Exploitation of Exascale Systems for Open-Source Computational Fluid Dynamics by Mainstream Industry
Keyvisual image main

exaFOAM is working to reduce bottlenecks in performance scaling for computational fluid dynamics (CFD) applications on massively parallel high-performance computing (HPC) systems.

It will identify the principal bottlenecks in mainstream CFD tools, and identify and implement improvements in algorithms in order to improve both their scalability on HPC systems and workflows for codes used in industrial software, with emphases on preprocessing, postprocessing, and I/O.

The project is also developing a comprehensive set of tools for measuring performance and scalability on different HPC platforms, demonstrating methods for improving performance and scalability, and identifying, implementing, and promoting best practices in coding, execution, performance, and scalability. The results should enhance Europe's leadership in industrial sectors in which the target code is applied, including air and space travel, power generation, the automotive industry, and disaster prevention, among others.

These activities will converge in an open source CFD code and open source tools, ensuring that the technological progress achieved during this project is easily accessible for European industry. This approach will also ensure that the resulting tools are verifiable during implementation and validation using a pre-established, well characterized, and easily understandable framework.


01. April 2021 -
31. March 2024


H2020-JTI-EuroHPC-2019-1 (EuroHPC-03-2019)

Project partners

  • ESI Group (FR, coordinator)
  • Cineca Consorzio Interuniversitario (IT)
  • E4 Computer Engineering SPA (IT)
  • Politecnico di Milano (IT)
  • Sveuciliste u Zagrebu, Fakultet Strojarstva I Brodogradnje (HR)
  • Upstream CFD GmbH (DE)
  • Technische Universität Darmstadt (DE)
  • Wikki Gesellschaft für numerische Kuntinuumsmechanik mbH (DE)
  • Universität Stuttgart (DE)
  • Barcelona Supercomputing Center (ES)
  • National Technical University of Athens (EL)
  • Universidade do Minho (PT)


Andreas Ruopp

Deputy Head, Department of Numerical Methods & Libraries

+49-711 685-87259 andreas.ruopp(at)hlrs.de