Schneider Electric has announced a strategic collaboration with Hon Hai Technology Group, better known as Foxconn, to design and scale the next generation of AI data centres. The deal, announced on 15 June, pairs the French energy-management group with the world’s largest contract electronics manufacturer.
The division of labour follows the two companies’ respective strengths. Foxconn brings compute platforms, AI rack integration, and manufacturing reach; Schneider brings power systems, cooling, and energy management. The aim, as the companies describe it, is to offer integrated, ready-to-deploy infrastructure that customers can stand up faster and run more predictably across regions.
In practice, that means co-developing reference architectures for AI data centres and working on closed-loop energy optimisation, modular power and cooling skids, and standardised design frameworks.
The skid approach, prefabricated blocks of power and cooling that ship as units, is the part aimed squarely at speed: it lets an operator assemble a facility from repeatable components rather than engineering each site from scratch. Production is expected to begin later this year, the companies said.
The collaboration sits in the part of the AI boom that gets less attention than the chips: the buildings, the power, and the heat. Training and serving large models needs dense racks drawing more electricity than conventional server halls were built for, and the constraint is increasingly the physical plant rather than the silicon inside it.
A power-and-cooling specialist tying up with a manufacturer that already integrates AI racks is a bet on that being where the next squeeze lands.
Both companies come to the deal from established positions. Schneider Electric, headquartered near Paris, is one of the largest suppliers of data-centre power and cooling equipment, a business that has grown with the AI build-out.
Foxconn, the assembler of much of the world’s consumer electronics, has been pushing into server and AI-rack manufacturing as that work has become a larger share of its order book. The partnership knits together two parts of the supply chain that customers have until now had to integrate themselves.
The closed-loop energy work named in the announcement points at the other constraint operators face, which is heat. As rack densities climb, air cooling gives way to liquid systems that capture and recirculate heat rather than simply venting it, and the engineering of those loops is one of the harder problems in a modern facility.
Standardising it into shippable modules is the efficiency the two companies are claiming, and the part most likely to appeal to operators racing to bring capacity online before demand outstrips it.
The companies did not disclose financial terms, nor did they name launch customers or specify which regions the first deployments will target. What they have committed to publicly is the joint engineering work and a timeline that starts this year.
Get the TNW newsletter #
Get the most important tech news in your inbox each week.