08:50 - 09:00 - Welcome to PMBS Workshop
09:00 - 10:00 - Technical Papers
-09:00 - 09:30 - Navigating An Evolutionary Fast Path to Exascale
Richard Barrett, Sandia National Laboratories,
Simon Hammond, Sandia National Laboratories,
Courtenay Vaughan, Sandia National Laboratories,
Douglas Doerfler, Sandia National Laboratories,
Justin Luitjens, NVIDIA Corporation and
Duncan Roweth, Cray Inc
-09:30 - 10:00 - Modeling a Million-Node Dragonfly Network using Massively Parallel Discrete Event Simulation
Misbah Mubarak, Rensselaer Polytechnic Institute,
Christopher Carothers, Rensselaer Polytechnic Institute,
Robert Ross, Mathematics and Computer Science Division, Argonne National Laboratory and
Philip Carns, Mathematics and Computer Science Division, Argonne National Laboratory
10:00 - 10:30 - Coffee
10:30 - 12:30 - Technical Papers
-10:30 - 11:00 - Performance Modeling of Algebraic Multigrid on BlueGene/Q: Lessons Learned
Hormozd Gahvari, University of Illinois at Urbana Champaign,
William Gropp, University of Illinois at Urbana Champaign,
Kirk Jordan, IBM Research,
Martin Schultz, Lawrence Livemore National Laboratory and
Ulrike Yang, Lawrence Livermore National Laboratory
-11:00 - 11:30 - Developing Performance-Portable Molecular Dynamics Kernels in OpenCL
John Pennycook, University of Warwick and
Stephen Jarvis, University of Warwick
-11:30 - 12:00 - Performance Tuning of Matrix Multiplication in OpenCL on Different GPUs and CPUs
Kazuya Matsumoto, University of Aizu,
Naohito Nakasato University of Aizu and
Stanislav G. Sedukhin, University of Aizu
-12:00 - 12:30 - Performance Modeling for Dense Linear Algebra
Elmar Peise, AICES RWTH Aachen and
Paolo Bientinesi, AICES, RWTH Aachen
12:30 - 13:30 - Lunch
13:30 - 15:00 - Technical Papers
-13:30 - 14:00 - Unprecedented Scalability and Performance of the new NNSA Tri-Lab Capacity Cluster (TLCC2)
Mahesh Rajan, Sandia National Laboratories,
Doug Doerfler, Sandia National Laboratories,
Paul Lin, Sandia National Laboratories,
Simon Hammond, Sandia National Laboratories,
Richard Barrett, Sandia National Laboratories and
Courtenay Vaughan, Sandia National Laboratories
-14:00 - 14:30 - Towards Performance Predictive Application-dependent Workload Characterization Waleed Alkohlani, New Mexico State University and
Jeanine Cook, New Mexico State University
-14:30 - 15:00 - Towards the Generation of Hard Disk Models Through Physical Geometry Discovery
Steven Wright, University of Warwick,
Simon Pennycook, University of Warwick and
Stephen Jarvis, University of Warwick
15:00 - 15:30 - Coffee
15:30 - 16:30 - Technical Papers
-15:30 - 16:00 - Improving the Accuracy and Efficiency of Time-Independent Trace Replay
Federic Desprez, INRIA
George S. Markomanolis INRIA/ENS-Lyon and
Frederic Suter, CC IN2P3/CNRS
-16:00 - 16:30 - Trace Driven Data Structure Transformations
Tomsilav Janjusic, University of North Texas
Christos Kartsaklis, Oak Ridge National Laboratory and
Krishna M. Kavi, University of North Texas
16:30 - 17:45 - Late Breaking Research Papers
-16:30 - 16:55 - Accelerating Hydrocodes with OpenACC, OpenCL and CUDA
Wayne Gaudin, AWE PLC,
Andy Herdman, AWE PLC,
Andrew Mallinson, University of Warwick,
David Beckingsale, University of Warwick,
Michael Boulton, University of Bristol,
Simon McIntosh-Smith, University of Bristol and
Stephen Jarvis, University of Warwick
-16:55 - 17:20 - Designing Configurable, Modifiable and Reusable Components for Simulation of Multicore Systems
Jun Wang, Georgia Tech,
Jesse Beu, Goergia Tech,
Sudhakar Yalamanchili, Georgia Tech and
Tom Conte, Georgia Tech
-17:20 - 17:45 - An Analytical Study of Loop Tiling for a Large-scale Unstructured Mesh Application
Mike B. Giles, University of Oxford,
Gihan R. Mudalige, University of Oxford,
Carlo Bertolli, Imperial College London and
Paul H.J. Kelly, Imperial College London
17:45 - Workshop Close
3rd International Workshop on
Performance Modeling, Benchmarking and Simulation of High Performance Computer Systems (PMBS12)
held as part of SC12, Salt Lake City, Utah, USA, November 2012
in cooperation with ACM SIGMETRICS