The new EuroTPC project has officially launched with backing from the EuroHPC Joint Undertaking, aiming to strengthen Europe’s influence within the global Trillion Parameter Consortium (TPC), an international initiative focused on advancing open AI for science.
The three-year programme began on 1 May 2026 and brings together leading European supercomputing and research organisations to coordinate Europe’s contribution to next-generation AI models.
Led by Barcelona Supercomputing Center, the EuroTPC project includes partners from Finland, Italy, France and Germany, including CSC – IT Center for Science, CINECA, Neovia Innovation and BADW-LRZ.
The initiative has secured nearly €1.5m in Horizon Europe funding following selection under the HORIZON-JU-EUROHPC-2025-INCO-01 call.
The project’s central objective is to ensure European priorities and standards are represented in the future development of extreme-scale AI systems.
Through closer collaboration between high-performance computing centres, AI researchers and industry, the EuroTPC project is expected to accelerate scientific discovery while reinforcing Europe’s digital sovereignty and competitiveness in advanced computing.
Europe expands its position in global AI research
The TPC was launched in 2023 to support open and collaborative development of massive AI models designed for scientific applications.
Unlike commercial large language models developed behind closed doors, the consortium focuses on transparent research, shared infrastructure and international cooperation.
The EuroTPC project will act as Europe’s coordination layer within that global effort. One of its first initiatives will be the creation of a dedicated European TPC Office to manage collaboration between research institutions, align projects with EU policy priorities and develop a long-term strategic roadmap for Europe’s participation.
This framework is intended to give European researchers a stronger voice in decisions surrounding AI infrastructure, datasets, evaluation methods and governance standards tied to scientific AI systems.
Building trustworthy AI for science
A major focus of the EuroTPC project is the development of trustworthy and reliable AI for science using open and scalable infrastructure.
The consortium plans to support work on large-scale AI models alongside the data ecosystems, benchmarking systems and computing resources required to operate them effectively.
These technologies could play an increasingly important role across scientific and industrial sectors.
Key areas include climate modelling, pharmaceutical research, materials science and advanced manufacturing, where large AI systems are already beginning to reshape research workflows.
The initiative also aims to strengthen links between Europe’s high-performance computing ecosystem and AI development efforts.
By combining supercomputing capacity with advanced machine learning research, the EuroTPC project intends to improve the efficiency and performance of trillion-parameter-scale AI systems.
Collaboration across Europe’s HPC ecosystem
The consortium will work closely with EuroHPC AI Factories, research centres, private companies and scientific communities across Europe.
Planned activities include technical working groups, hackathons and collaborative workshops designed to encourage knowledge sharing and broader stakeholder participation.
This stronger coordination is essential if Europe wants to remain competitive in the global race to develop advanced AI for science technologies.
With the United States and China investing heavily in large-scale AI infrastructure, European institutions are increasingly focused on maintaining technological independence while promoting open research standards.
Over the next three years, the EuroTPC project is expected to lay the groundwork for a more unified European approach to scientific AI development, helping position the region as a key contributor to the future of open, large-scale AI research.