Fostering International Academic Cooperation in Big Data and High Performance Computing

File 1 pdf 7.7MB

This presentation proposes to discuss emerging trends and lessons learned as part of cHiPSet "High-Performance Modelling and Simulation for Big Data Applications” http://chipset-cost.eu/, an on-going 4-year COST Action. COST is the “longest-running European framework supporting trans-national cooperation among researchers, engineers and scholars.” http://www.cost.eu/
Funded by the European Commission as part of the Horizon 2020 programme from 2015 to 2019, cHiPSet is fostering a novel endeavour to support information exchange, synergy, and coordination of activities among leading research groups. It nurtures cross-pollination between the high-performance computing (HPC) community and those scientific disciplines actively deploying Big Data applications.
On the one hand, complex systems do not naturally lend themselves to straightforward modular decompositions that support parallelisation and hence optimal HPC execution support. They often require a significant amount of computational resources with data sets scattered across multiple sources and different geographical locations. On the other hand, modelling and simulation are widely considered essential in science and engineering to substantiate the prediction and analysis of complex systems and natural phenomena. Consequently, data-intensive modelling and simulation make efficiency critical as they typically require coping with multi-dimensional and multi-level integration and model state explosion. Furthermore, their performance is nowadays arguably dominated by remote and local data movement overhead (network messages, memory, and storage accesses), which poses a challenge to different computing areas including data analytics, high-throughput infrastructures, cloud computing, and parallel programming.
cHiPSet systematically fosters interconnected research cooperation and mobility through the organisation of meetings, workshops, visits, and schools with the participation of academics and practitioners of leading industrial and higher-education institutions. The research groups involved in this Action represent a well-balanced set of competencies, ranging from HPC architectures through high-level parallel programming models, to a number of specific data-intensive domains. cHiPSet has been actively promoted as an effective platform for recommendation and support for the best possible use of publicly-funded HPC infrastructures, commercially-provided or community-run clouds, and volunteer computing systems in Europe, Asia, Australia, USA, and Latin America (Brazil).

Author(s):

Horacio González-Vélez    
NATIONAL COLLEGE OF IRELAND
Ireland