Abstract
Abstract. In the simulation of complex multi-scale flows arising in weather and climate modelling, one of the biggest challenges is to satisfy strict service requirements in terms of time to solution and to satisfy budgetary constraints in terms of energy to solution, without compromising the accuracy and stability of the application. These simulations require algorithms that minimise the energy footprint along with the time required to produce a solution, maintain the physically required level of accuracy, are numerically stable, and are resilient in case of hardware failure. The European Centre for Medium-Range Weather Forecasts (ECMWF) led the ESCAPE (Energy-efficient Scalable Algorithms for Weather Prediction at Exascale) project, funded by Horizon 2020 (H2020) under the FET-HPC (Future and Emerging Technologies in High Performance Computing) initiative. The goal of ESCAPE was to develop a sustainable strategy to evolve weather and climate prediction models to next-generation computing technologies. The project partners incorporate the expertise of leading European regional forecasting consortia, university research, experienced high-performance computing centres, and hardware vendors. This paper presents an overview of the ESCAPE strategy: (i) identify domain-specific key algorithmic motifs in weather prediction and climate models (which we term Weather & Climate Dwarfs), (ii) categorise them in terms of computational and communication patterns while (iii) adapting them to different hardware architectures with alternative programming models, (iv) analyse the challenges in optimising, and (v) find alternative algorithms for the same scheme. The participating weather prediction models are the following: IFS (Integrated Forecasting System); ALARO, a combination of AROME (Application de la Recherche à l'Opérationnel à Meso-Echelle) and ALADIN (Aire Limitée Adaptation Dynamique Développement International); and COSMO–EULAG, a combination of COSMO (Consortium for Small-scale Modeling) and EULAG (Eulerian and semi-Lagrangian fluid solver). For many of the weather and climate dwarfs ESCAPE provides prototype implementations on different hardware architectures (mainly Intel Skylake CPUs, NVIDIA GPUs, Intel Xeon Phi, Optalysys optical processor) with different programming models. The spectral transform dwarf represents a detailed example of the co-design cycle of an ESCAPE dwarf. The dwarf concept has proven to be extremely useful for the rapid prototyping of alternative algorithms and their interaction with hardware; e.g. the use of a domain-specific language (DSL). Manual adaptations have led to substantial accelerations of key algorithms in numerical weather prediction (NWP) but are not a general recipe for the performance portability of complex NWP models. Existing DSLs are found to require further evolution but are promising tools for achieving the latter. Measurements of energy and time to solution suggest that a future focus needs to be on exploiting the simultaneous use of all available resources in hybrid CPU–GPU arrangements.
Highlights
Numerical weather and climate prediction capabilities represent substantial socio-economic value in multiple sectors of human society, namely for the mitigation of the impact of extremes in food production, renewable energy, and water management, infrastructure planning, and for finance and insurance whereby weather-sensitive goods and services are traded
This paper presents an overview of the ESCAPE strategy: (i) identify domain-specific key algorithmic motifs in weather prediction and climate models, (ii) categorise them in terms of computational and communication patterns while (iii) adapting them to different hardware architectures with alternative programming models, (iv) analyse the challenges in optimising, and (v) find alternative algorithms for the same scheme
Simplified simulations of the messagepassing interface (MPI) communications performed in ESCAPE indicate that the strong scalability of the communication time for the spectral transform transpositions is better than for the halo communication required by semi-Lagrangian advection halos and the global norm computation commonly used in semi-implicit methods (Zheng and Marguinaud, 2018). These results indicate that halo communication will become almost as costly as the transpositions in the spectral transform method if a very large number of MPI processes is involved
Summary
Numerical weather and climate prediction capabilities represent substantial socio-economic value in multiple sectors of human society, namely for the mitigation of the impact of extremes in food production, renewable energy, and water management, infrastructure planning, and for finance and insurance whereby weather-sensitive goods and services are traded. If the envisioned increase in model fidelity is constrained by only marginally growing power envelopes and decelerating general-purpose processor speed, performance issues need to be addressed at the root, and a more radical redesign of the basic algorithms and their implementations used for weather prediction needs to be considered This is why ESCAPE investigates both HPC adaptation and alternative numerical formulations of physically and computationally clearly identifiable model components – introduced as Weather & Climate Dwarfs – followed by a detailed analysis of their respective computational bottlenecks and subsequent hardware- and algorithm-dependent optimisations.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
