News

HUMAIN, DataVolt Brings Nebius into NEOM’s Oxagon AI Giga Campus

HUMAIN and DataVolt have signed a preliminary agreement with Nebius for up to 50 MW of AI capacity at NEOM’s Oxagon giga campus, strengthening Saudi Arabia’s expanding artificial intelligence infrastructure plans.

Written By : Poulami Saha
Reviewed By : Achu Krishnan

Saudi Arabia’s artificial intelligence infrastructure plans have received a major boost. HUMAIN and DataVolt have signed a preliminary framework agreement with AI cloud company Nebius for data-centre capacity at NEOM’s Oxagon giga campus.

Under the agreement, Nebius will become an anchor customer at the facility. The company has been allocated up to 50 MW of capacity, equal to half of the initial 100 MW currently under construction at the campus. The deal was announced on the sidelines of NVIDIA GTC Berlin.  

Campus to Scale to 360 MW

The Oxagon AI giga campus is being developed in phases. The first phase is designed to deliver 100 MW of capacity, while the broader project is planned to reach 360 MW over time.

Nebius’s allocated capacity is expected to come online in phases from the fourth quarter of 2027. The arrangement gives the project a clear early demand signal and supports its transition from construction to commercial deployment.

HUMAIN, a company backed by Saudi Arabia’s Public Investment Fund, provides full-stack artificial intelligence capabilities. DataVolt develops, owns and operates sustainable digital infrastructure. Together, the two companies are building the physical foundation for AI services at Oxagon.

Nebius Gains Strategic Foothold

For Nebius, the agreement offers access to a large-scale AI infrastructure site in the Middle East. The company will use the capacity to deliver AI cloud services from the campus.

The preliminary nature of the framework means commercial terms and final commitments still need to be completed. However, the allocation of up to 50 MW signals strong interest from a global AI cloud provider in Saudi capacity.

The deal also reflects growing competition among countries and companies to secure power, land and data-centre capacity for AI training and inference. Large AI models require dense computing clusters, reliable power supply and advanced cooling systems.

Also Read: Technologies Making UAE Travel Easier: Best Travel-Friendly Tech for You

SoFi Launches Tokenized Deposits Offering 4.4% Interest

Tesla Cybercab Sales: Elon Musk's Robotaxi Plan Targets 2026 Launch

PHP Security Flaws Trigger Millions of Hacking Attempts Worldwide

Oil Prices Surge as US, Israel Reportedly Weigh Fresh Iran Strikes

Norway's Wealth Fund Posts USD 29.7 Billion Loss in Third Quarter