Imagine trying to predict exactly how air flows around a new plane wing or how a storm moves, but the math is too big for normal computers. This work creates a way to use the world's most powerful supercomputers to run these simulations much faster. It's like upgrading from a handheld calculator to a super-brain that can handle massive amounts of data without wasting energy.
Traditional fluid dynamics simulations are too slow and energy-intensive for complex real-world engineering. This prevents companies from iterating designs quickly and accurately.
Exascale-ready workflows and optimized algorithms for 5 European HPC codes. This includes mixed-precision calculations and GPU-accelerated solvers for fluid dynamics.
If you are an aircraft manufacturer dealing with slow wind-tunnel testing and simulation bottlenecks — this project developed exascale-ready CFD workflows that allow for more realistic simulations of aeronautical flows. This reduces the time needed to validate aerodynamic designs.
If you are a forecasting service dealing with imprecise atmospheric models — this project developed algorithms for atmospheric flows that leverage GPU acceleration. This enables higher resolution weather predictions using EuroHPC systems.
If you are a turbine designer dealing with complex multiphase flows in energy production — this project developed five European HPC codes that span the spectrum of fluid applications. This allows for more energy-efficient simulation of complex fluid dynamics.
Based on available project data, no pricing or cost structures are provided as the project is funded by the EU.
Yes, the project specifically targets exascale performance and is designed for use on EuroHPC systems to handle massive engineering workflows.
Based on available project data, specific licensing terms are not mentioned, though development involves five different European HPC codes.
The project is implementing exascale-ready workflows, but the specific integration timeline for external companies is not provided.
The project period runs from 2023-01-01 to 2026-12-31.
The consortium is heavily academic, consisting of 8 partners from 5 countries, with 6 universities and 2 research organizations. There is a 0% industry ratio, meaning the technology is currently being driven by scientific excellence rather than immediate commercial product requirements, which may indicate a need for industrial bridge-building to reach TRL 6+.
Contact the KUNGLIGA TEKNISKA HOEGSKOLAN (KTH) research office in Sweden.
Contact us to find a bridge between these exascale algorithms and your industrial CFD needs.
