Contents
Solar inverter maker SMA Solar Technology AG is bringing the power conversion and grid integration technology it developed for utility-scale plants to the data centre market. The company has introduced a portfolio of three power infrastructure architectures for AI and hyperscale data centres.
Grid access becomes a bottleneck for data centres
According to SMA’s statement of 26 August 2026, AI, cloud computing and digitalisation are rapidly increasing demand for data centre capacity, while long grid connection processes and growing stability requirements are slowing the commissioning of new facilities in many regions. SMA CEO Jürgen Reinert said access to reliable energy increasingly determines how quickly new capacity can come online.
Three solution architectures
- SMA GridAssist is a battery-based architecture that provides load balancing, power quality stability and grid code compliance for grid-connected data centres.
- SMA GridLink AC is a medium-voltage uninterruptible power supply (UPS) architecture that decouples data centre loads from the grid.
- SMA GridLink DC is a solution designed to support the shift to 800 V direct current in AI infrastructure, planned for launch in 2027.
The company says these solutions are based on the same power conversion platform, medium-voltage stations and control systems that are installed at gigawatt scale around the world. SMA’s expertise in this field covers grid-forming technology, battery storage integration, plant control and cybersecurity.
Revenue expectations for 2027
SMA expects the data centre business to contribute revenue in the mid-to-high double-digit millions of euros in fiscal year 2027 and to grow in the following years. Florian Bechtold, Executive Vice President for Large Scale and Project Solutions, said data centre developers need more than individual components; they need a partner that combines grid integration, power conversion, control and lifecycle support.
Sources
- SMA Solar Technology AG press release, “SMA introduces dedicated solutions to accelerate time-to-power for AI-driven data centers”, 26.08.2026 – sma.de
Image: SMA Solar Technology AG
Author
