Scalability is a core property of algorithms and solutions and, at the same time, a crucial computer science and especially computational challenge. Because in more and more scientific domains the ability to “go big” computationally sets the pace of insight and innovation, there is an urgent need for efficient methods addressing large-scale problems and handling large-scale data with large-scale software systems on large-scale computer systems: “to out-compute is to out-compete”. Hence, extreme scalability involves various fields of informatics and implies innovative and, in many cases, disruptive approaches.
Future supercomputers will consist of billions of cores arranged in complex and heterogeneous hierarchies. Powerful networking technologies (inside and across systems), with billions of devices and sensors being interconnected to an Internet of Things, will be crucial. There is a need to develop programming models and tools to help application developers write efficient code. New future-proof, highly efficient (asynchronous) algorithms must be able to prevent unwanted data transfer and communication as well as cope with occasional hardware errors. Energy-efficiency will become a guiding principle for system, algorithm, and infrastructure design. Algorithms and applications have to be re-designed and optimized for massive parallelism. Computing centers will see a larger variety of operation models – batch, interactive, and urgent computing. Classical computing centers (with the main task of High-Performance Computing) and data centers (with the main task of High-Performance Analytics) will either converge or develop on separate tracks, extreme scaling in any case being a common challenge. The management, storage, analysis, fusion, and processing of huge amounts of data including visualization – research data in general, which may stem from simulations, experiments, sensors, social networks, businesses, etc. – will be the key to the upcoming data- driven science, economy, and society.
Thus, extreme scalability is a door-opener for Exascale Computing and Big Data, since it helps to keep problems tractable even if they increase in size.
Further aspects of Extreme Scaling are IT Security, Big Data, Computer Graphics and Image Processing.
SeisSol is a highly scalable software to simulate earthquake scenarios, in particular for accurate simulation of dynamic rupture processes. SeisSol was optimized for the currently largest supercomputers of the world. Automatic code generation substantially increases the performance per processor. Further algorithmic improvements led to faster runtimes (by a factor of 20) and allow simulations that are bigger than before by a factor 100 – executed on the SuperMUC supercomputer at LRZ.
The Priority Programme (SPP) SPPEXA is different from other SPP with respect to its genesis, its volume, its funding via DFG's Strategy Fund, with respect to the range of disciplines involved, and to a clear strategic orientation towards a set of time-critical objectives. Therefore, despite its distributed structure, SPPEXA also resembles a Collaborative Research Centre to a large extent. Its successful implementation and evolution will require both more and more intense structural measures. The Coordination Project comprises all intended SPPEXA-wide activities, including steering and coordination, internal and international collaboration and networking, and educational activities.
The Transregional Collaborative Research Center InvasIC (TCRC 89) investigates dynamic resource management for invasive applications from highly parallel chip multiprocessors up to state-of-the-art supercomputers. The goal is to provide optimized execution and resource usage while maintaining a high level of predictability. In High Performance Computing this research will lead to the productive development of evolving applications based on MPI and OpenMP as well as to a system-level resource management beyond the current static space sharing approach.
The goal of the European Horizon 2020 project READEX is the dynamic tuning of the energy consumed by HPC applications. The project will extend the Periscope Tuning Framework (periscope.in.tum.de) developed at TUM according to the scenario-based tuning methodology from the embedded systems area. Application dynamism will be exploited by dynamically switching between precomputed system configurations.