NVIDIA opens AI tools, training and compute access to European researchers

NVIDIA joined the European Commission’s AI in Science Enablers Initiative on Oct. 15, pledging open courseware, research software, technical support and routes to computing resources for European researchers and startups.

The company named collaborations with AI2F in France, BSC AI Factory in Spain and Sweden AI Factory. NVIDIA also pointed researchers to free GPU access offered by the CINECA and Jülich supercomputing centers. Neither NVIDIA nor the Commission disclosed a monetary value, number of compute credits or total computing allocation for the pledge.

NVIDIA describes the initiative as a framework connected to RAISE — the Resource for AI Science in Europe — that brings private-sector support closer to the region’s research community. The Commission launched it at the AI in Science Summit in Dublin.

NVIDIA said its contribution covers accelerated-computing tools, models and platforms; cloud, data and scientific resources; training and technical support; research collaborations and pilots; and startup support. It is separate from NVIDIA’s U.S. science commitment.

The company said it has released more than 250 open models, 250 open datasets and 1,400 open-source software projects. The available sources do not independently audit those totals.

The research stack named in the pledge includes Nemotron models, BioNeMo Agent Toolkit, PhysicsNeMo libraries, Earth-2 models and ALCHEMI NIM microservices. NVIDIA says the tools can support work in biology, chemistry, molecular design, weather and climate, energy, engineering, physics and materials science. The announcement establishes access commitments, not discoveries produced with the tools.

NVIDIA’s Deep Learning Institute is making modular course materials freely downloadable under Apache License 2.0 and CC BY 4.0. The packages include lecture notes, notebooks, code samples and lab guides. NVIDIA says users may still have to pay for compute instances, depending on the environment they choose.

The three AI-factory collaborations are at different stages. AI2F completed a four-month collaboration with NVIDIA’s Inception startup program, but NVIDIA did not publish participant numbers or quantified outcomes. BSC AI Factory ran a pilot and is rolling out its first full accelerator program; the Spain-led factory also involves Portugal, Romania and Türkiye and is centered on an upgraded MareNostrum 5 system.

Sweden AI Factory’s accelerator will offer startups and small and midsize companies AI infrastructure, technical expertise, training and go-to-market support through NVIDIA Inception. The factory, led by NAISS and RISE, says it offers free AI-optimized supercomputing and expert support to startups, smaller businesses, academia and the public sector.

The collaborations sit within a public network that EuroHPC says comprises 19 AI Factories and 13 AI Factory Antennas, with free customized support for startups and small and midsize companies. NVIDIA said its support also includes office hours, bootcamps and hackathons, including an October event for Pharos-CY with European computing organizations and The Cyprus Institute.

NVIDIA separately says 35 NVIDIA-based AI and high-performance computing systems are in development across 23 European countries and would serve more than 3 million researchers. The company says the systems could support research in climate, drug development, materials, life sciences and energy. The available evidence does not independently measure that reach or document resulting scientific impact.