Skip to content

Du launches the sovereign Hyper AI platform in the Emirates: local Nvidia computing to move enterprises to production agents

Share
Du launches the sovereign Hyper AI platform in the Emirates: local Nvidia computing to move enterprises to production agents

Listen to this article

Read by Anchor

Through its "du tech" arm, du offers an institutional solution to the fragmented AI computing infrastructure in the UAE market by launching the National Hyper AI platform aimed at sovereign inference and the operation of independent agents, targeting government bodies, large corporations and small-and-medium enterprises within the country to move beyond isolated pilots and preparatory projects and to push generative solutions and smart agents directly into field-work environments at scale, in line with the national AI strategy.

The new platform bundles three essential services under a single contractual and commercial umbrella: sovereign AI inference, graphics processing as a service based on direct Nvidia hardware, and model-as-a-service using a curated suite of open models, together with workflows for independent agents and tools for managing the machine-learning lifecycle, a combination that addresses model deployment hurdles, infrastructure complexity and high operating costs while streamlining procurement for organisations subject to strict regulations.

The platform is built directly on the sovereign cloud National Hyper Cloud, officially approved by the UAE Cybersecurity Council.This integration enables organisations to connect critical workloads, sensitive databases and enterprise applications to high-efficiency inference capabilities without routing data to cloud environments outside the country, reducing integration complexity and moving compliance, governance and national security controls to the AI layer itself.

Unlike services that rely on rented intermediary servers and multiple providers, du owns the entire digital stack and operates it through five live data centres in the UAE, featuring large-scale liquid-cooled facilities, national fibre-optic networks, advanced 5G connectivity and technology-partner ecosystems, ensuring that customer data, model weights and compute capacity remain within the nation’s borders under a single accountable national operator to meet ministry and security-agency requirements.

Jassim Al-Awadi, chief executive of du’s information and communications technology division, explains that shifting to an economy driven by smart agents requires decisive guarantees of sovereignty, predictable costs and unified accountability after the initial trials succeed.The company confirmed that the platform is available immediately to all government entities and business organisations across the Emirates for practical deployment.

What this means for the region

The launch provides a practical benchmark for the infrastructure needed in Gulf and regional markets, where banks and government bodies have long hesitated to adopt AI models because of data localisation and confidentiality requirements; offering Nvidia computing via direct hardware and liquid-cooled data centres under a single local contract lowers inference costs and removes regulatory-audit risks associated with international public clouds.

This direction forces startups and tech teams in the Gulf and Egypt to adapt to environments that prioritise locally managed open-weight models rather than relying solely on closed-model APIs, and it pushes regional telecom providers to offer integrated compute layers that manage the lifecycle of smart agents instead of merely selling raw cloud capacity.

Don't miss the next story

Subscribe for updates