Antoine Delignat-Lavaud
Senior Principal Researcher -
MicrosoftAntoine Delignat-Lavaud is a researcher specializing in cryptographic protocols and cloud security. He contributes to the design and security architecture of new hardware, platforms, and services in support of the Confidential Computing initiative; in particular his current focus is verifiable privacy-preserving AI based on confidential confidential computing. He earned his PhD in Computer Science from ENS and Inria in Paris.
Session Presenting
Confidential AI with Ubuntu on Azure
Pre-recorded
Online Only
ODFP929
Enterprises are concerned about data privacy and intellectual property in cloud-based machine learning, which limits the use of valuable private data. Confidential AI provides a hardware-rooted execution environment for CPU and GPU, protecting AI data and code against privileged software. On Microsoft Azure, this uses Ubuntu confidential VMs with AMD 4th Gen EPYC processors and NVIDIA H100 GPUs. Join us to explore memory confidentiality, integrity, and remote attestation in securing AI workloads
