Skip to content
AI.info

The Pulse

Bull Doubles Production Capacity at Its €80M Angers Factory

Bull opened a new production building in Angers, France, as part of an €80 million project that doubles the company’s manufacturing capacity. The expanded site is set to build supercomputers and AI infrastructure, including Alice Recoque an

Bull Doubles Production Capacity at Its €80M Angers Factory

AI.info Team · Written with AI ·

“This new industrial facility will support the next generation of supercomputers, AI infrastructures and advanced computing technologies, helping nations and organisations strengthen their strategic autonomy.”

Emmanuel Le Roux, CEO, Bull

Bull opened a new production building in Angers, France, on October 2, completing a major stage of an €80 million reconstruction of its historic industrial site. The company says the project doubles production capacity and adds space to build and test supercomputers, AI infrastructure and enterprise servers, with quantum technologies also in scope.

€80 million for a 20,000-square-metre site

The redeveloped site covers 20,000 square metres, including 10,700 square metres devoted to production. Bull projects a 30% productivity increase and a 30% reduction in building maintenance costs. The wider reconstruction and modernisation project is scheduled for completion in 2028.

The new building uses production areas that can be reconfigured as manufacturing needs change, alongside plans for more automation, virtualisation and collaborative robots. Bull says the site includes more than 3,500 photovoltaic panels across 7,000 square metres, with a combined peak capacity of 1,500 kilowatts. A heat-recovery system is designed to send 5 gigawatt-hours of recovered heat each year into Angers’ district heating network.

Two large systems a year, then four

The capacity increase has a specific measure in supercomputer output. Le Monde reported that CEO Emmanuel Le Roux said the old factory could produce two systems the size of Alice Recoque each year; the expanded plant is expected to handle four. That comparison describes large exascale-class machines, not every product assembled at Angers.

Bull says the factory is scheduled to produce Alice Recoque, a French-hosted exascale supercomputer, and LUMI-AI, a system for the Finnish LUMI AI Factory, in 2027. The company’s announcement identifies Angers as Europe’s only industrial site capable of assembling supercomputers at large scale. Its work has also contributed to JUPITER, Europe’s first exascale supercomputer.

Bull opens the factory to outside projects

The inauguration also marks the launch of Bull’s Open Factory programme, which offers manufacturing capabilities to companies developing AI, advanced-computing and quantum technologies. Bull says the programme is intended to help firms move from research and development into industrial-scale production and to bring more infrastructure manufacturing to Europe.

The company points to its partnership with Foxconn as the programme’s first concrete step. Bull says the partners aim to design and manufacture AI servers in Europe based on Nvidia’s Vera Rubin NVL72 platform, for AI factories and cloud providers. The announcement does not give a production start date or an output target for those servers.

More assembly capacity, not yet full chip independence

The factory expansion strengthens Bull’s ability to assemble and validate large computing systems in Europe, but it does not by itself remove reliance on foreign components. In its October 2 report, Le Monde said Bull estimates that foreign technologies account for about 60% of its most critical parts, including processors and networking equipment.

That gap sets a limit on what doubling factory capacity means for European autonomy: more systems can be built at Angers, while key components still come from outside Europe. Bull’s immediate plans span the next generation of supercomputers, enterprise servers and Nvidia-based AI servers; the new building gives those projects more production space, but the company has not published a unit-by-unit capacity forecast.

Sources

Explore

More articles