Amazon Web Services (AWS) announced on Tuesday the introduction of "AI Factories," a new offering designed to enable large corporations and government entities to operate AWS's artificial intelligence systems directly within their own data centers. This initiative, developed in partnership with Nvidia, targets market demand for enhanced data control and localized AI processing, according to statements from both companies.
The "AI Factories" product is engineered to address stringent data sovereignty requirements. Under this model, customers supply the necessary power and data center infrastructure, while AWS is responsible for the deployment and management of the AI system. This arrangement facilitates the processing of AI workloads without proprietary data leaving the customer's controlled environment, thereby preventing exposure to external model makers or shared hardware. AWS indicated these on-premise systems can be integrated with existing AWS cloud services, including networking, storage, databases, and security components.
The technological foundation of AWS "AI Factories" integrates innovations from both AWS and Nvidia. Customers deploying these systems have hardware flexibility, with options for Nvidia's latest Blackwell GPUs or Amazon's recently introduced Trainium3 chips. The platform also leverages AWS's homegrown infrastructure services and is designed to connect with Amazon Bedrock, AWS's service for AI model selection and management, and AWS SageMaker, its tool for building and training AI models. Nvidia's blog post confirmed the collaboration, detailing the joint effort to bring advanced AI capabilities to enterprise customers.
This strategic announcement by AWS aligns with a broader industry trend among leading cloud providers to offer robust on-premises and hybrid cloud solutions for advanced AI workloads. Microsoft, a significant competitor, previously highlighted its adoption of Nvidia AI Factory technology in October for its global data centers, specifically to power OpenAI workloads. Microsoft also has outlined "Azure Local," a managed hardware solution for customer sites, as part of its comprehensive sovereign cloud offerings, particularly for European organizations. Analysts note that these developments represent a significant investment by major cloud providers in corporate private data centers and hybrid clouds, a shift that re-emphasizes on-premises infrastructure for cutting-edge AI deployments.
