There has been of late a renewed interest in technologies that improve the productivity of HPC software development, such as parallel programming languages and libraries, Object Oriented scientific programming, and parallel run-time systems and tools. The success of these activities requires a good understanding of common patterns used in the development of HPC software: patterns used for the coding of parallel algorithms, their mapping to various architectures, and performance tuning activities. The purpose of the workshop is to bring together people that are interested in HPC patterns: people interested in documenting patterns for the development of HPC software, people that are involved in the development of HPC software, and people that are involved in the development of languages, libraries, programming environments and tools that support the development of HPC software. The workshop will enable researchers to discuss their work and to initiate a dialogue on ways for leveraging the methods and insights of the design pattern community to advance the state of HPC software. Depending on the success of this first event, we expect it to develop into a regular workshop.
The workshop focuses on the documentation of patterns for the development of High Performance Computing software and on the use of parallel design pattern for the production of high-performance computing software. Topics of interest include but are not restricted to:
Laxmikant Kale, Ralph Johnson and Marc Snir, University of Illinois at Urbana-ChampaignProgram Overview
Please consider joining us for an in-depth look at recent advances in this area!
There is a link to the workshop Wiki site here.
For any other questions/comments about the workshop, or about this website, please send us an e-mail message