Challenges and Goals
Computational Fluid Dynamics (CFD) maturely used in engineering design and analysis across several industry sectors is ubiquitous; estimated annual value in the order of €1billion, used by 10,000s of organisations and 100,000s of users. The algorithmic technology-base is known, well established and universally applied. Continued widespread exploitations, in the next decade and beyond, are hampered by performance scaling bottlenecks on massively parallel HPC (High Performance Computing) systems, rendering this technology as not fully fit to take best advantage of recent major developments in High Performance Computing.
Euros of CFD Annual Value
Lines of code in OpenFOAM
Project — exaFOAM
Computational Fluid Dynamics performance on HPC is notably worse than idealised algorithms, due to inherent bandwidth needs, three-dimensional and time-date handling with non-sparse matrix dependencies, spatial domain decomposition requirements, I/O challenges. Independent software vendors (ISVs) have rightly concentrated their efforts on different customer demands such as complex physics modelling (turbulence, multi-phase, combustion/reactions, particulates, heat transfer). However, we see increasingly now that there is a commercial need to improve performance of industrial software and codes. There is an increasingly stated need to redress the balance between performance and functionality, perhaps even to re-learn the lessons of parallelism and vectorisation explored during the 1980s, due to the approaching of the hybrid pre-exascale era in HPC.
The project focus on transitioning CFD from its current near-petascale performance towards (pre)-exascale deployment by understanding and overcoming the current limitations in algorithms and production-delivery processes. Therein, they will consolidate the position of European organisations at the leading edge of exploitation of CFD for engineering needs in energy efficiency, safety, clean environment and comfort.