Amazon Web Services (AWS) has unveiled AI Factories – a new service that turns ordinary corporate data centres into high-octane AI hubs.
Instead of shipping sensitive data to the cloud, companies can now tap into cutting-edge AI hardware and software right on-site.
Customers provide the infrastructure – space, power, and connectivity – while AWS handles the setup. The service combines Trainium chips with NVIDIA GPUs and integrates networking, storage, databases, and AI tools such as Amazon Bedrock and SageMaker AI.
AWS says the AI Factories help organisations deploy AI projects faster than building the same capability in-house, without sacrificing control over sensitive information.
The offering targets enterprises and government agencies that must meet strict regulatory and data sovereignty requirements.
“Large-scale AI requires a full-stack approach – from advanced GPUs and networking to software and services that optimise every layer of the data centre,” said Ian Buck, Vice President and GM of Hyperscale and HPC at NVIDIA.
“Together with AWS, we’re delivering all of this directly into customers’ environments.”
Data Sovereignty and Compliance
Keeping data in-house is a major driver for on-premises AI.
Enterprises and governments often require sensitive workloads to remain on-site, particularly in regulated industries such as finance, healthcare, and defence.
AWS AI Factories operate in isolated environments, providing low-latency access to compute, storage, and AI services without moving data off-site.
The company says the infrastructure can handle workloads across all classification levels – Unclassified, Sensitive, Secret, and Top Secret.
AWS is collaborating with HUMAIN, a Saudi Arabian company, to build an AI “zone” with up to 150,000 AI chips in a purpose-built local data centre.
“The AI factory AWS is building in our new AI Zone represents the beginning of a multi-gigawatt journey for HUMAIN and AWS,” said Tareq Amin, CEO, HUMAIN.
“From inception, this infrastructure has been engineered to serve both the accelerating local and global demand for AI compute.”
Technical Capabilities
The AI Factories combine NVIDIA’s Grace Blackwell and Vera Rubin GPUs with AWS’ Trainium chips and high-speed interconnects such as the Elastic Fabric Adapter and Nitro virtualisation.
AWS positions the AI Factory as a private environment similar to a dedicated AWS Region. It provides integrated AI services and support for large-scale workloads.
By managing the infrastructure, the company aims to remove much of the burden of procurement, hardware installation, networking configuration, and software integration – tasks that can normally take organisations months or years to complete independently.
Broader Market Context
AI Factories are part of a wider shift toward hybrid and on-premises AI infrastructure.
Public cloud adoption has dominated enterprise IT for over a decade. However, AI workloads are prompting organisations to reconsider fully cloud-based deployments.
Sensitive data, regulatory compliance, and the demands of modern AI models make on-site deployments increasingly attractive.
Microsoft has followed a similar approach. The company has deployed AI Factories in its global data centres to run OpenAI workloads.
It also offers Azure Local, a managed on-premises hardware solution for organisations that need strict data sovereignty.
These developments suggest hybrid models – combining cloud scalability with on-premises control – are becoming the standard for enterprise AI.
NVIDIA remains central, providing the GPU acceleration and software stack for these deployments.
Enterprises can pair this hardware with cloud management and AI services to achieve performance that would be difficult to replicate independently.
Implications for Enterprise IT
AI “factories” could give IT leaders faster access to AI infrastructure but also introduce new considerations.
Organisations must plan for costs, staffing, and integration with existing systems.
They need to decide which workloads should remain on-site and which can run in the cloud, while factoring in compliance, latency, and performance requirements.
Although AWS manages the infrastructure, organisations still require expertise in AI model deployment, system monitoring, and security.
Upskilling teams or hiring specialists will be crucial for making the most of these environments.
Looking Ahead
AWS and NVIDIA’s AI Factories highlight a strategic shift in enterprise computing.
AI is no longer just another application – it affects infrastructure, compliance, and corporate strategy.
Organisations that deploy AI effectively on-premises while retaining flexibility and control could gain a competitive advantage.
The HUMAIN partnership in Saudi Arabia demonstrates the potential for large-scale regional AI deployments.
Other countries are likely to pursue similar initiatives, balancing AI adoption with data security and regulatory requirements.
By offering a full-stack, on-premises AI solution, AWS and NVIDIA are signalling that the future of enterprise AI may be hybrid – combining cloud expertise, managed services, and cutting-edge hardware with the control and compliance benefits of on-site deployment.
