At the AMD Advancing AI event, AWS EC2 compute is evolving to meet demand from agentic AI, physical AI, and general-purpose cloud infrastructure, according to Art Baudo, principal product marketing manager and head of EC2 product marketing at Amazon Web Services Inc. Customers are using AMD-based instances for AI inference and high-performance computing, with Baudo noting that as performance has increased, so has the price performance delivered to customers. AWS introduced AMD EPYC processors into EC2 in 2018 and has shipped every subsequent generation, including Turin-based instances launched in 2026. The shift from multithreaded to single-threaded instances between the sixth and seventh generation delivered substantial performance gains for AI and EDA workloads. AWS has also introduced high-frequency instances combining 5GHz clock speeds with expanded memory configurations. Baudo stated that some of the most recent instances continue to see a reduction in price performance. The AWS Nitro System underpins security and performance throughout the full EC2 portfolio by offloading hypervisor functions to dedicated hardware and software, allowing faster deployment at greater scale while enforcing zero operator access, and extends naturally into AI workloads across AMD, Graviton, Trainium, and Inferentia instances.