Energy-Aware Sustainable Computing on Future Technology – Paving the Road to Exascale Computing
Activity Code: ICT-2011.9.8
Project Reference: 318793
Coordinator:Karlsruher Institut für Technologie
Description:
The EXA2GREEN project aims at developing a radically new energy aware computing paradigm and programming methodology for exascale computing. The key aspect of the proposed approach is that the issue of energy sumption and the resulting trade-off with the performance and the accuracy of the overall simulation process will be taken into account in all simulation levels: from the kernel, numerical/combinatorial building blocks to the application level by means of the sidered mathematical models. The proposed approach of Energy-Aware Numerics goes beyond the standard hardware level or operating software stack usually sidered for energy issues and puts the application in the centre of the scene for all aspects related to energy efficiency.
The EXA2GREEN project takes up this multidisciplinary challenge by bringing together HPC experts, computer scientists, mathematicians, physicists and engineers. The project team is part of an emerging, multidisciplinary European research community and covers all essential fields of expertise, which allow opening absolutely new perspectives in the area of energy-aware numerics in the exascale era.
The overall goal of this project is to develop unventional ideas in order to cope with the issue of power sumption. Reducing the power requirement by a factor of at least 100 is the challenge which needs to be addressed in order to be able to use this technology in a meaningful way. This is one of the reason why making the transition to exascale computing requests radical transformation in the current perception of numerical simulation in high performance computing. The viability of the proposed approach will be investigated sidering a proof of cept where the energy footprint of a large and operational meteorological model for atmospheric and aerosol simulation (COSMO-ART) will be analysed.
Partners:
- Universitaet Hamburg, Deutschland
- Universitat Jaume I De Castellon, España
- Steinbeis Innovation Ggmbh, Deutschland
- Ibm Research Gmbh, Schweiz
- Eidgenoessische Technische Hochschule Zurich, Schweiz