EuroHPC Partnership

EuroHPC Partnership
Name	EuroHPC Partnership
Formation	2018
Type	Public-public partnership
Region served	European Union, European Economic Area
Headquarters	Brussels

EuroHPC Partnership is a collaborative European initiative established to develop a pan-European supercomputing ecosystem integrating funding, procurement, and operation of high-performance computing (HPC) resources across European Union member states and associated countries. Launched through a political agreement among institutions including the European Commission, the Partnership aligns with strategic programs such as Horizon Europe, the Digital Europe Programme, and national initiatives to address computational demands in science, engineering, and industry. The initiative coordinates procurement, deployment, and co-funding of pre-exascale and exascale systems while fostering collaborations with research infrastructures like PRACE and policy frameworks such as the European Cloud Initiative.

Overview

The Partnership builds a distributed supercomputing network connecting national centers, research infrastructures, and industry stakeholders across Belgium, France, Germany, Italy, Spain, Netherlands, Poland, Portugal, Greece, Sweden, Finland, Denmark, Austria, Czech Republic, Slovakia, Romania, Hungary, Lithuania, Latvia, Estonia, Bulgaria, Croatia, Slovenia, Luxembourg, Ireland, Malta, and associated countries. It brings together funding and procurement mechanisms modelled after partnerships involving organizations such as CERN, EMBL, ESA, EIB, EBRD, and European Investment Bank projects, while aligning with standards and roadmaps from bodies like IEEE, ETSI, ISO, and OpenMP. Operational coordination involves national HPC centres, university consortia, and technology vendors that previously collaborated on projects with Intel Corporation, NVIDIA Corporation, AMD, Arm Holdings, IBM, Cray Inc., Hewlett Packard Enterprise, Atos SE, Dassault Systèmes, and Siemens AG.

Objectives and Scope

Primary objectives include procuring pre-exascale and exascale systems, expanding access to public research and industry, and fostering an HPC ecosystem supporting applications in domains such as climate modelling, bioinformatics, aeronautics, and energy. The Partnership’s scope covers the deployment of supercomputers, development of machine learning and data analytics stacks, and support for software toolchains used in collaborations with institutions like Max Planck Society, CNRS, CERN, Fraunhofer Society, Wellcome Trust, European Molecular Biology Laboratory, European Space Agency, Joint Research Centre, and national academies. It aims to accelerate innovations linked to multinational programmes such as the European Green Deal and initiatives by organizations including EIT Digital, Shift2Rail, Clean Sky, and the European Defence Agency for dual-use technologies.

Governance and Funding

Governance is structured around a legal framework involving participating states, the European Commission, and national agencies, with oversight committees modelled after governance seen in European Investment Fund projects and intergovernmental research organisations. Funding combines contributions from national budgets, Digital Europe Programme allocations, and co-funding mechanisms reminiscent of Horizon 2020 and Horizon Europe grant frameworks, with procurement managed through competitive tenders akin to those used by European Space Agency procurements and multinational contracts with firms such as Atos, HPE, IBM, and NVIDIA. Decision-making engages representatives from national ministries, research councils like UK Research and Innovation (where applicable), funding agencies such as ANR and DFG, and regional development banks.

Infrastructure and Systems

The Partnership coordinates the acquisition and operation of tiers of HPC systems including pre-exascale and exascale-class platforms, interconnected via high-bandwidth networks operated by research and education networks such as GÉANT, JANET, SURFnet, RedIRIS, RENATER, DFN, and NorduNet. Installed systems integrate accelerators, processors, and storage technologies from suppliers like NVIDIA, AMD, Intel, Fujitsu, and leverage software ecosystems including Linux, Kubernetes, Slurm Workload Manager, OpenStack, Singularity (software), and libraries such as MPI, OpenCL, and CUDA. Centres hosting systems include national supercomputing centres, university-managed data centres, and collaborations with pan-European facilities such as PRACE and the European Open Science Cloud.

Research, Development and Industrial Impact

Research and development activities supported by the Partnership enable projects across computational chemistry, genomics, materials science, artificial intelligence, and weather forecasting, engaging laboratories and companies such as Max Planck Institute for Plasma Physics, Institut Pasteur, University of Oxford, ETH Zurich, Technical University of Munich, Siemens Healthineers, BASF, Airbus, Schneider Electric, and startups incubated by EIT Digital Accelerator. Industrial uptake targets small and medium enterprises through competence centres and programmes comparable to CATRENE and EUREKA, while academic collaborations foster reproducibility and open science practices endorsed by organisations like FAIRsharing and the Research Data Alliance.

Membership and Participating Entities

Participants include EU member states, associated countries, national research councils, supercomputing centres, universities, and private partners. Key institutional participants and stakeholders map to entities such as the European Commission, Council of the European Union, European Parliament, national ministries for research and digital affairs, regional authorities, major academic institutions (e.g., University of Cambridge, Sorbonne University, Politecnico di Milano), and technology providers including Hewlett Packard Enterprise, Atos SE, IBM, NVIDIA, AMD, Fujitsu, and cloud providers like Amazon Web Services, Microsoft Azure, and Google Cloud Platform that collaborate on hybrid HPC-cloud solutions.

Challenges and Future Directions

Challenges comprise sustaining multi-year funding, securing supply chains influenced by export controls and trade policies, skills shortages in high-performance computing, and integration with emerging paradigms such as quantum computing and neuromorphic processors developed by organisations like IBM Research, Google Quantum AI, D-Wave Systems, and Rigetti Computing. Future directions emphasize exascale deployments, interoperability with the European Open Science Cloud, enhanced support for artificial intelligence workflows, and strengthening partnerships with international initiatives such as the US National Science Foundation programmes, PRACE, and bilateral research agreements with countries like Japan and Canada. Ongoing policy alignment will involve dialogues with institutions including the European Court of Auditors and stakeholders across science and industry to ensure traceability, impact assessment, and long-term sustainability.

Category:Supercomputing