Enabling Technologies ...
In the early days of HPC, the challenge was primarily hardware- and algorithm-related. Since then, additional topics have appeared, some with a tradition in computational sciences (visualisation, e.g.), some with rather fresh (data exploration, e.g.) or basically even without relations so far (systematic software engineering, e.g.). It is the first main idea underlying MAC to bring together all these enabling technologies for the crucial step from just simulating to a modern integral concept of Advanced Computing. TUM and its partners in MAC are very well prepared to start this endeavour, since the MAC consortium has all the necessary expertise and infrastructure at its disposal – in an almost unique way.
... Plus Applications
Although the methodological aspects of Advanced Computing already represent demanding fields of research by themselves, we learned from CSE’s many success stories, but also disappointments, that fundamental technologies and challenging applications must be considered together to exploit the full potential – which is MAC’s second main idea. Neither a generic development of algorithms nor a merely application-specific approach can meet the requirements for CSE. And we need more than one application – to detect similarities, to avoid the wheel to be reinvented, to profit from synergies, and to allow for the broadest scope and utmost benefit of the various methodological ideas. MAC will, therefore, tackle challenging applications from both science and engineering and also adopt the doctoral education philosophy envisaged in TUM’s International Graduate School of Science and Engineering (IGSSE), successful in the first round of the German Excellence Initiative.
In CSE, everyone talks about interdisciplinarity, but only few really follow through.
Silo-like academic systems and attitudes persisting in established structures are the main roadblocks for an environment Advanced Computing needs. To implant the trans-disciplinary approach as a mainstreaming philosophy is MAC’s third main idea.