Linda Yao of Lenovo joins Patrick Moorhead and Daniel Newman to discuss why AI inferencing depends on services, operations, and architectural choices, not just models or hardware.