Schneider Electric and AMD released the first co-engineered reference design for Helios rack-scale AI clusters, enabling up to 10.4 MW of IT capacity with 246 kW per rack, centralized around liquid cooling and a target PUE of about 1.12 to simplify and accelerate high-density AI deployments. Meanwhile, Schneider Electric is broadening AI data center designs through a collaboration with Nvidia to provide publicly available reference designs, reflecting a broader industry push to align compute, power, and cooling across AI infrastructure.