The push for faster token serving is increasing demand for larger inference clusters, matching the same coordinated, high-performance networking as training clusters. Scale-out connects racks and pods into larger training clusters, where InfiniBand has historically been a common choice for high-performance fabrics. But AI fabric operations is not a straight extension of enterprise networking. What Ethernet must deliver for AITo earn its place as the common AI fabric, Ethernet must handle what makes AI traffic different. AI clusters increasingly run multiple tenants and multiple jobs side by side, and a fault or noisy neighbor in one must never degrade another’s performance.