Platform-driven e-infrastructure innovation

Deadline: Mar 29, 2017  
CALL EXPIRED

Topic Description

Prepare the capacity required to future generations of e-infrastructure is the key challenge. e-Infrastructure platforms and services need to evolve through innovation actions to respond to the long-term needs of research and education communities (e.g. in case of large RIs entering in functions in a 5 to 10 years' timeframe). Platforms and services are first designed, prototyped and piloted with "supply and demand-side" approaches triggered by to the most demanding cases. The innovative developments bringing state-of the-art technology need to evolve and mature to be integrated and offered as dependable e-infrastructures.

Proposals will address parts (a) or (b), but not both:

(a) Support to Public Procurement of innovative HPC systems, PPI (proposals should address all points below):

(1) procurement of innovative HPC solutions supporting the deployment in Europe of world-leading HPC capability infrastructure

(2) ensuring and reinforcing European access to European leading-edge supercomputing Tier-0 infrastructures and services, by making available a substantial percentage of the new systems to European researchers in the frame of the Pan-European High Performance Computing infrastructure and services (see EINFRA-11-2016)

(3) diversify the available leading-class HPC capabilities through a rich set of HPC architectures featuring the most advanced technology made available by R&I (Research and Innovation) in Europe, in order to satisfy the needs of a wider range of users in very different key application areas

(4) contribute to the coordination of plans and procurements for the provision of leading-class HPC capabilities at European and national level in view of the implementation the European supercomputing strategy , encompassing funding and technical specifications

(b) Research and Innovation Actions for e-Infrastructure prototypes:

Proposals will address only one of the points below. At least one proposal for each point will be selected:

1. Universal discoverability of data objects and provenance (proposals should address all points below):

Prototyping an e-infrastructure service, based on standards and best-practices, for the uptake of a Digital Identifier e-infrastructure for digital objects (articles, datasets, collections, software, nomenclature, etc.), researchers and contributors, which cuts across geographical, temporal, disciplinary, cultural, organisational and technological boundaries, without relying on a single centralised system but rather federating locally operated systems to ensure interoperability. The requirements of all relevant stakeholder groups (researchers, libraries, data centres, publishers, etc.) should be addressed as well as global interoperability through agreed mechanisms (e.g. in consensus building through the Research Data Alliance).

The Commission considers that proposals requesting a contribution from the EU between EUR 4 and 5 million would allow this specific challenge to be addressed appropriately. Nonetheless, this does not preclude submission and selection of proposals requesting other amounts. It is expected that one proposal will be selected.

2. Computing e-infrastructure with extreme large datasets (proposals should address all points below):

Develop service prototypes to cope with very large data resources. It should include the basis software layers supporting applications such as modelling, simulation, pattern recognition, visualisation, etc. The developments should be supported by robust mathematical methods and tools. Prototypes should follow an open source approach and aim at common interfaces to access and analyse underlying data collected/stored in different platforms, formats, locations and e-infrastructures and be tested against requirements of very large or highly heterogeneous research data sets. Clean slate approaches to high-performance computing and data management (e.g. HPC-through-the cloud, support of most innovative server’s architectures for distributed computing in particular high Memory/Cores ratios allowing “in memory” processing) targeting 2020+ 'data factory' requirements of research communities and large scale facilities (e.g. ESFRI projects) are encouraged.

The Commission considers that proposals requesting a contribution from the EU between EUR 2.5 and 3 million would allow this specific challenge to be addressed appropriately. Nonetheless, this does not preclude submission and selection of proposals requesting other amounts.

(a) Support to Public Procurement of innovative HPC systems, PPI: This action will contribution to the European HPC strategy through the creation of a European procurement market for the benefit of the HPC actors in Europe (in particular technology suppliers) and catalysing the efforts to vitalise the European HPC ecosystem. It position Europe as a world-class HPC hub with more leading-class HPC computing resources and services available at European level for European academia and industry, independently of the location of users or HPC systems. It will foster adoption and use of innovative world-class HPC solutions featuring the most advanced results of the R&I in Europe, widening the access to more users, in particular for and industry (including SMEs). It will improve effectiveness of public procurement in leading-class HPC systems through joint procurement and pooling of European and national resources, contributing to sustainability. Benefits will also translate in better coordination between demand and supply in the European HPC ecosystem, with improved collaboration of the users and procurers with technology suppliers.

(b) Research and Innovation Action for e-Infrastructure prototypes

Universal discoverability of data objects and provenance: the successful set up of such service based on standards and best-practices will support interoperability of e-infrastructure services. The use of Digital Identifiers opens new prospects for advanced services for science and education and for encouraging openness and building trust. Data and other resources become discoverable and easy to use which will facilitate access to resources and collaboration between scientists. It has the potential to be used as a core service across Europe and globally. Duplication of efforts for developing services common to many e-infrastructures is reduced.

Computing e-infrastructure with extreme large datasets: The successful prototyping of this action services will support the evolution of e-infrastructure services based on exascale data resources. It will prepare data and computing infrastructure to absorb needs of communities that push the envelope in terms of data and computing intensive science while softening the learning curve for scientific communities that will be using new services.