Research Infrastructure
New exascale-class supercomputer Alice Recoque is on its way
Key facts
What: New exascale-class supercomputer Alice Recoque.
Name: The supercomputer is named after a pioneering French computer scientist and one of the first engineers in artificial intelligence.
Made possible by: The Jules Verne project, carried out by the Jules Verne consortium.
Where: The supercomputer will be housed at TGCC (Très Grand Centre de Calcul) located in Bruyères-le-Châtel, near Paris. This location is part of CEA’'s supercomputing centre.
Solution: It will address the challenges posed by the explosion of data generated by scientific instruments.
Aim: The installation and deployment of Alice Recoque will start in 2026. The system’s capacity will be tested at the end of 2027. We expect that the first researchers will be able to start using the system by the end of 2027.
The Jules Verne project is carried out by the Jules Verne consortium. This brings together France, represented by the Grand Équipement National de Calcul Intensif (GENCI) as the hosting entity, in collaboration with the Commissariat à l'énergie atomique et aux énergies alternatives (CEA) as the hosting site, the Netherlands, represented by SURF, the Dutch national supercomputing centre and GRNET, the Greek national infrastructures for research and technology.
The supercomputer will be housed at TGCC (Très Grand Centre de Calcul) located in Bruyères-le-Châtel, near Paris. This location is part of CEA’s supercomputing centre. The system’s capacity will be tested at the end of 2027.
Objectives of the consortium
The main objective of this consortium is to deploy a world-class Exascale supercomputer, partly based on European hardware and software technologies. It will make it possible to respond to major societal and scientific challenges. Via the convergence at scale of numerical simulations, massive data analysis and artificial intelligence.
Investing in our future
Through SURF, the Netherlands is one of the investors in this supercomputer. This will allow Dutch researchers to make direct use of some of the computing time on Alice Recoque in the future. This is in addition to the computing time that can already be requested via SURF on the national supercomputer Snellius and the European supercomputer LUMI.
In addition to the supercomputer itself, the Jules Verne consortium, in collaboration with other EuroHPC consortia, will provide user support and training to European researchers in optimising their applications on the supercomputer.
Ready for the future data explosion
The supercomputer will act as a sovereign gas pedal in the finer modelling of the effects of climate change, in the development of new materials, energies and decarbonized mobility solutions, in the creation of digital twins of the human body enabling personalized medicine, or in training the next generation of generative AI or multimodal models.
It will also address the challenges posed by the explosion of data generated by scientific instruments (such as telescopes, satellites, sequencers, microscopes, sensor networks), IoT/Internet devices or large-scale multi-digital simulations. This avalanche of data makes the use of these supercomputers crucial for science, industry and decision-makers, to process this data in competitive timescales and in the most energy-efficient way possible.
And we call her Alice Recoque
The supercomputer is named Alice Recoque, a pioneering French computer scientist and one of the first engineers in artificial intelligence. She worked on the designs of mini computers in the 1970s and led research focused on artificial intelligence.
Funding
This new European Exascale system is acquired by the EuroHPC Joint Undertaking (EuroHPC JU) and co-funded by France, the Netherlands and Greece, represented respectively by GENCI, together with CEA, by SURF and by GRNET. The project represents a total investment of € 554 million over five years of operations.
Technical specifications
Jules Verne will provide European, French and Dutch researchers with an unprecedented computing capacity of over 1 Exaflop/s – a billion billion ("1" followed by 18 zeros) floating point operations per second, equivalent to over 5 million modern laptops, and over 300 PB of storage at startup.
Alice Recoque will be powered by next-gen AMD EPYC CPUs, codenamed “Venice”, AMD Instinct™ MI430X GPUs engineered for sovereign AI and scientific applications, as well as a dedicated scalar partition based on SiPEARL Rhea2 processors, all being interconnected with next-gen Eviden BXI v3 European high-performance network system. This supercomputer will deliver a computing capacity exceeding one Exaflop/s (one billion billion operations per second) in double precision floating point for HPC and AI processing power. Featuring 432 GB of HBM4 and 19.6 TB/s of memory bandwidth [1], AMD Instinct MI430X GPUs will enable Alice Recoque to be among the most powerful and energy-efficient scientific instruments ever built and made available in Europe.
About EuroHPC JU
EuroHPC JU is a legal and funding entity that brings together the European Union and participating countries to coordinate efforts and pool resources. With the objective of making Europe a world leader in supercomputing.
To equip Europe with a cutting-edge supercomputing infrastructure, EuroHPC JU has already procured 11 supercomputers, distributed across Europe including JUPITER in Germany, Europe’s first exascale system.
An exascale supercomputer in 2027
The installation and deployment of Alice Recoque will start in 2026. The system’'s capacity will be tested at the end of 2027. It is expected that the first researchers will be able to start using the system by the end of 2027.
More information
Do you have a question about Alice Recoque? Get in touch with Servicedesk Data & Computing Services at servicedesk@surf.nl