Workshop Program

Opening Remarks (8:45am - 9:00am)

Pavan Balaji, Argonne National Laboratory [slides]

Session 1 (9:00am - 10:00am): Invited Keynote

Session Chair: Pavan Balaji, Argonne National Laboratory

Speaker: Dr. Barbara Chapman, University of Houston

Title: Exascale: Why It Is Different [slides]

Abstract: Shortly after the first petascale system was installed (or maybe before), researchers began to discuss the likely characteristics of exascale systems and to develop plans to create them. The US Department of Energy has already made a commitment to build and deploy a small number of platforms of this scale. In order to do so, there will need to be a concerted effort to develop and implement the requisite software support for application developers, from the programming model to the execution environment. In this presentation, we discuss the challenges of exascale, and where we need to consider new solutions.

Bio of the Keynote Speaker: Dr. Chapman has been engaged in research on parallel programming languages and their implementation for more than 15 years. She is currently a Professor of Computer Science at the University of Houston, where she designs parallel programming interfaces, explores enhancements to compiler and runtime technology, and creates tool support for parallel application development. Her interests cover a wide range of areas from high performance computing through support for embedded systems design, and consider the challenges posed by a diverse set of architectures.
In 2001, Dr. Chapman founded cOMPunity, a not-for-profit organization that enables research participation in the development and maintenance of the OpenMP industry standard for parallel programming. Since that time, she has been actively involved in the evolution of the OpenMP API. Today, she is also engaged in the creation of OpenSHMEM, a new standard for parallel programming based upon single-sided communication, as well as in implementation techniques for Partitioned Global Address Space (PGAS) languages and system support for high-level programming models on heterogeneous systems. Her research group has developed OpenUH, a freely-available Open64-based compiler that demonstrates many of her research ideas.

Session 2 (10:30am - 12:20pm): Programming Models and Runtime Systems

Session Chair: Martin Schulz, Lawrence Livermore National Laboratory

  • "Recomposing An Irregular Algorithm Using a Novel Low-Level PGAS Model", Megan Vance and Peter Kogge [slides]
  • "A Middleware for Concurrent Programming in MPI Applications", Tobias Berka, Helge Hagenauer and Marian Vajtersic [slides]
  • "Kangaroo: Reliable execution of scientific applications with DAG programming model", Kai Zhang, Kang Chen, Wei Xue
  • "JETS: Language and System Support for Many Parallel Task Computing", Justin Wozniak and Michael Wilde [slides]

Session 3 (1:30pm - 3:20pm): Scheduling and Workflows

Session Chair: Vinod Tipparaju, Advanced Micro Devices (AMD)

  • "Energy-Constrained Dynamic Resource Allocation in a Heterogeneous Computing Environment", B. Dalton Young, Jonathan Apodaca, Luis Diego Briceno, Jay Smith, Sudeep Pasricha, Anthony A. Maciejewski, Howard Jay Siegel, Bhavesh Khemka, Shirish Bahirat, Adrian Ramirez, and Yong Zou [slides]
  • "Job Co-Scheduling on Coupled High-End Computing Systems", Wei Tang, Narayan Desai, Venkatram Vishwanath, Daniel Buettner, Zhiling Lan [slides]
  • "Integrating Scientific Workflows and Large Tiled Display Walls: Bridging the Visualization Divide", Hoang Nguyen, David Abramson, Blair Bethwaite, Minh Ngoc Dinh, Colin Enticott, Stephen Firth, Slavisa Garic, Ian Harper, Martin Lackmann, A.B.M. Russel, Stefan Schek, Mary Vail [slides]
  • "Restricted Admission Control in View-Oriented Transactional Memory", K. Leung and Z. Huang [slides]

Session 4 (3:40pm - 5pm): Communication and I/O

Session Chair: Brice Goglin, INRIA

  • "CellPilot: A Seamless Communication Solution for Hybrid Cell Clusters", Natalie Girard, William Gardner, John Carter and Gary Grewal [slides]
  • "Interval Based I/O: A New Approach to Providing High Performance Parallel I/O", Jeremy Logan and Phillip Dickens [slides]
  • "Improving Performance of the Irregular Data Intensive Application with Small Computation Workload for CMPs", Zhimin Gu