of Technology, Germany) Presentation Slides
- The Lightweight Protocol CLIC on Gigabit Ethernet, A. F. Díaz, J. Ortega, A. Cañas, F. J. Fernández, M. Anguita, and A. Prieto (Univ. of Granada, Spain)
- Implementing TreadMarks over GM on Myrinet: Challenges, Design Experience, and Performance Evaluation, Ranjit Noronha and Dhabaleswar K. Panda (Ohio State Univ.) Presentation Slides
System ServicesEach presentation is for 15 minutes. The last 10 minutes of the session is for discussion.
Session Chair: Adolfy Hoisie, Los Alamos National Laboratory
- Intelligent Architecture for Automatic Resource Allocation in Computer Clusters, Sophia Corsava and Vladimir Getov (Univ. of Westminister, U.K.) Presentation Slides
- AmpNet -- A Highly Available Cluster Interconnection Network, Amy Apon (Univ. of Arkansas) and Larry Wilbur (Belobox Networks) Presentation Slides
Performance EvaluationEach presentation is for 15 minutes. The last 10 minutes of the session is for discussion.
Session Chair: Pete Wyckoff, Ohio Supercomputer Center
- Performance Evaluation of COWs under Real Parallel Applications, Jose C. Sancho, Juan C. Martínez, Antonio Robles, Pedro López, Jose Flich, and Jose Duato (Universidad Politechnica de Valencia, Spain)
- A Performance Analysis of 4X InfiniBand Data Transfer Operations, Ariel Cohen (Topspin Communications) Presentation Slides
Title: Cluster Interconnect Technologies: What are the Top Three Limitations?
Description: Despite recent advancements, cluster interconnects continue to be a limiting factor in building and using large-scale clusters. The performance of cluster interconnects has only recently begun to match the performance of custom interconnects deployed in parallel machines nearly ten years ago. However, there are many other factors that can affect the performance and usability of cluster interconnects. In order for large-scale clusters to remain a viable computational platform, these limitations must be addressed.
The following questions were posed to the panelists: Given the current state of cluster interconnect technology, what are the top three technologies, or parts thereof, that are limitating cluster interconnects (networks, communication layers, libraries, programming models, etc.). What features and improvements are needed in communication subsystems to build next generation clusters? Assuming full freedom, what improvements need to be done to the communication subsystem to build next generation clusters?
Each panelist was asked to indicate three items in each catergory and provide justification for it. After the panelists asserted their positions, the attendees (as well as the panelists) cross-examined the panelists. Towards the end of this panel, we took a vote from the attendees to see what list they came up with and how it agreed with the opinion by the panelists.
Moderator: Ron Brightwell, Sandia National Labs
Mitchell Gusat, IBM Zurich Presentation Slides
Jarek Nieplocha, PNNL
Dan Reed, NCSA/UIUC Presentation Slides
Tom Sterling, CalTech/JPL Presentation Slides
Thomas Stricker, ETH Zurich Presentation Slides
Deadline for hotel reservation is January 31, 2003 and deadline for advance registration is April 7, 2003.
A large number of research groups from academia, industry, and research labs are currently engaged in the above research directions. The goal of this workshop is to bring together researchers and practitioners working in the areas of communication, I/O, and architecture to discuss state-of-the-art solutions as well as future trends for designing scalable, high-performance, and cost-effective communication and I/O architectures for clusters.
The first two workshops in this series (CAC '01 and CAC '02) were held in conjunction with IPDPS conferences, and they were very successful. The CAC '03 workshop plans to continue this tradition.
Paper submission: November 4, 2002 Notification of acceptance: December 20, 2002 Camera-ready due: January 24, 2003
Last updated May 4, 2003