AWS, Nvidia and Humain: Data Centres Behind Saudi AI Goals

Saudi Arabia’s ambitions to expand its AI infrastructure have advanced with Humain announcing a series of partnerships with major technology companies, including Nvidia and AWS. 

The agreements were revealed at the US-Saudi Investment Forum in Washington DC and focus on new high-performance data centre capacity that will support the country’s developing AI sector.

Jensen Huang, founder and CEO of Nvidia, says: “AI is essential infrastructure: like electricity, every industry will use it and every country will build it. 

“Saudi Arabia intends to be a leading global AI hub. With Nvidia-powered infrastructure deployed across the US and Saudi Arabia, Humain and its partners are building the innovation engine for the new industrial revolution.”

A new AI Zone supported by AWS

A major element of the plan is the development of an AI Zone in Riyadh that will run on Nvidia GB300 infrastructure and AWS Trainium chips. The facility is designed for high intensity AI workloads, including model training and inference and forms part of a wider roadmap to deploy 150,000 AI accelerators.

Customers in the AI Zone will also gain access to AWS services such as Amazon Bedrock, Amazon AgentCore and Amazon SageMaker, giving organisations immediate availability of foundation models for rapid deployment.

“This marks a pivotal moment in our commitment to our partnership with Humain,” says Tanuja Randery, Managing Director for Europe, Middle East & Africa at AWS. 

“By combining Humain's local expertise and investment with AWS AI solutions – including our advanced infrastructure, hardware partnerships with Nvidia, the transformative AI platform Amazon Bedrock and AI solutions for business users including Amazon Quick Suite – we're establishing a world-class innovation hub that will serve customers across Saudi Arabia and around the globe.”

Tareq Amin, CEO of Humain, says the launch of the AI Zone represents the start of a multi-gigawatt growth plan for both companies.

Expanding Nvidia-powered data centre operations

Humain’s partnership with Nvidia extends beyond the Riyadh development. The company plans to deploy up to 600,000 of Nvidia’s latest AI systems in the next three years, strengthening the compute capacity needed for its growing data centre footprint. This includes the Nvidia GB300 platforms, which are among the most in-demand accelerators for large-scale AI workloads.

The partnership also enables Humain to use Nvidia’s Nemotron open technologies to support Humain Chat, a conversational AI service powered by Humain’s own large language model (LLM), ALLAM. The system is intended to serve the world’s 400 million Arabic speakers, addressing a gap in regional language models.

Humain will also draw on Nvidia Omniverse libraries to develop sovereign AI infrastructure, including advanced digital twin environments for modelling and simulation. These capabilities are expected to play a role in data centre planning, operational optimisation and future capacity expansion.

Tareq says: “We are working together to advance global AI infrastructure. By expanding our compute capacity in both the US and Saudi Arabia with the latest Nvidia AI infrastructure, Humain and our partners at Global AI, xAI and AWS, secure an even stronger strategic foothold. This integrated presence gives us the strength and scale to fuel the future of global AI innovation.”

Strengthening international data centre capacity

Alongside its developments in Saudi Arabia, Humain intends to expand its Nvidia-powered data centre operations into the US.

The goal is to create a distributed network of high-density compute facilities that can support its international AI partnerships and upcoming enterprise AI services.

The company’s approach centres on building scalable infrastructure that can accommodate increasing accelerator density while supporting high-performance cooling and power delivery.

The partnerships with AWS and Nvidia form the foundation of this expansion, giving Humain access to hardware, platforms and model development frameworks required for large-scale AI deployment.