The EU-LAC Supercomputing Network on HPC and AI
Chairs Workshop
- Bernd Mohr, Jülich Supercomputing Centre (JSC), Germany
- Rui Carlos Oliveira, INESC TEC, Portugal.
Description of the workshop
The new EU-LAC Supercomputing Network on HPC and AI began in summer 2026. This new bi regional supercomputing network connects existing infrastructures, builds skills and opens up opportunities for joint innovation. It will enhance the use of existing high-performance computing (HPC) resources in Latin America and the Caribbean (LAC) and promote links with Europe’s advanced supercomputing ecosystem. It will enable scientists from both regions to collaborate on running simulations, analyzing big data, and testing cutting-edge processing models to address urgent global challenges, ranging from climate modelling and drug discovery to artificial intelligence.
The EU-LAC Supercomputing Network for AI will: (1) Promote shared HPC infrastructures; (2) Enhance skills with capacity building and training for researchers; (3) Boost scientific collaboration by developing joint use-cases such as drug development and climate simulations; (4) Demonstrate the capacity of networked infrastructures to work together on developing and deploying AI applications; and (5) Increase the use of the BELLA cable, the high-speed digital link connecting the two regions.
In the first stage, with coordination from the European High Performance Computing Joint Undertaking, the network will see fourteen EU and LAC countries bringing together their existing computing capacity: Argentina, Brazil, Chile, Colombia, Costa Rica, Finland, France, Germany, Italy, Mexico, Slovenia, Spain, Portugal and Uruguay. Over time, it will expand to include more countries that are member of the EU-LAC Digital Alliance.
One outcome of the project will be the creation of shared use cases that showcase the practical applications and societal advantages of HPC and AI collaboration between the EU and LAC. The use cases will cover a variety of areas, including climate modelling, natural hazard prediction, large language models (LLMs) and the energy transition.
The workshop will introduce the new initiative and the planned use cases.