Execution Models for Energy-Efficient Computing Systems - EXCESS
Computing technology is currently at the beginning of the disruptive transition from petascale to exascale (2010 – 2020), posing a great challenge on energy efficiency. Future computing systems will be constrained by power/energy and will feature massive parallelism, heterogeneity, and heavy exploitation of data locality. Existing approaches to current system design tend to focus on individual layers of the system stack, leaving performance and energy potentials unused that could be leveraged by coordinated system-wide optimisation.
The EXCESS project aims at providing radically new energy execution models forming foundations for energy-efficient computing paradigms that will enable significant improvements in energy efficiency for computing systems. A holistic approach that involves both hardware and software aspects together has the best chances to successfully address the energy efficiency problem and discover innovative solutions. EXCESS proposed models will bridge embedded processor models with general purpose ones.
EXCESS will take a holistic approach and will introduce novel-programming methodologies to drastically simplify the development of energy-aware applications that will be energy-portable in a wide range of computing systems while preserving relevant aspects of performance. EXCESS aims at bridging the gaps between embedded computing and HPC by novel models that will allow for holistic optimisation across all the layers of the computer system. The models “bridge” is supported by four “pillars” that are described in the picture below.