The International Workshop on OpenMP (IWOMP) is the annual workshop dedicated to the promotion and advancement of all aspects of parallel programming with OpenMP. This pioneering workshop has been attracting an international audience of leading academic and industrial experts since 2005 and is the premier forum to present and discuss issues, trends, recent research ideas, and results related to parallel programming with OpenMP.

We solicit quality submissions of unpublished technical papers that detail innovative, original research and development related to OpenMP.

IWOMP 2022 will be hosted by the University of Tennessee, Chattanooga and will be colocated with EuroMPI 2022.

Background

As computing hardware has evolved from simple core reproduction to advanced SIMD units, deeper memories, and heterogeneous computing, OpenMP has also evolved and extended its application interface to harness new capabilities throughout the spectrum of hardware advances. The 5.0, 5.1 and 5.2 versions of the OpenMP specification have established OpenMP as the leading API for on-node heterogeneous parallelism that supports all versions of the C/C++ and Fortran base programming languages.

Advances in technologies, such as multicore processors and OpenMP devices (accelerators such as GPGPUs, DSPs or FPGAs), Multiprocessor Systems on a Chip (MPSoCs), and recent developments in OpenMP itself (e.g., metadirectives and variants for selecting device- and architecture-specific directives) present new opportunities and challenges for software and hardware developers. Recent advances in the C, C++ and Fortran base languages also offer interesting opportunities and challenges to the OpenMP programming model.

Theme for 2022:

OpenMP in a Modern World: From Multi-device Support to Meta Programming

As most HPC systems are now comprised not only of multicore CPUs but also feature multiple accelerators, such as GPGPUs and FPGAs, programming model support for handling multiple devices and multiple levels of parallelism within a node is growing in importance. Further, the growing diversity in node designs makes programming mechanisms that enable seamless optimization for multiple node types are also essential. This year’s theme highlights OpenMP extensions, implementations and applications that facilitate using such systems and papers that detail them are particularly welcome.

Topics

All topics related to parallelism in OpenMP are of interest, including:

    Accelerated computing and offloading to devices
    Applications (in any domain) that rely on OpenMP
    Data mining and analysis or text processing and 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
    Vectorization