Advancing OpenMP for Future Accelerators 20th International Workshop on OpenMP, IWOMP 2024, Perth, WA, Australia, September 23–25, 2024, Proceedings

This book constitutes the refereed proceedings of the 20th International Workshop on OpenMP: Advancing OpenMP for Future Accelerators, IWOMP 2024, in Perth, WA, Australia, during September 23–25, 2024.

The 14 full papers presented in this book  were carefully reviewed and selected from 16 submissions. They are grouped into the following topics: current and future openMP optimization; targeting more devices; best practices; tools; and simplifying parallelization.

.- Current and Future OpenMP Optimization. 

.- Towards Locality-Aware Host-to-Device Offloading in OpenMP.

.- Performance Porting the ExaStar Multi-physics App Thornado On Heterogeneous Systems - A Fortran-OpenMP Code-base Evaluation.

.- Event-Based OpenMP Tasks for Time-Sensitive GPU-Accelerated Systems.

.- Targeting More Devices. 

.- Integrating Multi-FPGA Acceleration to OpenMP Distributed Computing.

.- Towards a Scalable and Efficient PGAS-based Distributed OpenMP.

.- Multilayer Multipurpose Caches for OpenMP Target Regions on FPGAs.

.- Best Practices. 

.- Survey of OpenMP Practice in General Open Source Software.

.- CI/CD Efforts for Validation, Verification and Benchmarking OpenMP Implementations.

.- Evaluation of Directive-based Programming Models for Stencil Computation on Current GPGPU Architectures.

.- Tools. 

.- Finding Equivalent OpenMP Fortran and C/C++ Code Snippets Using Large Language Models.

.- Visualizing Correctness Issues in OpenMP Programs.

.- Developing an Interactive OpenMP Programming Book with Large Language Models.

.- Simplifying Parallelization. 

.- Automatic Parallelization and OpenMP Offloading of Fortran Array Notation.

.- Detrimental Task Execution Patterns in Mainstream OpenMP Runtimes.