Karman and Legrand Partnered on AI Power Infrastructure
The collaboration integrates AI-driven power orchestration into data center distribution hardware to boost efficiency.
Updated on Oct. 6, 2026 in Data Centers
Live Poll
Do you believe new AI-driven infrastructure investments will make digital services more affordable for you?
Karman has partnered with Legrand to embed artificial intelligence into data center power distribution units. This integration aims to optimize energy usage for high-demand GPU operations.
Why it matters
Data centers often maintain significant amounts of idle power, which this technology seeks to reclaim. By providing real-time visibility, operators can manage energy budgets more effectively to support intensive computational tasks.
The Karman platform utilizes a custom NVIDIA Jetson Orin Nano to provide microsecond power visibility. Legrand will incorporate this into its existing infrastructure, including AC rack PDUs and DC power shelves.
The players
Karman
Karman is an engineering firm that specializes in providing embedded AI solutions for power infrastructure.
Legrand
Legrand is a multinational manufacturer of electrical and digital building infrastructure products that reported 9.5 billion euros in sales in 2025.
NVIDIA
NVIDIA is a technology company whose Jetson Orin Nano hardware serves as the foundation for the Karman AI platform.
The details
The partnership enables data center operators to automate power scheduling and gain predictive control over their electrical environment. Legrand, which owns brands like Raritan and ZPE Systems, will integrate the platform across its global product lines to help lower the price per GPU hour.
Timeline
In 2025, Legrand reported annual sales of 9.5 billion euros.
The partnership between the two companies was announced on October 6, 2026.
The first Karman-enabled Legrand products are expected to launch in mid-2027.
The Tech Race
This partnership reflects a broader industry shift toward maximizing the utility of contracted data center power. It follows the growing trend of embedding high-performance edge computing components into standard utility hardware to compete in the expanding AI infrastructure market.
Data center operators will gain the ability to increase AI token generation output without needing additional power provisioning. This efficiency move aims to reduce operational costs, potentially lowering the per-hour price for AI compute services.
The takeaway
This collaboration signals an important move toward smarter, AI-integrated infrastructure within the data center ecosystem. Operators looking to maximize performance should track the mid-2027 release to assess how these controls might impact their own power management strategies.
Further reading
Learn more about the latest innovations in Data Centers.
Live Poll
Do you believe new AI-driven infrastructure investments will make digital services more affordable for you?
