Introduction to sustainable ultrascale computing systems and applications

Excerpt

Ultrascale systems are envisioned as large-scale complex systems joining parallel and distributed computing systems that will be two to three orders of magnitude larger than today’s systems. Targeting sustainable solutions for ultrascale computing will need cross fertilization among HPC, large-scale distributed systems, and big data management, and the reformulation of many algorithms and applications used today in different areas of research. Such a reformulation has to address different challenges that arise from the different application areas, algorithms, and programs. Challenges include the scalability of the applications programs to use a large number of system resources efficiently, the usage of resilience methods to include mechanisms to enable application programs to react to system failures, and the inclusion of energy-awareness features into the application programs to be able to obtain an energy-efficient execution. As a large number of resources has to be controlled when using ultrascale systems, the availability of suitable programming models and environments also plays an important role for ultrascale applications. The programming models for ultrascale computing should provide enough abstractions, such that the application programmer does not need to deal with all low-level details of an efficient execution of the (parallel) programs and the control of the execution resources of the platform. On the other hand, the programming models should enable the application programmer to concentrate on the algorithmic aspects and problem-specific issues of the specific application area, such that program development is supported as far as possible. …