Here's the latest news from the OpenMP Architecture Review Board...

Webinar: Mastering OpenMP Performance

Join us on March 18 for a webinar about how to tune your OpenMP code to get the best performance. This presentation by Ruud van der Pas will dispel myths and include guidelines on how to get the best performance using the OpenMP API.

DATE: Thursday, March 18, 2021
TIME: Noon Pacific (1900 UTC)

Register for this free webinar now!

This webinar is part of a series. Watch for announcements of new webinars coming soon! Next up: The Leaders of OpenMP Discuss its Future

Announcing the IWOMP 2021 CFP!

IWOMP is the annual workshop for the promotion and advancement of parallel programming with OpenMP. They are looking for quality submissions of unpublished technical papers on original R&D related to OpenMP.

IWOMP 2021 will be hosted by the High Performance Computing research group at the University of Bristol, UK.

Submission deadline is May 14, 2021

Submit your paper for IWOMP 2021

New Book: High Performance Parallel Runtimes

A new book, High-Performance Parallel Runtimes, has been released by Michael Klemm and Jim Cownie.

Focus: The theoretical and practical aspects of parallel programming systems.
Audience: Students of compiler construction, parallel programming, or programming systems. Also software developers interested in the core algorithms used to implement a parallel runtime system.

The source code for a small, incomplete, LLVM compatible, OpenMP runtime (LOMP) and all of the benchmarks used in the book are available on GitHub

Learn more about the Book
Get LOMP on Github

Webinar on debugging tools

Check out the POP HPC webinar: “Debugging Tools for Correctness Analysis of MPI and OpenMP Applications” presented by Joachim Protze of RWTH Aachen IT Center.

Friday 26th March 2021 15:00hrs-15:40hrs CET.

Register for this free webinar now!

Monthly OpenMP telecons from ECP

The Exascale Computing Project (ECP) is holding monthly teleconferences for OpenMP users.
These monthly calls allow users to share their OpenMP tips and experiences with the community and bring issues or concerns to the attention of the compiler developers and OpenMP language committee members.

The next telecon is on March 26!

Learn more and participate in the next call!

Intro to OpenMP videos from EuroFusion!

The European Consortium for the Development of Fusion Energy (Eurofusion) has been organizing webinars on a range of topics, including how to use GPUs with OpenMP.
Two videos on OpenMP are now available.

Check out the videos

LLVM: OpenMP offloading to AMD Radeon GPU

The mainline LLVM compiler’s AMDGPU back-end now supports OpenMP offload to AMD Radeon GPUs.

Read the full story

Nvidia SDK 21.1 with OpenMP support for target directives

The latest update to the NVIDIA HPC Software Development Kit (SDK) is available for download. It now has beta support for a subset of OpenMP target offload on Volta or newer GPUs.

Read the full story

Improved OpenMP support in MS Visual Studio

Starting with Visual Studio 2019 version 16.9, Microsoft has begun to add experimental support for new OpenMP features in a more systematic way. As a first step, they have added the option to generate code compatible with LLVM’s OpenMP runtime library (libomp) on the x64 architecture. This is the runtime they will use in the future to enable OpenMP in their compilers.

Learn more