IWOMP 2026 Opens Call for Papers Ahead of October Workshop in Vienna

The 22nd International Workshop on OpenMP (IWOMP) has officially opened its call for papers ahead of the October 7–9, 2026 event in Vienna, Austria. The workshop will be hosted at Austrian Scientific Computing (ASC) at TU Wien, continuing its tradition as the premier forum for parallel programming research with the OpenMP API.

This year's theme centers on "OpenMP: Adaptability for Heterogeneous Multi-Device Systems." The focus reflects the growing complexity of modern computing architectures where developers must balance ease of use against raw performance across an expanding range of hardware configurations.

According to the official IWOMP 2026 call for papers, the OpenMP API has evolved through versions 5.0, 5.1, 5.2, and 6.0 to become the leading programming model for on-node heterogeneous parallelism. The standard now supports C/C++ and Fortran base programming languages while targeting increasingly diverse computing platforms.

Reputable tech media outlet HPCwire reported the announcement on April 30, 2026, confirming the timeline and scope of the workshop. The coverage notes that IWOMP serves as the annual gathering dedicated to promoting and advancing all aspects of parallel programming with OpenMP.

The hardware landscape has shifted dramatically from simple cores to advanced SIMD units, deeper memory hierarchies, and heterogeneous computing environments. OpenMP has extended its application interface to harness these capabilities throughout the spectrum of hardware advances. Developers now face the reality of multicore processors, OpenMP devices (accelerators such as GPGPUs, DSPs, or FPGAs), and Multiprocessor Systems on a Chip (MPSoCs).

Recent developments in the OpenMP API itself—metadirectives and variants for selecting device- and architecture-specific directives—present new opportunities and challenges for software and hardware developers. The growing prevalence of multi-device nodes and tightly integrated heterogeneous components further underscores the need for adaptable and scalable programming abstractions.

Submission guidelines require papers to be limited to 12 pages (not counting references). Authors of accepted papers will prepare a final version of up to 15 pages including references. The submission deadline is Friday, May 29, 2026 (AoE). Acceptance notifications arrive Friday, July 10, 2026 (AoE). Camera ready copies are due Friday, July 31, 2026 (AoE).

Submissions go through easychair.org/conferences/?conf=iwomp2026. Papers must follow LNCS Guidelines found at the Springer website. As in previous years, IWOMP 2026 will publish formal proceedings of accepted papers in Springer's Lecture Notes in Computer Science (LNCS) series.

The topics list covers accelerated computing and offloading to devices, implementation of OpenMP for new devices and architectures, applications in any domain that rely on OpenMP, data mining and analysis or text processing with OpenMP, machine learning and OpenMP, memory model, memory policies and management, performance analysis and modeling, performance portability, proposed OpenMP extensions, runtime environment, scientific and numerical computations, tasking, tools, and vectorization.

Recent advances in the C, C++, and Fortran base languages offer interesting opportunities and challenges to the OpenMP programming model. More complex applications of OpenMP tasks, usage of heterogeneous computing platforms, and performance portability across different architectures are topics of particular interest to the organizing committee.

Technical University of Vienna (TU Wien) serves as the host institution. Founded in 1815, it is Austria's largest institution for technology and natural sciences. The university offers programs in engineering, computer science, and natural sciences across eight faculties with a focus on research and innovation.

The workshop aims to bring together researchers, practitioners, and tool developers to discuss innovations, experiences, and future directions for OpenMP. This happens in an era where flexibility, performance, and portability must coexist across an ever-expanding range of computing architectures (which is easier said than done when your code needs to run on everything from embedded systems to exascale supercomputers).

Looking at the physical reality of this work, developers spend hours wrestling with compiler flags, debugging race conditions that appear only under specific load patterns, and measuring performance differences that might be fractions of a percent but matter at scale. The tactile experience involves staring at terminal output, clicking through profiler visualizations, and waiting for builds that take longer than a coffee break.

IWOMP 2025 took place in Charlotte, NC, USA. IWOMP 2024 was held in Perth, AU. IWOMP 2023 occurred in Bristol, UK. IWOMP 2022 happened in Chattanooga, TN, USA. IWOMP 2021 was held in Bristol, UK (online). The workshop has maintained its annual cadence despite global disruptions.

The OpenMP API has been instrumental in advancing parallel programming, enabling portability across both traditional and emerging heterogeneous computing systems. As the standard continues to introduce new capabilities, the OpenMP API continues to evolve, offering solutions to the growing complexity of high-performance computing (HPC) environments.

The workshop welcomes papers on how the OpenMP API balances ease of use and performance, particularly in the context of heterogeneous systems, exascale computing, and real-time workloads. The emphasis remains on adaptability across heterogeneous and multi-device ecosystems.

Whether the community can deliver practical solutions that work across the fragmented hardware landscape remains the real question. Theoretical advances mean little if developers cannot implement them without spending weeks on architecture-specific tuning. The bar for acceptance will likely reflect this pragmatic reality.

Artūras Malašauskas is an AI Systems Integrator with 20+ years of production-grade web engineering experience. He has designed, shipped, and scaled enterprise Python/PHP systems for logistics, SaaS, and public-sector clients. For the past year, he has focused exclusively on AI integrations: deploying open-source LLMs, building generative media pipelines (image, audio, video), and engineering multi-agent workflows for real production environments. His standard: reproducibility, security, cost-efficient inference—no vaporware. He documents and evaluates emerging AI tooling, separating verified capabilities from marketing noise. Technical editor at: muza-ai.eu, ai-verslas.lt, ai-naujinos.lt Connect on LinkedIn