NVIDIA joins EU AI in Science initiative with tools, training and compute

NVIDIA said on Oct. 15, 2026, that it is joining the European Commission’s AI in Science Enablers Initiative, adding open models, training resources, compute access and startup support to a continentwide effort aimed at scientific research. The announcement matters because the company framed the pledge as support for the practical hurdles that often sit before research itself: tools, expertise and infrastructure, especially for smaller labs, early-career scientists and science-based startups.

Evidence and context

In the blog post published Thursday, NVIDIA said the initiative is connected to RAISE, or the Resource for AI Science in Europe, and is intended to bring private-sector support closer to Europe’s research community. European Commissioner for Startups, Research and Innovation Ekaterina Zaharieva said the campaign turns the EU’s ambition in AI for science into “concrete commitments” that give researchers access to the tools, expertise and resources they need.

NVIDIA described four areas of support: access to accelerated tools, models and platforms; compute, cloud resources, data and scientific resources; skills, training and technical support; and collaborative projects and pilots with support for researchers and talent. The company said the pledge begins with open-source technology, citing more than 250 open models, 250 open datasets and over 1,400 open-source software projects spanning large language models, healthcare, robotics, climate science, physical AI, autonomous vehicles and scientific research. Those materials are available on Hugging Face and GitHub, according to the company.

The company also pointed to scientific software and model families including NVIDIA Nemotron, BioNeMo Agent Toolkit, PhysicsNeMo, Earth-2 models and ALCHEMI NIM microservices, which it said can be used across biology, chemistry, molecular design, weather and climate, energy, engineering, physics and materials science. NVIDIA separately said its Deep Learning Institute has launched a new open-source courseware hub with freely downloadable, openly licensed materials that can be adapted for university curricula, supercomputing center bootcamps or developer community events.

Compute access and European partnerships

On compute, NVIDIA said it is working with cloud infrastructure partners and European national AI factories in France, Spain and Sweden to create new routes for startups and small and medium-sized enterprises to access sovereign European compute and services. The company said its Inception program for startups provides access to compute resources and technical support. It added that AI2F in France has worked with Inception for four months, while the Barcelona Supercomputing Center’s AI factory has moved from a pilot to a first full accelerator program. NVIDIA also said the Sweden AI Factory has announced a collaboration to open infrastructure and expertise to Inception startups and SMEs across life and materials sciences, autonomous systems and gaming.

The company said EuroHPC supercomputing centers such as CINECA and JSC provide academia and startups with free access to NVIDIA GPU infrastructure and services for open research. It also said it will support office hours, bootcamps and hackathons, and named an European AI hackathon this month with Austrian Scientific Computing, the Leibniz Supercomputing Centre, the High-Performance Computing Center Stuttgart and The Cyprus Institute for Pharos-CY.

Broader infrastructure buildout

NVIDIA tied the pledge to its wider European infrastructure push, saying 35 supercomputers are currently in development across the continent. In related company reporting on June 22, 2026, NVIDIA said those systems would support more than 3 million researchers. The company has described the supercomputers as part of a larger effort to expand AI infrastructure, though the systems under development should not be treated as fully operational.

Separately, NVIDIA announced on Oct. 13, 2026, that it would commit $250 million to UK-hosted compute access over five years for agentic AI safety and security research. That earlier pledge is distinct from Thursday’s EU initiative and is a compute-access commitment rather than an unrestricted cash grant.

The company’s latest announcement adds to a sequence of Europe-focused infrastructure and research partnerships, but the source material does not provide independent measurements of research outcomes or adoption rates. It also does not specify funding amounts for the European Commission initiative itself. The timing of any next verified catalyst was not supplied.