Global AI growth is increasingly colliding with electricity limits, a shift that could slow data center buildouts and reshape infrastructure planning from the US to Asia. Delta Electronics chairman Ping Cheng said the bottleneck is already delaying projects, pushing operators toward self-owned power systems and off-grid microgrids.

Quanta Computer executive vice president and Quanta Cloud Technology (QCT) president Mike Yang said GPU applications are expanding across training, edge computing, and storage, as well as in the development of application-specific integrated circuits (ASICs). In response to future demand, the company is preparing for power needs and production capacity, planning to add three more plants in the US by the end of 2026.

Lightmatter has joined Nvidia’s NVLink Fusion ecosystem, a move that could accelerate the rollout of high-performance optical links for AI infrastructure worldwide. The collaboration is aimed at easing data-center bottlenecks, improving bandwidth, and giving global customers more options for building energy-efficient AI systems at scale.