A “one size fits all” approach to the deployment of Confidential Computing does not reflect the many different demands Fortanix customers experience. When deploying Confidential Computing for customers, Fortanix needs to provide the flexibility and operational efficiency necessary for them to adapt to changes in their business landscape or the technical requirements of their individual projects.
Fortanix has previously supported Microsoft Azure Virtual Machines (VMs) with a user-configurable secure memory allocation of up to 256 GB for the Trusted Execution Environment (TEE). This capability enabled customers to specify the machine requirements for their workloads, according to their individual needs.
Fortanix is pleased to announce the introduction of a new, serverless, infrastructure model through support for Confidential Containers deployed using Azure Container Instances (ACI). The introduction of Confidential Containers on the ACI platform provides customers with important advantages when working with workloads that require high scalability and consumption-based infrastructure costs.
Instead of purchasing VMs on a time-based pricing model, where these compute resources may occasionally be left idle, customers can implement their Confidential Computing infrastructure on-demand and scale according to the demands of their protected data and applications. This is an important capability for those working in the development and deployment of AI/ML models, where model training cycles and requisite sample data sizes may exhibit significant differences between different workloads.
Confidential Containers on ACI deliver several useful benefits to Fortanix customers, and we are, initially, introducing this new infrastructure support through the Fortanix Confidential AI platform.The first benefit is increased transparency with respect to the cost of Confidential Computing infrastructure for the deployment of containerized applications.
The serverless architecture of Microsoft ACI removes the concern of costly VMs laying idle—maximizing the utilization of resources. A second benefit is the scalability provided by elastic bursting to meet the changing demands of deployed workloads. When you add in the convenience of containerized workload deployment, the new infrastructure support significantly reduces the investment required to adopt Confidential Computing.
Vikas Bhatia, Head of Product, Azure confidential computing, Microsoft said, “Fortanix continues to offer support for confidential computing solutions. Their collaboration has been invaluable in testing our Microsoft Azure confidential computing offerings early on to refine and scale to meet our joint customer needs. With the growth of AI and needs to protect both data and IP, Fortanix benefits from confidential containers on ACI to bring privacy protection to their Confidential AI solution.”
For AI developers and data science teams who need to efficiently train and run models using Confidential Computing to protect their sensitive data and intellectual property, the Fortanix Confidential AI solution with Microsoft ACI provides the right balance of security and efficiency.