This project provides: Asynchronous task-based runtime initially targeting HPC applications
The value of the Open Community Runtime (OCR) project
The Open Community Runtime project is creating a runtime framework that explores new programming methods for machines with high core count. The initial focus is on HPC applications. Its goal is to create a framework and reference implementation to help developers explore programming methods to improve the power efficiency, programmability, and reliability of HPC applications while maintaining app performance.
OCR will help app developers with the complex process of writing multi-core apps by masking the effort to manage event-driven tasks, events (which embody dataflow and code flow dependencies), memory data blocks (with semantic annotations for runtime use), machine description facilities, and more.
Who it’s for
This project is for system developers, testers, debuggers, and other contributors working on high performance systems.
This is a large open source project distributed under the BSD open source license. OCR was originally unveiled at Supercomputing Conference 2012 (SC12) with a major new release (v0.8) introduced at Supercomputing 2013 (SC13). Community participation is encouraged, both for runtime enhancement as well as exploration of algorithm/application decomposition for new programming models. Version 1.0 was released in May 2014 and development has moved to its new home:
Please refer to that website for more up-to-date resources, including mailing lists and GIT repository.
About Intel's and other contributers' involvement
Intel is a major contributor to the Open Community Runtime project, enabling the generation-to-generation ‘it just works’ the ecosystem has come to expect. This project is optimized through open source for Intel® Atom™ processors, Intel® Core™ processors, Intel® Xeon® processors and Intel® Xeon® Phi™ coprocessors.
Partial support for the OCR v0.8 release was provided through the XStack program of the U.S. Department of Energy, Office of Science, Advanced Scientific Computing Research (ASCR). Components of OCR were also supported by the UHPC program of the U.S. Department of Defense's Advanced Research Projects Agency (DARPA), Intel Corporation/Intel Labs, and Rice University.