The European Council has unveiled plans to amend the existing regulations governing the European Consortium for High Performance Computing (EuroHPC), which oversees supercomputers for artificial intelligence (AI) development. The update, announced on May 24 by the European Union, includes goals for establishing and operating local artificial intelligence factories. These commitments aim to bolster infrastructure for AI supercomputing services within the EU.
The proposed amendment, initiated by the committee, builds upon the foundation laid by EuroHPC, an entity established in 2018 to spearhead European supercomputing initiatives. Presently, EuroHPC manages nine supercomputers across Europe, developed since the program's inception. The introduction of new artificial intelligence factories, as outlined in the amendment, will further this mission. The increased supercomputing capacity will provide EU AI startups with enhanced resources to innovate and refine their models.
Willy Borsus, Vice-President of the Wallonia Region and Minister of Economy, Research and Innovation, emphasized the collaborative effort required to reach this milestone. He underscored the strength of unity within Europe, highlighting achievements in the development of world-class supercomputers and the ongoing commitment to bolstering their capabilities with reliable artificial intelligence.
The Commission is urging the EuroHPC Governing Council to establish special access conditions for these AI supercomputers, with priority given to startups and small and medium-sized enterprises. Under the proposed regulations, hosting entities can receive up to 50% of AI supercomputer procurement and operating costs from the consortium. Ownership of these machines can be transferred to the hosting entity five years after acceptance testing.
The primary focus of the AI supercomputers will be on developing, testing, evaluating, and validating large-scale AI training models and emerging AI applications. This initiative aims to advance AI solutions within the consortium. Before implementation, the proposed regulation will undergo legal review and subsequent publication in the Official Journal of the European Union, becoming effective twenty days after publication. These regulatory efforts align with the broader AI initiative outlined by EU President Ursula von der Leyen, demonstrating a strategic commitment to strengthening Europe's AI capabilities and fostering safe and ethical AI development within the region.
