Technical Report 12 is a preview of OpenMP 6.0
Denver, Colorado – November 9, 2023 – The OpenMP® Architecture Review Board (ARB) has released Technical Report 12, the second preview of version 6.0 of the OpenMP API, which will be released in 2024.
New Features included in the Technical Report
The key features that are added in Technical Report 12 include:
- Improved tasking support: Support for free-agent threads, which are threads that are not assigned to any team that is executing a parallel region; this support extends the set of threads that may execute tasks;
- Improved device support: The coexecute directive divides the execution of array notation into separate units of work, enhancing device support for Fortran, while the mapping of data to devices has been extended to facilitate control of the memory used for that data;
- Improved support for C & C++: Full support for C23, including C attribute syntax, and C++23. New C/C++ attributes have been introduced;
- Extended loop transformations: Addition of loop transformations that ensure the use of loop fusion, reversal and interchange are applied; and
- Removal of Deprecated features: Features that were deprecated in versions 5.0, 5.1, and 5.2 have been removed.
Technical Report 12 is an important milestone on the path to OpenMP 6.0. Users and implementers are encouraged to provide feedback on the proposed API. Feedback can be sent to email@example.com.
OpenMP 6.0 planning
“This technical report provides a preview for the OpenMP community, our users, and the implementers to comment on the features planned for OpenMP 6.0,” says Michael Klemm, CEO of the OpenMP ARB. “Version 6.0 will be released in November 2024 and will be a major leap forward to ready the OpenMP programming model for exascale computing.”
“TR12 assures that OpenMP 6.0 will provide several significant advances for programming large-scale, heterogeneous systems,” said Bronis R. de Supinski, Chair of the OpenMP Language Committee. “It adds critical features that include free-agent threads, which simplify the OpenMP tasking model and improve its ability to provide high resource utilization, and the coexecute construct, which will make the use of GPUs and other devices from Fortran applications that use array syntax easier. Several other key features have already been added with more planned over the next year.”
The mission of the OpenMP Architecture Review Board (ARB) is to standardize high-level, directive-based,multi-language parallelism that is performant, productive and portable. Jointly defined by a group of major computer hardware and software vendors and users, the OpenMP API is a portable, scalable model that gives parallel programmers a simple and flexible interface for developing parallel applications for platforms ranging from embedded systems and accelerator devices to multicore systems and shared-memory systems. The OpenMP ARB owns the OpenMP brand, oversees the OpenMP specification and produces and approves new versions of the specification. Further information can be found at https://www.openmp.org.
Matthijs van Waveren