News
DealsCorvex agrees USD 103 million take-or-pay deal to perform AI inference with Mid-Atlantic GPUsSovereignNeysa scales AI infrastructure in India with AMD Instinct MI350X GPUs on TytanInferenceMicrosoft Foundry adds Baseten, DeepSeek AI inference (MSFT:NASDAQ)ClustersDataOne and Rowan University sign engineering co-op program agreement for DataOne's Vineland AI factoryClustersLook inside Cal Poly's $3 million AI Factory (Video)DealsOpenAI Taps AMD Chips for 457 MW of AI Capacity in Six-Year TensorWave DealClustersCal Poly launches an AI data center on campus. Take a look insideInferenceArista takes aim at the next AI network battleground: distributed inferenceDealsOpenAI Signs 457 MW AI Data Center Deal as AMD Expands Its Challenge to NvidiaClustersBehind Dell AI Factory’s NVIDIA & AMD Enterprise ExpansionDealsCorvex agrees USD 103 million take-or-pay deal to perform AI inference with Mid-Atlantic GPUsSovereignNeysa scales AI infrastructure in India with AMD Instinct MI350X GPUs on TytanInferenceMicrosoft Foundry adds Baseten, DeepSeek AI inference (MSFT:NASDAQ)ClustersDataOne and Rowan University sign engineering co-op program agreement for DataOne's Vineland AI factoryClustersLook inside Cal Poly's $3 million AI Factory (Video)DealsOpenAI Taps AMD Chips for 457 MW of AI Capacity in Six-Year TensorWave DealClustersCal Poly launches an AI data center on campus. Take a look insideInferenceArista takes aim at the next AI network battleground: distributed inferenceDealsOpenAI Signs 457 MW AI Data Center Deal as AMD Expands Its Challenge to NvidiaClustersBehind Dell AI Factory’s NVIDIA & AMD Enterprise Expansion
EveryNeocloud
Blue Lion

Blue Lion

German supercomputer for AI and climate science

Blue Lion is a high-performance supercomputer being built for the Leibniz Supercomputing Centre in Germany. It uses Nvidia Vera Rubin chips to deliver thirty times more power than its predecessor, supporting climate modelling and physics research across Europe.

Share on WhatsApp

  • Delivers 30x more power than SuperMUC-NG
  • Runs on Nvidia Vera Rubin architecture
  • Built by HPE with liquid cooling
  • Supports kilometre-scale climate models
  • Budget of €250 million

What it is

Blue Lion is a new supercomputer designed for the Leibniz Supercomputing Centre (LRZ) in Germany. It replaces the existing SuperMUC-NG system with significantly higher computational capacity. The machine is built by HPE using Cray technology and runs on Nvidia’s upcoming Vera Rubin platform. This architecture combines custom CPUs and GPUs to handle complex scientific simulations and artificial intelligence tasks. The system supports researchers working on climate models, turbulence, and physics. It also enables collaborative projects across Europe, allowing scientists to share resources and data efficiently.

How it came about

The LRZ, part of the Gauss Centre for Supercomputing, previously operated SuperMUC-NG. Plans for its successor initially cited only next-generation Nvidia accelerators. Confirmation later identified the specific hardware as the Vera Rubin architecture. This marks a shift from traditional Intel-based processors to a unified Nvidia CPU and GPU design. The project has a budget of €250 million. Construction aims for completion in early 2027. This timeline follows similar deployments in the United States, such as the Doudna system at Lawrence Berkeley National Lab.

The compute

Blue Lion uses the Nvidia Vera Rubin superchip. This platform pairs the Rubin GPU with the Vera CPU, Nvidia’s first custom central processing unit. The two components share memory and communicate via high-bandwidth links. The system employs HPE Slingshot interconnect for fast data movement between nodes. Storage and networking are optimized for low-latency workflows. The hardware supports both classic simulation and modern machine learning. Jobs can scale across the entire cluster, blending data analysis with predictive modelling in real time.

Where the GPUs are

The supercomputer resides at the Leibniz Supercomputing Centre in Garching, near Munich. HPE builds the system using its direct liquid-cooling technology. This method uses warm water circulating through pipes to remove heat without fans. The captured thermal energy is reused to warm nearby buildings, improving overall efficiency. The site serves as a hub for European scientific collaboration. Researchers from various institutions can access the machine remotely to run large-scale experiments in climate science and materials discovery.

Customers and deals

Primary users include scientists at the LRZ and partners within the Gauss Centre for Supercomputing. The system supports broader European research networks. Key applications involve creating generative climate foundation models with kilometre-scale precision. These models help predict long-term weather patterns and environmental changes. Physicists and engineers also use the cluster for turbulence studies and fusion energy research. The platform allows multiple teams to work simultaneously on diverse scientific problems.

Money and ownership

The project operates under a €250 million budget. This funding covers the acquisition of Vera Rubin chips and associated infrastructure. The final number of processing units depends on component availability and pricing. HPE delivers the hardware as part of the contract. The investment reflects the growing cost of high-performance computing infrastructure. No public stock listing or private equity details apply, as the facility is a public research institution.

Its place in AI

Blue Lion represents a major step in European sovereign compute capability. It displaces Intel processors in favour of an all-Nvidia stack, a notable shift in supercomputing trends. The system aims to lead in energy-efficient scientific computing. By reusing waste heat, it addresses sustainability concerns common in large data centres. Its performance targets support advanced climate modelling, positioning Germany as a key player in global environmental science.

Questions and criticism

Power consumption remains a critical factor. While the system is more efficient per calculation, total energy use will rise compared to older machines. The budget limits the number of chips that can be purchased, potentially capping peak performance. Dependence on a single vendor for both CPUs and GPUs raises supply chain considerations. Delays in chip production could affect the planned 2027 launch date.

Questions people ask about Blue Lion

What chips does Blue Lion use?

It uses Nvidia Vera Rubin superchips. This platform combines the Rubin GPU and Vera CPU into a single coherent system. It replaces previous Intel-based designs with an all-Nvidia architecture for better AI and simulation performance.

When will Blue Lion be ready?

The system is planned to come online in early 2027. It is currently being built by HPE for the Leibniz Supercomputing Centre in Germany. The launch follows the release of the underlying Vera Rubin hardware in late 2026.

What is Blue Lion used for?

It supports scientific research in climate modelling, physics, and machine learning. Users can run generative climate models with high precision. The system also aids in turbulence studies and fusion energy research for European scientists.

How is Blue Lion cooled?

It uses HPE’s direct liquid-cooling system. Warm water flows through pipes to remove heat from the racks without fans. This waste heat is then reused to warm nearby buildings, improving the facility’s overall energy efficiency.

Good reading and sources

    Where this page comes from

    There is no English Wikipedia article on Blue Lion yet, so the facts on this page come from the sources listed above: blogs.nvidia.com/blog/blue-lion-vera-rubin/, eenewseurope.com/en/vera-rubin-chips-displace-intel-for-blue-lion-supe. Some are the subject's own pages, so its claims about itself are reported as its claims. EveryNeocloud wrote the text in its own words, with AI assistance, and compares it with the sources to make sure no passage is copied.

    It may contain mistakes, and figures change over time. Nothing here is investment or buying advice. If something is wrong, tell us. Last refreshed 2026-11-14.

    Photo credits on this page

    The small pictures above come from Wikimedia Commons under free licences. Some are cropped to fit.

    The weekly digest

    Every Monday: the provider of the day, new pages on EveryNeocloud, and the conferences and anniversaries of the week ahead.