Schneider Electric launches standardised AI power Modules to cut deployment time to 6 months.
Schneider Electric recently announced a new range of standardised Prefabricated Power Modules and Power Skids, aimed at hyperscale, neocloud and colocation data center operators building out AI infrastructure. The modules can deliver up to 2.5 megawatts of capacity each, which Schneider describes as the largest single-container prefabricated power solution it has brought to market.
The new units are built around Schneider’s Galaxy VXL UPS technology and combine switchgear, lithium-ion batteries, busway and power distribution systems inside a factory-tested enclosure. A second format, power skids, houses the same core electrical infrastructure in an open, skid-based design for indoor deployment. Both are available in 2MW, 2.25MW and 2.5MW configurations – Schneider advise that the modules were designed using reference designs developed with NVIDIA for its GB300 NVL72 platform, with integrated power management and liquid cooling controls built in and multiple units can be deployed together as repeatable “power blocks” to scale capacity to tens of megawatts across a campus, they can also be combined with their modular IT pods for a complete AI-ready infrastructure platform.
Adressing the deployment time bottleneck
According to Schneider, this standardised approach is intended to address a specific industry bottleneck: customised prefabricated power solutions typically carry lead times of 12 to 14 months – a timeline that has struggled to keep pace with hyperscale and AI providers looking to expand capacity quickly. By using pre-engineered, standard SKUs and optimised manufacturing and supply chain processes, Schneider claim they can manufacture and deploy the new modules in as little as six months.
Tarunjeet Sarao, SVP of Data Center Systems at Schneider Electric comments: “By standardising our prefabricated power systems, and incorporating our leading Galaxy VXL UPS technology, we are enabling hyperscale, neocloud and colocation providers to deploy scalable, high-density power infrastructure in months rather than years. This represents a major step forward in helping customers build the next generation of AI factories, AI hubs and large-scale digital infrastructure.”
Why this matters to data center operators
Power, not compute, has increasingly become the pacing item in hyperscale AI deployment, as operators wait longer for interconnection, permitting and on-site electrical build-out than they do for GPU delivery. Prefabricated, factory-tested power infrastructure has been one industry response to that bottleneck, and several vendors have moved toward standardised, repeatable designs over custom-engineered ones for exactly this reason: standardisation is easier to manufacture at scale and faster to certify than a bespoke build. Schneider’s tie-in to NVIDIA’s GB300 NVL72 reference architecture reflects a broader pattern this year, with power and cooling vendors increasingly designing around specific GPU platforms rather than generic high-density specs.
