A substantial increase of 800 AI exaflops is now enhancing research capabilities across Europe, as NVIDIA announced the development of 35 new AI high-performance computing supercomputers. These systems, spanning national centers and academic institutions, will equip over 3 million researchers with infrastructure for advancements in areas from climate science to quantum computing. The expansion represents Europe’s largest one-year increase in supercomputing power, driven by NVIDIA’s Blackwell and Hopper platforms, alongside technologies like NVIDIA Quantum InfiniBand networking. “AI is a new instrument for science, and Europe is building the infrastructure to put it in the hands of millions of researchers,” said Jensen Huang, founder and CEO of NVIDIA. Supercomputers like Barcelona Supercomputing Center’s EuroHPC MareNostrum5 AI upgrade and BavariaAI’s Blue Swan are central to this continental effort.
Europe Deploys 35 NVIDIA AI Supercomputers
Europe is rapidly becoming a major force in artificial intelligence infrastructure with the deployment of 35 new NVIDIA AI high-performance computing supercomputers across the continent. These systems, designed to equip over 3 million researchers, signify Europe’s largest one-year expansion of this specialized computing technology and represent a substantial investment in the future of scientific discovery and industrial innovation. This buildout focuses on accelerating progress in critical fields like climate science, healthcare, and clean energy, with a particular emphasis on decarbonization efforts. The scale of this expansion is demonstrated by the 800 AI exaflops that have been deployed or announced in Europe since last year. This is a significant increase in AI capability, providing researchers with the tools to tackle increasingly complex problems.
Systems like BavariaAI’s Blue Swan, IT4LIA, and HLRS’s HammerHAI are leveraging advanced NVIDIA AI infrastructure, demonstrating a coordinated, continental approach to building this essential resource. This investment extends beyond national supercomputing centers to encompass AI factories and academic research institutions, broadening access to technology. Mateo Valero Cortés, director of the Barcelona Supercomputing Center, stated that “BSC is committed to building AI infrastructure that advances science, industry and society.” The MareNostrum5 upgrade, a collaborative effort involving Spain, Portugal, and Türkiye, will provide European researchers with the resources needed to address global challenges ranging from climate modeling to biomedical discovery. Bavaria is also contributing significantly with a project aimed at creating a multimodal AI foundation model.
Bavarian Minister of Science Markus Blume explained, “With the project ‘Blue Swan Platform,’ Bavaria is working on an innovative and independent, multimodal AI foundation model for important application areas like health and robotics.” The project centers around a large GPU cluster at Friedrich-Alexander University in Erlangen, currently the largest of its kind in Germany. This expansion isn’t limited to model training and inference; initiatives are underway to integrate AI with quantum computing, with CINECA, EuroHPC, and Pasqal integrating a neutral-atom QPU at the CINECA supercomputing center, utilizing the NVIDIA CUDA-Q platform. Fraunhofer FOKUS is also facilitating the integration of CUDA-Q with the quantum programming language Eclipse Qrisp, aiming to simplify the development of complex quantum algorithms. This multifaceted approach, combining classical and quantum computing resources, positions Europe for next-generation scientific exploration and technological advancement.
With the upgrade to MareNostrum5 and NVIDIA accelerated computing, the consortium composed of Spain, Portugal and Türkiye will make available to European researchers the tools to tackle some of the world’s most complex challenges, from climate modeling to biomedical discovery.
Mateo Valero Cortés, director of the Barcelona Supercomputing Center
NVIDIA Blackwell & Hopper Power European AI Factories
The expansion of artificial intelligence infrastructure across Europe is accelerating, evidenced by the deployment of advanced computing systems at institutions focused on both fundamental research and industrial application. A network of national supercomputing centers and specialized AI factories are undergoing significant upgrades, driven by demand for increased computational power to address complex challenges in fields ranging from climate modeling to drug discovery. This build-out isn’t simply about adding more processing capacity; it’s a strategic investment in a full-stack AI ecosystem, encompassing hardware, software, and networking technologies designed to facilitate end-to-end AI workflows. A key component of this expansion is the adoption of NVIDIA’s latest platforms, with the NVIDIA Blackwell and NVIDIA Hopper architectures powering the majority of Europe’s AI factory development. This surge in capacity is facilitating a broad range of research initiatives, including advanced simulation, scientific AI model training, and the development of agentic AI systems.
The infrastructure extends beyond providing raw compute; NVIDIA’s Quantum InfiniBand networking, CUDA-X libraries, NIM microservices, and AI Enterprise software are all integrated to create a comprehensive platform for scientific advancement. Several prominent institutions are at the forefront of this effort. CINECA’s IT4LIA project is strengthening Europe’s AI and high-performance computing ecosystem, while the High-Performance Computing Center Stuttgart’s HammerHAI is building secure, national AI infrastructure to accelerate simulation and scientific discovery, solidifying Europe’s position in advanced computing and real-world breakthroughs.
With HammerHAI, Germany’s first AI factory, we are building on that foundation with secure, national AI infrastructure that will help researchers and industrial users accelerate simulation, inference and scientific discovery, strengthening Europe’s ability to turn advanced computing into real-world breakthroughs.
Michael Resch, director of the High-Performance Computing Center Stuttgart
CUDA-Q Advances Hybrid Quantum-Classical Computing
The pursuit of practical quantum computing received a boost with advancements in hybrid approaches, particularly through NVIDIA’s CUDA-Q platform, now being integrated across several European supercomputing centers. Researchers are no longer solely focused on building fault-tolerant, universal quantum computers; instead, they are leveraging near-term quantum processors in conjunction with classical high-performance computing resources to tackle complex problems. This collaborative effort demonstrates a shift toward utilizing quantum processors as co-processors, accelerating specific computational tasks within a larger classical workflow. CUDA-Q, described as an open, qubit-agnostic platform for hybrid computing, is central to this strategy. It provides a unified software environment for programming and controlling both quantum and classical resources, streamlining the development of hybrid algorithms. For Pasqal and CINECA, the platform offers a means of developing and running quantum applications focused on optimization and materials science use cases.
Qrisp, initially developed at Fraunhofer FOKUS and now under the Eclipse Foundation’s umbrella, allows researchers to construct complex quantum algorithms that can be simulated, optimized, and ultimately executed with NVIDIA CUDA-Q. The Barcelona Supercomputing Center is also actively involved, having recently deployed an analog quantum computer from Qilimanjaro Quantum Tech, as part of the EuroHPC JU initiative. Qilimanjaro has integrated NVIDIA CUDA-Q into its quantum software development kit, QiliSDK, further expanding the ecosystem of tools and resources available to researchers. The center is also working toward making Qilimanjaro’s QPU available within the NVIDIA CUDA-Q platform, aiming for seamless control of quantum-accelerated workflows. Beyond these specific integrations, researchers at the Jülich Supercomputing Centre, in collaboration with NVIDIA, achieved a record by fully simulating a universal 50-qubit quantum computer using classical high-performance computing resources.
Jülich’s JUQCS-50 quantum simulator allows researchers to test the largest possible quantum problems on supercomputers, effectively scaling quantum computing capabilities beyond the limitations of current quantum hardware. This achievement highlights the crucial role of classical simulation in validating quantum algorithms and exploring the potential of larger quantum systems. The combined efforts across Europe demonstrate a coordinated approach to harnessing the power of hybrid quantum-classical computing, positioning the continent at the forefront of this rapidly evolving field.
AI is the new instrument of science, and Europe is building the infrastructure to put it in the hands of millions of researchers.
Jensen Huang, founder and CEO of NVIDIA
AI Accelerates Climate Modeling & Industrial Simulation
These systems aren’t simply adding capacity; they represent Europe’s largest one-year expansion of this technology, fundamentally altering the scale of possible research. This expansion is particularly crucial for climate science, where increasingly complex models are required to accurately predict future scenarios and inform mitigation strategies. Researchers are leveraging NVIDIA technologies, including the Blackwell and Hopper platforms, to simulate Earth systems with unprecedented fidelity. Siemens Energy, for example, is utilizing the Siemens Xcelerator portfolio, accelerated by NVIDIA technologies, including NVIDIA Omniverse libraries, CUDA-X and AI infrastructure, to unify design, computational fluid dynamics simulation and manufacturing for gas turbines built to run on up to 100% hydrogen. This complex physics challenge, involving extreme heat, fluid dynamics and combustion behavior, previously demanded extensive time and resources; the new workflow cuts simulation times by up to 77% to advance hydrogen-capable, low-carbon gas turbines. Beyond climate, the new infrastructure is driving innovation in industrial engineering.
Accelerated simulation allows for rapid, iterative design processes, reducing development cycles and optimizing performance. This capability extends to materials science, where AI-powered simulations are helping researchers discover and design new materials with tailored properties. “IT4LIA marks a strategic step in strengthening Europe’s AI and high-performance computing ecosystem, providing a high-performance infrastructure to the research and innovation ecosystem,” said Gabriella Scipione, high-performance computing director of CINECA. This coordinated continental effort, fueled by 800 AI exaflops deployed or announced since last year, signals a clear intention to remain at the forefront of scientific discovery and technological innovation.
IT4LIA marks a strategic step in strengthening Europe’s AI and HPC ecosystem, providing a high-performance infrastructure to the research and innovation ecosystem.
Gabriella Scipione, high-performance computing director of CINECA
