Secure AI Workloads with Intel Virtualization

When AI/ML models are deployed on edge devices, like a Kiosk, they run alongside business-critical applications. Workload consolidation presents a new set of data, privacy, and intellectual property (IP) challenges. Key user and device assets must be secured. This paper proposes an architecture to virtualize AI workload at Windows-based Kiosks and Intelligent Edge devices while protecting the assets using Intel hardware-based virtualization technology (VT-x, MKTME, HLAT, CET, PTT).