OpenMP: Balancing Productivity and Performance Portability

This book constitutes the refereed proceedings of the 21st International Workshop on OpenMP, IWOMP 2025, in Charlotte, NC, USA, during October 1 October 3, 2025

The 14 full papers included in this book were carefully reviewed and selected from 19 submissions. They were organized in topical sections as follows:New Directions: Communication and I/O; New Directions: Handling New Constraints and Devices; New Performance and Correctness Tooling ; Advanced Capability Evaluation and New Directions: Programming Advances.

Inhalt

.- New Directions: Communication and I/O.
.- Discussion of Device-Device Collective Communication in OpenMP Target Offloading.
.- OMPCCL: Extending OpenMP with Portable Collective Operations for Multiple Devices.
.- OpenMP Meets I/O: Portable and Runtime-Managed File Tasks.
.- New Directions: Handling New Constraints and Devices.
.- Parallelizing Irregular DOACROSS Loops Using ChatGPT and Transactional Memory in OpenMP.
.- OpenMP-RT: Pragma Support for Scheduling Periodic Real-Time Tasks.
.- OpenMP-Q: Quantum Task Offloading in OpenMP.
.- New Performance and Correctness Tooling.
.- Profile Generation for GPU Targets.
.- Data Race Satisfiability on Array Elements.
.- Predicting Performance for OpenMP GPU Parameter Choices.
.- Advanced Capability Evaluation.
.- Evaluating LLVM OpenMP Offload Optimizations on NVIDIA GH200 Grace Hopper Superchip and AMD Instinct™ MI300A Accelerator Architectures.
.- Evaluating OpenMP on Aurora's Intel GPU Max Series 1550.
.- Demonstrating OpenMP Offload Performance with the STREAmS-2 Application and the AMD Next-Gen Fortran Compiler.
.- New Directions: Programming Advances.
.- ChatPORT: Fine-tuned LLM for Easy Code {PORT}ing.
.- Programming GPUs with OpenMP and Python.