Enterprise AI Inference at Scale with AMD and CVS Health

Abstract

CVS Health is building a heterogeneous AI inference architecture using AMD GPUs across on-premises and cloud environments. This session shows how CVS Health uses a vLLM-based Semantic Router to direct workloads across AMD MI350p, MI350x, and hosted models to optimize cost, compliance, and performance. Attendees will also learn how the AMD Enterprise AI stack and AIMs architecture simplify deployment of efficient, scalable, production-ready AI infrastructure.