During a discussion with Dark Reading’s Terry Sweeney, Microsoft Azure’s Mark Russinovich and Nvidia’s Daniel Rohrer explained the importance of hardware-based confidential computing in protecting data during AI calculations, particularly when dealing with sensitive and protected information. The key to enhancing security in AI workloads, they noted, lies in including accelerators like GPUs in the confidential computing framework.
The collaboration between Microsoft and Nvidia aims to bring about confidential GPUs that can safeguard data throughout its lifecycle while enabling secure data exchange between multiple parties. This development is especially beneficial for industries with highly sensitive data, such as pharmaceutical, financial services, and insurance companies, who are now able to securely transfer their protected data to the public cloud for the first time.
The concept of confidential computing involves safeguarding data in use, a crucial aspect when dealing with AI calculations that involve highly sensitive information. By integrating accelerators like GPUs into the confidential computing framework, Microsoft Azure and Nvidia are working towards enhancing security in AI workloads.
During a discussion with Dark Reading, Microsoft Azure’s Mark Russinovich and Nvidia’s Daniel Rohrer elaborated on how their collaboration is improving security for AI workloads. They emphasized the importance of protecting data throughout its lifecycle, particularly when sharing sensitive information between multiple parties.
The partners highlighted that confidential GPUs play a vital role in securing data end-to-end, providing a level of confidentiality that was previously lacking in data exchange processes. This development is particularly significant for industries like pharmaceutical, financial services, and insurance, which deal with highly sensitive data and are now able to securely transfer this information to the public cloud.
In summary, the collaboration between Microsoft Azure and Nvidia aims to enhance security in AI workloads by incorporating accelerators like GPUs into the confidential computing framework. This development ensures that data is protected throughout its lifecycle and enables secure data exchange between multiple parties, benefiting industries with highly sensitive information.
Article Source
https://www.darkreading.com/cloud-security/nvidia-gpu-enabled-confidential-computing-on-microsoft-azure
