Exascale Technology Centre: Parallelism for extreme scale systems
The largest supercomputers today can perform over a Petaflop of calculations per second on real scientific applications. However Exascale Systems are planned for 2016-2018 which will perform an Exaflop.
An Exaflop is a million million million calculations per second, a thousand times faster than a Petaflop. Extreme scale systems are needed for global challenges such as climate change, medical research and sustainable energy sources. However, these systems require a huge increase in parallelism with many innovations in supercomputer hardware, software and applications which represents an immense challenge to those wishing to exploit them.
The International Exascale Software Project (IESP) leads the way in researching this challenge. EPCC is part of this initiative and has established the Exascale Technology Centre. The centre is exploring new ideas and technologies to meet the challenge of delivering an Exaflop within the next decade in particular areas such as massive scalability, new programming models, fault tolerance and accelerated node design. To scale to the extreme, programming models will need to provide excellent data locality management, lightweight asynchronous communications and improved synchronisation. The centre is therefore looking at hybrid, single-sided and PGAS models. Exascale architectures will meanwhile contain large numbers of multi-socket, multi-core nodes. Accelerators that increase node performance will require new efficient programming models.
EPCC and Cray
This is a joint project with Cray's European Exascale Research Initiative.