Workshop on Software Development Environments for High-Performance Computing
DEHPC '15
Co-located with the 24th International Conference on Parallel Architectures and Compilation Techniques • October 18, 2015 • San Francisco, CA

Call for Papers

High-Performance Computing platforms are increasingly challenging to program. The hardware is becoming more complex because of a large number of concurrent threads per node, heterogeneous cores, deep memory hierarchies, increased variability in computation and communication time and the need to reduce energy consumption and handle frequent faults. The software is becoming more complex as it combines codes developed by independent teams using different programming languages, and combines simulation with data analysis. The need for top performance and the need for portability are conflicting forces that are hard to reconcile.

The HPC community also suffers from a lack of good development environments that can be used across platforms, and handle well the unique problems of coding for HPC. This lack reduces the productivity of HPC programmers.

The workshop will focus on research that aims at solving these problems and advance the state of the art in HPC Software Development Environments.

Topic of interest include, but are not limited to:

  • Integrated development environments for parallel computing and HPC
  • Refactorings for performance and portability of parallel code
  • Case studies on the adoption and use of IDEs for parallel computing and HPC
  • Software development environments for GPUs and accelerators
  • Software development environment support for emerging parallel programming models and environments
  • Software development environment support for scientific domain specific languages
  • Program analysis and transformation for HPC
  • Support for the development of community codes
  • Software development environments for scientific computing
  • Performance and debugging tools for parallel computing and HPC
  • Software development environments for data analytics problems
  • Support for virtualization, and the management of virtual system images
  • Support for cloud development

The proceedings will be published digitally, together with the conference proceedings.

Submissions

All submissions must be made electronically through EasyChair. Submissions must include the full list of authors, their contact information, and their affiliations. Full paper submissions must be in PDF, formatted for US letter-size paper. They must not exceed 10 pages (all inclusive) in standard ACM two-column conference format (preprint mode, with page numbers). Templates for ACM format are available for Microsoft Word, and LaTeX on the ACM SIGPLAN Web site (use the 9 pt template). Over-length submissions will not be accepted. Submissions are not blind. Submissions will be judged on correctness, relevance, originality, significance, and clarity. For additional information regarding paper submissions, authors should contact the Program Chair.

Important Dates

Paper Submissions Due: August 25, 2015, Anywhere on Earth (extended)
Author Notification: August 31, 2015
Workshop: October 18, 2015

Agenda

Sunday, October 18, 2015

Submitted Papers 9:00–9:30 Tony Saad and James C. Sutherland. Wasatch: Addressing Multiphysics and Hardware Complexity in a High-Performance Computing Environment (Paper, Slides)
9:30–10:00 Jeffrey Overbey and Mitchell Price. Toward a Cloud IDE for HPC (Paper, Slides)
10:00–10:30 Alessandro Capotondi and Andrea Marongiu. On the Effectiveness of OpenMP Teams for Programming Embedded Manycore Accelerators (Paper, Slides)
  10:30–11:00 Break
News from the Trenches 11:00–11:20 Greg Watson: The View from IBM
11:20–11:40 David Lombard: The View from Intel (Slides)
11:40–12:00 Dale Southard: The View from NVIDIA (Slides)
12:00–12:20 Jay Alameda: The View from NCSA (Slides)
  12:20–2:00 Lunch (on your own)
Round Table 2:00–4:00 Marc Snir, Jeff Overbey et al. The Exascale Computing Initiative:
  1. What technology is needed for an exascale software development environment?
  2. What research program is needed to create this technology?
  3. What organization is needed to succeed?

Committee

Organizing Committee

  • Jay Alameda, National Center for Supercomputing Applications
  • Paul Hovland, Argonne National Laboratory
  • Lois Curfman McInnes, Argonne National Laboratory
  • Jeff Overbey, Auburn University
  • Marc Snir, Argonne National Laboratory
  • Greg Watson, IBM