Extreme scale computing and data driven technologies

to develop world-class extreme scale, and energy-efficient High Performance Computing (HPC) and data driven technologies, that use software engineering techniques, programming tools and libraries that can be adapted and retargeted to rapidly evolving HPC architectures, in view of maximising application performance and efficiency in next generation supercomputers. Actions should allow leveraging the efforts on the European low power processing technologies as well as the Centres of Excellence and build a sustainable exascale HPC ecosystem in Europe, enabling collaborations among the relevant stakeholders.

Scope

Proposals should address the development of extreme scale computing technologies and system architectures, programming models and tools, resource management, fault tolerance, and mathematical methods and algorithms in an increasingly complex and heterogeneous computing environment with memory and storage hierarchies. The approaches should respond to critical demands of application performance, energy efficiency, scale, resilience, scheduling, programmability, etc., across the levels of the compute stack, including compute elements, networking, data storage and data handling. A co-design approach should be followed, covering from the application to the hardware, answering user challenges of industry and research centres/academia and involving the work on the European low power processing technologies and results of Centres of Excellence, significant previous European and national projects in the area, and the data provided e.g. by the Digital Innovation Hubs.

The approaches should ensure that they contribute to the realisation of future exascale system architectures based on European technologies. In particular, proposals have to demonstrate that such approaches align with the efforts of European low power processing technologies, by describing the mechanisms that will be used for that purpose.

Proposals should address one or more of the following sub-topics (proposals should clearly indicate the sub-topic which is their main focus):

a) Technologies to increase sustained application performance at node and system level, improve energy efficiency and open new usage domains (e.g. High-Performance Data Analytics and mixed precision technologies for existing computing models as well as for Artificial Intelligence).

b) Technologies to manage data volumes generated and consumed, to minimize data movement and to increase flexibility to store, manipulate and access extremely large data sets in very heterogeneous computing environments.

c) Networking capabilities allowing low latency and high bandwidth communication between large numbers of extreme computing and data components (e.g. nodes, storage, memory, accelerator etc.).

d) Programming models, associated run-time systems, system software and compilers to ensure code scalability, maintainability, functionality and performance portability and optimisation across existing and possible future architectures and systems.

e) New mathematical methods and algorithms to ensure efficient usability and improve energy efficiency, featuring high robustness and enhanced scalability.

The JU considers that for sub topics a), b) and c) proposals requesting a contribution from the JU of up to EUR 4 million, respectively requesting a contribution from the JU of up to EUR 7.5 million for sub-topic d) and requesting a contribution from the JU of up to 1.5 million for sub topic e), and , matched by the Participating States with a similar amount, and a duration of 3 years would allow this specific challenge to be addressed appropriately[1]. Nevertheless this does not preclude submission and selection of proposals with another duration or requesting other amounts. At least one proposal per sub-topic will be selected for sub topics a), b), c) and d).

Successful proposals are expected to coordinate their approaches to ensure coherence of the expected outcomes with the EuroHPC exascale roadmap.

Proposals should describe how the proposed work will contribute to the listed corresponding expected impacts and include baseline, targets and metrics to measure impact.

• Contribution to the realisation of the EuroHPC overall and specific objectives[2]
• Strengthening scientific leadership as well as the competitiveness and innovation potential of European industry, contributing to a sustainable exascale HPC supply ecosystem in Europe and ensuring European technological autonomy in this field
• Leveraging the efforts on the European low power processing technologies (in particular the European Processor Initiative) and contributing to the realisation of future exascale system architectures based on such technologies
• Creation and promotion of European IP
• Maturity of solutions and potential for commercial exploitation in future European exascale HPC systems

[1]Please refer to the Annex of the Work Plan for the funding rate and the national contribution to this topic

[2]Council Regulation 2018/1488 of 28 September 2018 establishing the European High Performance Computing Joint Undertaking (EuroHPC) https://eurohpc-ju.europa.eu/documents/Regulation.pdf

News flashes