AMD is leading the UALink Consortium following the ratification of its next-generation open accelerator interconnect standards. This initiative establishes an open and scalable infrastructure for AI data centers and cloud operators.
The updated UALink specification introduces new capabilities including In-Network Compute and a Chiplet Definition. These standards are designed to lower latency and reduce reliance on proprietary vendor ecosystems.
This move signals AMD’s transition toward becoming a full-stack AI compute provider. The company is now influencing large-scale system architecture rather than solely supplying CPUs and GPUs.