Enterprises often assume that scaling AI requires expensive GPU-heavy infrastructure, but this session challenges that assumption. It demonstrates how SAS Viya combined with Intel Xeon 6 processors can deliver efficient, cost-effective AI performance using CPU-based architectures. By leveraging SAS Retrieval Agent Manager (RAM), organizations can build context-aware AI agents that process both structured and unstructured data while maintaining trust, traceability, and actionability. The result is a scalable, “sovereign AI” approach that reduces costs while still enabling advanced agentic workflows and enterprise decisioning.
Presenters: Imran Sheik Mohamed, Intel and Sundaresh Sankaran, SAS
Watch the recording
