Periodic Reporting for period 2 - CEEC (Center of Excellence for Exascale CFD)
Periodo di rendicontazione: 2024-01-01 al 2025-06-30
For many centuries, scientific discovery relied on performing experiments and the subsequent deduction of new theoretical models. The advent of powerful computers, coupled with new and ever more efficient numerical algorithms, makes it possible to simulate complex systems with increasing realism, and to automatize even model discovery using artificial intelligence (AI) technologies. Computational Fluid Dynamics (CFD) is one of the most prominent areas that clearly requires, and even motivate exascale computing to be part of the engineering and academic workflows. Given the physical scaling and the availability of highly efficient simulation codes, CFD has the potential of reaching exascale performance, as one of the few application areas. This center will implement exascale ready workflows for addressing relevant challenges for future exascale systems, including those procured by EuroHPC. The significant improvement in energy efficiency will be facilitated through efficient exploitation of accelerated hardware architectures (GPUs) and novel adaptive mixed-precision calculations. Emphasis is furthermore given to new or improved algorithms that are needed to exploit upcoming exascale architectures. The efforts of the center are driven by a collection of five different lighthouse cases of physical and engineering interest, ranging from aeronautical to atmospheric flows, with the goal of reaching TRL 4 and even 5 for selected cases. All development is done in five European HPC codes which span the entire spectrum of CFD applications, including compressible, incompressible and multiphase flows.
In the second reporting period, CEEC has achieved the following main results:
- Lighthouse cases and their needs have been further refined, and previously defined benchmark cases have been used to steer the development and track performance and scalability of associated codes and complete toolchains for each specific case. Some lighthouse cases have started the first simulations of configurations close to the final setup on EuroHPC systems.
- performance assessments and optimisations targeting various accelerators, analysing the readiness of CEEC's codes regarding code generation techniques and continuous integration, testing and performance tracking.
- Development of mixed-precision methods for the consortium codes has progressed using mini-apps. Efficient multigrid-based solvers and preconditioners have been developed, and work on fault-resilient algorithms and h-type mesh refinement algorithms for accelerators has continued.
- Developing the techniques and technologies required for supporting the lighthouse cases. Different workflow and machine-learning strategies and frameworks for multi-fidelity uncertainty quantification have been developed. Methods for visualisation, data management, and dynamic resource management have been developed.
- An initial exploratory analysis was carried out to assess the suitability of each code on the EPI Software Development Vehicles, and in collaboration with the original developers, simplified versions of the codes were prepared to run on EPI RISC-V prototypes. Work has continued on algorithms suitable for quantum computers towards solving simplified Navier-Stokes equations, like Burgers'.
- Lighthouse cases and their needs have been further refined, and previously defined benchmark cases have been used to steer the development and track performance and scalability of associated codes and complete toolchains for each specific case. Some lighthouse cases have started the first simulations of configurations close to the final setup on EuroHPC systems.
- performance assessments and optimisations targeting various accelerators, analysing the readiness of CEEC's codes regarding code generation techniques and continuous integration, testing and performance tracking.
- Development of mixed-precision methods for the consortium codes has progressed using mini-apps. Efficient multigrid-based solvers and preconditioners have been developed, and work on fault-resilient algorithms and h-type mesh refinement algorithms for accelerators has continued.
- Developing the techniques and technologies required for supporting the lighthouse cases. Different workflow and machine-learning strategies and frameworks for multi-fidelity uncertainty quantification have been developed. Methods for visualisation, data management, and dynamic resource management have been developed.
- An initial exploratory analysis was carried out to assess the suitability of each code on the EPI Software Development Vehicles, and in collaboration with the original developers, simplified versions of the codes were prepared to run on EPI RISC-V prototypes. Work has continued on algorithms suitable for quantum computers towards solving simplified Navier-Stokes equations, like Burgers'.
CEEC aim at being at the forefront of exascale computational fluid dynamics with consortium codes supporting all existing major pre- and exascale systems. Algorithmic activities ensured portability, extreme-scale scalability, and sustained performance of the consortium codes, which were successfully demonstrated with scale-out pre-exascale machine runs. CEEC and its codes are more and more recognized not only in Europe but also in the US and Japan. The six demanding exascale lighthouse cases will further drive the global impact generation.