NVIDIA announced at the ISC High Performance 2026 conference in Hamburg, Germany, a record expansion of artificial intelligence infrastructure across Europe, with 35 new AI HPC supercomputers currently in development. Spanning 23 countries, the initiative equips more than three million researchers with next-generation computing resources, marking Europe’s largest annual supercomputing deployment. The systems, built on NVIDIA Blackwell and Hopper architectures, will collectively deliver 800 AI exaflops of capacity since last year, underpinning the continent’s push toward AI-driven scientific discovery and industrial innovation. The expansion relies on a comprehensive NVIDIA AI stack, integrating Quantum InfiniBand networking, CUDA-X libraries, NIM microservices, and AI Enterprise software to streamline model training, simulation, and agentic AI workflows. Major national and institutional projects driving this buildout include the EuroHPC AI Factory at the Barcelona Supercomputing Center, which will upgrade its MareNostrum5 system with approximately 20 exaflops of AI training and 33 exaflops of inference performance. In Bavaria, the Blue Swan platform will deploy 1,000 GPUs to power a regional foundation model initiative focused on health, robotics, and public administration. Italy’s IT4LIA factory will utilize over 8,000 GPUs to foster open AI development across agritech, cybersecurity, and climate modeling, while Germany’s HammerHAI and Sweden’s Mimer AI Factory will establish secure, high-throughput environments for engineering simulation and life sciences research. Beyond traditional HPC, the infrastructure is accelerating targeted breakthroughs in climate science, biomedical research, and clean-energy decarbonization. Siemens Energy is leveraging NVIDIA-accelerated simulation to design hydrogen-capable gas turbine burners, reducing computational fluid dynamics modeling time by up to 77 percent and enabling rapid validation of low-carbon combustion systems. Concurrently, European institutions are pioneering hybrid quantum-classical computing through NVIDIA CUDA-Q. Partnerships involving CINECA, Fraunhofer FOKUS, the Barcelona Supercomputing Center, and the Jülich Supercomputing Centre are integrating neutral-atom and analog quantum processors with GPU clusters, extending Europe’s leadership in scalable quantum algorithms for materials science and optimization. Jensen Huang, founder and CEO of NVIDIA, emphasized the strategic shift in scientific computing, noting that AI has become the primary instrument for modern research. He highlighted that accelerated computing will enable scientists to simulate complex systems, train domain-specific AI models, and deploy autonomous workflows that transform Europe’s existing data assets into global breakthroughs. The coordinated rollout underscores a broader continental effort to achieve technological sovereignty, ensuring that European academia, government, and industry maintain competitive access to sovereign AI infrastructure. As these systems transition from deployment to active research phases, the expanded network is expected to standardize high-performance computing across European science, reducing bottlenecks in generative AI adoption, climate modeling, and industrial engineering. The infrastructure investment positions Europe to sustain long-term leadership in AI HPC while fostering cross-border collaboration on pressing environmental and health challenges.