People

Publications

Software

Sponsors

• Understanding important performance issues in MPI collective operations
• Developing topology and architecture specific communication algorithms
• Developing practical and accurate performance models for MPI collective communication algorithms
• Investigating adaptive MPI library implementation techniques
• Investigating the integrated compiler and library approach (compiled communication) for MPI optimizations.

• Xin Yuan

Current Students:

• Joshua Lawrence
• Wickus Nienaber
• Pitch Patarasuk
• Matthew Small
• Chi Zhang

Past students:

• Ahmad Faraj (PHD 2006, currently at IBM)

Equipment

• draco.cs.fsu.edu (since Summer 2007)
  • Number of nodes: 18
  • Node: Dell Poweredge 1950 (two 2.33 GHz Quad-core Intel Xeon E5345, 8 cores per node).
  • Memory: 8GB per node
  • Connectivity: (1) Infiniband DDR (20Gbps), (2) 1Gbps managed Ethernet, (3) 100Mbps Ethernet.
• cetus.cs.fsu.edu (since Fall 2004)
  • Number of nodes: 36
  • Node: Dell dimension 2400 (2.8GHz P4)
  • Memory: 640MB per node
  • connectivity: 1 Gbps and 100 Mbps reconfigurable Ethernet (8 Ethernet switches in disposal)

Journal publications

• A. Faraj, P. Patarasuk, and X. Yuan, "A Study of Process Arrival Patterns for MPI Collective Operations," International Journal of Parallel Programming, accepted.

• A. Faraj, P. Patarasuk and X. Yuan. "Bandwidth Efficient All-to-all Broadcast on Switched Clusters," International Journal of Parallel Programming, Accepted.

• P. Patarasuk, A Faraj, and X. Yuan. " Techniques for Pipelined Broadcast on Ethernet Switched Clusters," Journal of Parallel and Distributed Computing, 68(6):809-824, June 2008.

• R. G. Lane, S. Daniels and X. Yuan, " An Empirical Study of Reliable Multicast Protocols over Ethernet-Connected Networks," Performance Evaluation Journal, Vol. 64, No. 3, pages 210-228, March 2007

• A. Faraj, X. Yuan, and P. Patarasuk, "A Message Scheduling Scheme for All-to-All Personalized Communication on Ethernet Switched Clusters," IEEE Transactions on Parallel and Distributed Systems, Vol. 18, No. 2, pages 264-276, Feb. 2007

• A. Karwande, X. Yuan, and D. K. Lowenthal, "An MPI Prototype for Compiled Communication on Ethernet Switched Clusters." Journal of Parallel and Distributed Computing (JPDC), special issue on Design and Performance of Networks for Super-, Cluster-, and Grid-Computing, Vol. 65, No. 10, pages 1123-1133, October 2005.

• X. Yuan, R. Melhem and R. Gupta, "Algorithms for Supporting Compiled Communication," IEEE Transactions on Parallel and Distributed Systems, Volume 14, No. 2, pages 107-118, February 2003.

Conference Publications

• P. Patarasuk and X. Yuan, "Efficient MPI_Bcast across Different Process Arrival Patterns," the 2008 International Parallel and Distributed Processing Symposium (IPDPS), Miami, FL, April 14-18, 2008.

• J. Lawrence and X. Yuan, "An MPI Tool for Automatically Discovering the Switch Level Topologies of Ethernet Clusters," the IPDPS Workshop on System Management Techniques, Processes, and Services, Miami, FL, April 14, 2008.

• W. Nienaber, X. Yuan, Z. Duan, "On LID Assignment in Infiniband Networks," ACM/IEEE Symposium on Architectures for Networking and Communications Systems, pages 97-106, Orlando, FL, December 3-4, 2007.

• A. Faraj, P. Patarasu, and X. Yuan, "A Study of Process Arrival Patterns for MPI Collective Operations," the 21th International Conference on Supercomputing (ICS), Seattle, WA, June 16-20, 2007.

• X. Yuan, W. Nienaber, Z. Duan, and R. Melhem, "Oblivious Routing for Fat-Tree Based System Area Networks with Uncertain Traffic Demands," (full paper), ACM Sigmetrics, San Diego, CA, June 12-16, 2007.

• P. Patarasuk and X. Yuan, "Bandwidth Efficient All-reduce Operation on Tree Topologies," IEEE IPDPS Workshop on High-Level Parallel Programming Models and Supportive Environments (HIPS 2007), Long Beach, CA, March 2007.

• A. Faraj, X. Yuan, and D.K. Lowenthal, "STAR-MPI: Self Tuned Adaptive Routines for MPI Collective Operations." The 20th ACM International Conference on Supercomputing (ICS), pages 199-208, June 2006.

• P. Patarasu, A. Faraj, and X. Yuan, "Pipelined Broadcast on Ethernet Switched Clusters." The 20th IEEE International Parallel & Distributed Processing Symposium (IPDPS), Rhodes Island, Greece, April 25-29, 2006.

• A. Faraj, P. Patarasuk, and X. Yuan, " Bandwidth Efficient All-to-all Broadcast on Switched Clusters." The 2005 IEEE International Conference on Cluster Computing (Cluster 2005), Boston, MA, September 27-30, 2005.

• A. Faraj and X. Yuan, "Automatic Generation and Tuning of MPI Collective Communication Routines." The 19th ACM International Conference on Supercomputing (ICS), June 2005.

• A. Faraj and X. Yuan, "An Empirial Approach for Efficient All-to-All Personalized Communication on Ethernet Switched Clusters." The 2005 International Conference on Parallel Processing (ICPP), Oslo, Norway, June 14-17, 2005.

• A. Faraj and X. Yuan, "Message Scheduling for All-to-all Personalized Communication on Ethernet Switched Clusters." The 19th IEEE International Parallel & Distributed Processing Symposium (IPDPS), Denver, Colorado, April 4-8, 2005.

• A. Karwande, X. Yuan, and D. Lowenthal, "CCMPI: A Compiled Communication Capable MPI Prototype for Ethernet Switched Clusters." ACM SIGPLAN Symposium on Principles and Practice of Parallel Programming (PPoPP), San Diego, California, June 11-13, 2003.

• A. Faraj and X. Yuan "Communication Characteristics in the NAS Parallel Benchmarks," Fourteenth IASTED International Conference on Parallel and Distributed Computing and Systems (PDCS 2002), Cambridge, MA, November 4-6, 2002.

• X. Yuan, S. Daniels, A. Faraj and A. Karwande, "Group Management Schemes for Implementing MPI Collective Communication over IP-Multicast.", The 6th International Conference on Computer Science and Informatics, Durham, NC, March 8-14, 2002.

• R. G. Lane, D. Scott and X. Yuan, "An Empirical Study of Reliable Multicast Protocols over Ethernet-Connected Networks," International Conference on Parallel Processing (ICPP'01), pages 553-560, Valencia, Spain, September 3-7, 2001.

Thesis

• Ahmad Faraj, " Automatic Empirical Techniques for Developing Efficient MPI Collective Communication Routines," PHD Dissertation, Department of Computer Science, Florida State University, Fall 2006.

• Ahmad Faraj, "Communication Characteristics in the NAS Parallel Benchmarks," MS. Thesis, Department of Computer Science, Florida State University, Fall 2002.

• Amit Karwande, "CC-MPI: A Compiled Communication Capable MPI Prototype for Ethernet Switched Clusters," MS. Thesis, Department of Computer Science, Florida State University, Spring 2003.