Menu

GPU-accelerated aerodynamics simulations with CODA

Within EXCELLERAT, the CFD code CODA, developed by DLR, ONERA and Airbus, was successfully optimised for modern GPU-based High-Performance Computing (HPC) systems. By significantly improving the scalability of the Sparse Linear Systems Solver (Spliss), CODA can now efficiently scale across hundreds of GPUs, achieving speedups of up to 8.6× compared to state-of-the-art CPU systems. This enables larger, higher-fidelity aerodynamics simulations in weeks instead of months and prepares CODA for current and future heterogeneous and exascale HPC systems, supporting the design of quieter, safer and more fuel-efficient aircraft.

Explaining Scientific Code, Assisted by Large Language Models

Understanding an unfamiliar codebase is a recurring challenge in scientific computing. In a blog article, CERFACS introduces WalkingPrompt: a workflow for code explanation assisted by language models that systematically traverses a repository, links dependencies, and generates structured summaries across modules and directories. By keeping the human developer in control, WalkingPrompt reduces repetitive effort, produces more consistent explanations, and creates a coherent first layer of documentation for large scientific codes, while addressing key concerns such as trust, privacy, and reproducibility.

Enabling GPUs for Energy Efficient Aerodynamics Simulations of an Aircraft

In this success story, GPU acceleration is used to enable energy-efficient, high-fidelity aerodynamics simulations of complex aircraft geometries. By optimizing the Alya and Sod2D CFD codes with OpenACC and NVIDIA HPC compilers, the Barcelona Supercomputing Center achieved excellent scalability on MareNostrum V, delivering up to 20× faster simulations and 10–12× lower energy consumption compared to CPU-based approaches. This work supports more accurate and sustainable aircraft design and demonstrates the broader potential of GPUs for large-scale scientific simulations.

EXCELLERAT at EuroHPC Summit 2025: Engaging in the Future of High-Performance Computing

The EuroHPC Summit 2025, held in Krakow from March 18 to March 20, brought together key stakeholders in the European high-performance computing (HPC) ecosystem. EXCELLERAT was represented at the conference by several team members participating in various sessions. Additionally, EXCELLERAT had a dedicated poster on display throughout the event, highlighting the project’s latest achievements and ongoing efforts in advancing computational engineering.

Automated workflow for the Hi-fidelity simulation of propulsion devices: application to after burners

The decarbonation and depollution efforts of Safran Group are pushing the innovation of new burner technologies forward. However, the volume of after-burner flames and the unpredictability of newer fuels such as hydrogen flames are challenging the meshing practices born of more classical, no longer computationally affordable configurations.
Learn more about how this challenge could be solved in our success story.

Advanced scalable workflow of ray tracing kernel for radiative heat loads assessment

In the engineering design and digital twin of the first wall of a tokamak fusion reactor, one of the challenges is to address the physical phenomena with sufficient complexity to arrive at a digital solution that is comparable to an actual experiment. This complexity comes at the cost of time and memory heavy computation. It is thus important to optimise the CPU/GPU solver for the codes being used and to demonstrate their capability to run them at exa-scale level. EXCELLERAT provides access to HPC infrastructure and expertise to achieve this goal.