NVIDIA has announced that 35 new AI supercomputers are under development across Europe, spanning 23 countries and delivering up to 800 exaflops of AI compute. The announcement, made at the ISC High Performance 2026 conference in Hamburg, Germany, represents the largest coordinated AI supercomputing expansion in European history.
According to NVIDIA's official newsroom, the systems range from national-scale AI factories to regional research clusters, collectively establishing Europe as a major hub for AI computing infrastructure. The announcement was widely covered by specialist publications including HPCwire, Neowin, The Quantum Insider, and Investing.com, all of which confirmed the details. For more context on this story, see our ongoing more AI stories.
JUPITER: Europe's First Exascale System
At the center of the announcement is JUPITER, Europe's first exascale supercomputer, located at Germany's Forschungszentrum Jülich research center. NVIDIA reported that JUPITER runs on NVIDIA Grace Hopper Superchips and NVIDIA Quantum-X800 InfiniBand networking, and has been operational and heavily utilized throughout 2026.
JUPITER represents a milestone for European high-performance computing, providing researchers with exascale capabilities — performing at least one quintillion calculations per second — for the first time on the continent. The system supports a wide range of scientific applications, from climate modeling and weather prediction to drug discovery and materials science. NVIDIA described it as demonstrating "what exascale science looks like," highlighting the system's busy year of scientific output since coming online.
The Vera Rubin Platform
NVIDIA also used the ISC conference to announce its Vera Rubin platform, designed to deliver what the company described as "world-class supercomputers for science." The platform combines native double-precision (FP64) performance — critical for scientific computing that requires high numerical accuracy — with NVIDIA's CUDA-X libraries and specialized tools for scientific workloads.
In a separate announcement from Los Alamos National Laboratory (LANL), three new supercomputers named Mission, Vision, and Veritas will be built using HPE systems powered by NVIDIA Vera CPUs. These systems are designed to accelerate scientific discovery and unlock agentic AI capabilities for research, representing a significant investment in AI-powered scientific computing by the U.S. Department of Energy.
Additionally, NVIDIA introduced new software tools at ISC designed to accelerate AI for science. The NVIDIA DAQIRI library and new NVIDIA ALCHEMI NIM services target applications ranging from chemistry and materials discovery to the search for dark matter, demonstrating the company's commitment to building a comprehensive software ecosystem around its hardware.
European Sovereignty in AI Compute
The expansion of AI supercomputing capacity across Europe reflects growing political and strategic concern about technological sovereignty. Wccftech reported that the 35 systems span 23 countries, ensuring broad geographic distribution of compute resources rather than concentration in a single nation or region. This distributed approach is deliberate, aligning with European Union policy objectives to build domestic AI infrastructure and reduce dependence on foreign cloud providers.
The systems will serve universities, research institutions, and government agencies across the continent, from established computing centers in Germany, France, and the Netherlands to emerging facilities in Eastern and Southern Europe. By distributing compute capacity widely, the initiative aims to ensure that researchers throughout the EU have access to world-class AI computing resources regardless of their location.
The Quantum Insider noted that the announcement also encompasses quantum-GPU computing integration, signaling convergence between traditional high-performance computing and emerging quantum technologies. This hybrid approach could position European research institutions at the forefront of next-generation computing paradigms.
Scale and Investment
The 35 systems collectively represent a massive investment in European computing infrastructure. With combined performance reaching up to 800 exaflops of AI compute, the deployment rivals the largest AI computing installations in the United States and China. For context, one exaflop equals one quintillion floating-point operations per second — and 800 exaflops of aggregate AI compute represents an extraordinary concentration of processing power.
Investing.com reported that the initiative involves collaboration between NVIDIA, European governments, and research organizations, with funding coming from a combination of national budgets, EU programs, and institutional investments. The scale of the deployment reflects the priority that European policymakers have placed on closing the AI infrastructure gap with the United States and China.
AI for Science
Beyond raw compute power, NVIDIA emphasized the scientific applications enabled by this infrastructure. The NAIRR (National AI Research Resource) program, powered by NVIDIA infrastructure, was highlighted for driving innovative research across more than 700 projects over the past two years. These projects span protein prediction, infectious disease research, and a wide range of other scientific domains.
Eco Wave Power, a company featured in a separate NVIDIA announcement at ISC, demonstrated how AI infrastructure and digital twins can be used to optimize wave energy conversion — turning ocean waves into electricity using AI-powered optimization systems. This type of application illustrates the broad impact that expanded AI computing capacity can have beyond traditional tech industry use cases.
Global Competition
The announcement underscores NVIDIA's central role in the global AI infrastructure buildout. While the company faces increasing competition from custom AI chips developed by cloud providers like Google, Amazon, and Microsoft, as well as from semiconductor companies like AMD and Intel, its Grace Hopper and Vera Rubin platforms remain the foundation for most large-scale AI supercomputer deployments worldwide.
For Europe, the expansion represents a significant step toward technological autonomy in AI. By building domestic computing capacity at this scale, European nations reduce their reliance on U.S.-based cloud providers for critical AI research and development work. This strategic shift could have far-reaching implications for the global balance of AI research, potentially enabling European institutions to compete more effectively with their American and Chinese counterparts in the years ahead.
As AI continues to transform scientific research, industrial processes, and public services, access to world-class computing infrastructure will be a decisive factor in determining which regions and institutions lead the next wave of innovation. NVIDIA's 35-system European deployment ensures that the continent will have a seat at the table.
---
Stay Ahead of AIGet the latest AI news, analysis, and breakthroughs — all in one place.
Read more AI news →
