Salishan 2011

Theme: Addressing Some Major Challenges for the Exascale Programming Model through Co-design among Hardware, Middleware, and Applications

Keynote Address: A Graphics Perspective on Co-Design?, Pat Hanrahan, Stanford University

Session 1: Bridging Disparate Architectures

• Are We Exploring the Right Enabling Technologies to Support End Applications in
  the Push to Exascale?, Robert Webster, Los Alamos National Laboratory
• Supporting Existing Programming Models while Allowing Exploitation of New
  Technologies, Robert Wisniewski, IBM Thomas J. Watson Research Center
• From Here to Exascale: Challenges and Potential Solutions, Bill Dally, Stanford University
• *Programming at Exascale, Michael Wolfe, The Portland Group

Session 2: Leveraging Hierarchy

• One Two Tree…Infinity: Facts and Speculations on Exascale Software, Pete Beckman, Argonne National Laboratory
• Explicit SW/HW Hierarchies for Efficient, Resilient, and Programmable Systems, Mattan Erez, University of Texas-Austin
• Opportunities for Chemistry on the Path to Exascale, Robert Harrison, Joint Institute of Computational Sciences, UT/ORNL
• Leveraging Hierarchy: Is This Our Undiscovered Country?, John Daly, National Security Agency (DOD)

Working Dinner/Speaker

• Lessons Learned: Relevant to Exascale--or not?, Paul Messina, Argonne National Laboratory

Session 3: Beyond the Software Pipeline

• Challenges Beyond the Software Pipeline, Bert Still, Lawrence Livermore National Laboratory
• Building the Next Generation of Parallel Applications: Co-design Opportunities and
  Challenges, Michael Heroux, Sandia National Laboratories
• Data Processing in Exascale-Class Computing Systems, Chuck Moore, AMD Corporate Fellow and CTO of Technology Group
• Performance Principles and Practice, Rich Vuduc, Georgia Tech University

Session 4: Exploiting Asynchrony

• Data-oriented Programming Models with Asynchronous Tasking: Yielding Some
  Control in a Community of Control Freaks, Brad Chamberlin, Cray, Inc.
• *System Co-design: The Potential for Greatness or Exceptional Mediocrity; the
  Choice is Ours, Stephen Poole, Oak Ridge National Laboratory
• *Unintelligent Design for Asynchronous Exascale Systems, Timothy Mattson, Intel Corporation
• Exploiting Asynchrony for Exascale Computational Materials Science, Timothy Germann, Los Alamos National Laboratory

Session 5: Co-design and Exposing Hidden Execution Costs

• Co-design and Exposing Hidden Execution Costs, Steve Wallach, Convey Computer
• A Case for More Intuitive Performance Analysis, Martin Schulz, Lawrence Livermore National Laboratory
• No Free Lunch, No Hidden Cost, Sharon Hu, University of Notre Dame
• Holistic Co-design, Arun Rodrigues, Sandia National Laboratories

*will not be posted