NVIDIA GPU-Enabled Confidential Computing on Microsoft Azure

Hardware-based confidential computing protects data in use, as it is in AI computations of what is often highly sensitive and protected data. The key with AI, according to Microsoft Azure’s Mark Russinovich and Nvidia’s Daniel Rohrer, is to bring accelerators like GPUs under the confidential computing umbrella.

In this Dark Reading Fast Chat, Mark Russinovich of Microsoft Azure and Daniel Rohrer of Nvidia explain to Dark Reading's Terry Sweeney how the Microsoft-Nvidia collaboration enhances the security of AI workloads.

Confidential GPUs can protect data end-to-end while also unlocking multi-party confidentiality in their data sharing, according to the partners. Among the early use cases are pharmaceutical, financial services, and insurance companies pushing protected data into the public cloud for the first time, knowing that their data is well protected in those environments.