France’s Bull Doubles Supercomputer Production as Europe Pushes for AI Independence
Europe’s ambitions in artificial intelligence depend on something that is easy to overlook when discussing AI models: computing infrastructure.
Powerful AI systems require enormous amounts of computing capacity.
That means data centres, processors, networking equipment and supercomputers are becoming strategic assets.
France is now strengthening one part of that infrastructure.
French supercomputer manufacturer Bull has reopened an expanded factory in Angers, doubling its production capacity as Europe works to increase its computing power and reduce its dependence on foreign infrastructure.
The factory can now produce approximately 12 supercomputer racks per month, compared with six previously, with plans to potentially increase capacity further. Investing.com
The expansion comes as the European Union attempts to close the computing gap between Europe and major technology powers such as the United States and China.
The EU has committed around €7 billion between 2021 and 2027 to develop a network of supercomputers across Europe. Investing.com
That investment is increasingly connected to Europe’s AI ambitions.
Artificial intelligence development requires massive computing resources.
Training advanced models can involve thousands of specialised processors operating together for extended periods.
Running those models at commercial scale also requires substantial infrastructure.
Europe therefore faces a strategic question.
Can it become a major AI power while depending heavily on computing infrastructure controlled by companies outside the continent?
European policymakers increasingly believe that some degree of technological independence is necessary.
Bull is part of that effort.
The company operates what Reuters describes as Europe’s only factory dedicated to building supercomputers.
Its systems are being deployed across European research and industrial projects, including work involving major organisations such as Airbus. Investing.com
Bull has also become an important supplier to Europe’s supercomputing programme.
The company reportedly won 15 of 18 tenders issued by EuroHPC, the European organisation responsible for coordinating and co-funding European supercomputing infrastructure. Investing.com
One important project is Finland’s LUMI-AI system, while France is also ordering additional systems.
The company is additionally preparing the Alice Recoque supercomputer for France’s nuclear agency.
These machines are not simply academic research tools.
High-performance computing is increasingly important for artificial intelligence, scientific research, climate modelling, engineering, pharmaceuticals, defence and industrial development.
That makes Europe’s supercomputing capacity an economic and strategic issue.
However, building a European supercomputer does not automatically make the entire system European.
Modern machines depend on components manufactured across the global technology supply chain.
Processors and memory remain major areas of dependence.
Bull CEO Emmanuel Le Roux has said that European-made components now represent around 70% of the company’s systems, compared with roughly 20% to 30% five years ago. Investing.com
That is a significant improvement, but it also demonstrates how difficult technological sovereignty can be.
Europe can assemble sophisticated machines while still depending on international suppliers for critical components.
Memory is one particular weakness.
Europe currently lacks a major domestic memory manufacturer capable of supplying the needs of the continent’s supercomputing industry at scale.
This means that Europe’s sovereignty strategy cannot focus on one technology alone.
It needs a complete ecosystem covering chips, memory, networking, data centres, energy and software.
The Bull factory expansion is therefore part of a much bigger European strategy.
Europe is trying to ensure that it has enough infrastructure to develop its own AI capabilities rather than simply becoming a consumer of technologies developed elsewhere.
That does not mean Europe intends to completely isolate itself from American or Asian technology.
Instead, the goal is to have enough domestic capability to maintain strategic choices.
For European AI companies, researchers and governments, that could become increasingly important.
The future AI race will not be determined entirely by whoever builds the smartest model.
It will also be influenced by who controls the infrastructure capable of running those models.
By doubling supercomputer production in France, Bull is helping Europe strengthen one important piece of that infrastructure.
