H2020: FET Extreme scale computing technologies, methods and algorithms for key applications and support to the HPC ecosystem

H2020: FET Extreme scale computing technologies, methods and algorithms for key applications and support to the HPC ecosystem

To develop world-class extreme scale, power-efficient and highly resilient High Performance Computing and data technologies, and to provide support for a sustainable exascale HPC ecosystem in Europe, enabling collaborations among the relevant stakeholders.

Proposals should clearly identify and address at least one of the following areas:

a. System software and management, addressing adaptive and dynamic scheduling; heterogeneity of system components; efficient data access, transfers and communication, novel execution models for emerging HPC and High Performance Data Analytics (HPDA) usages, etc.

b. Programming environments, reducing programming complexity and increasing scalability through advancements throughout the programming model and system software stack, and addressing code maintainability and functional portability across existing and future architectures and systems. Interoperability throughout the programming environment should be addressed.

c. I/O and storage environment for data-centric extreme scale computing addressing overall system performance predictability, feature-rich and flexible data access and storage system API’s, backup and retrieval of extreme volumes of data and systems operation in virtualised operating environment.

d. Data-intensive supercomputing and emerging HPC use modes addressing efficient implementation of established Big Data software frameworks and workloads on extreme-scale HPC systems, including the integration of Big Data and HPC programming models; algorithmic research addressing Machine Learning on HPC systems; interactive use of HPC resources for real time data analysis.

e. Mathematical methods and algorithms for extreme scalability of computing and data with impact in system energy reduction and resilience, and addressing the usability and the efficient implementation on different HPC architectures. Work should link to HPC and extreme scale data architectures and technologies as well as to relevant applications (e.g. challenges identified by the European Centres of Excellence on HPC).

The Commission considers that proposals for Research and Innovation Actions requesting a contribution from the EU of between EUR 5 and 10 million and a duration of 3 years would allow this area to be addressed appropriately. Nonetheless, this does not preclude submission and selection of proposals requesting other amounts and duration.

Established Researcher
EU
Consortium
Elina Humala, Senior Grant Advisor
Sisältö

Tämä kansio on tyhjä.