Saudi Arabia’s push to become a regional artificial intelligence hub advanced this week as HUMAIN, the Kingdom’s state-backed AI company, confirmed it has begun operating production-grade AI compute infrastructure built on hardware from AMD and networking technology from Cisco. The rollout marks one of the most concrete steps yet in translating Gulf AI ambitions into operational data centre capacity.
The infrastructure combines AMD’s Instinct MI355X graphics processing units, designed for large-scale AI training and inference, with AMD EPYC server processors and Cisco’s Silicon One networking silicon, which is built to move data between thousands of GPUs with the speed and reliability that modern AI workloads require. Together, the companies describe the stack as capable of supporting next-generation AI systems at scale, from large language models to enterprise AI applications.
According to the companies, this is not a pilot or proof-of-concept deployment but a live, production environment now processing workloads within Saudi Arabia. The move signals that the Kingdom’s AI infrastructure buildout has moved past the planning stage and into active operation, a milestone regional officials have been working toward as part of broader economic diversification efforts under Vision 2030.
Scaling Toward 250 Megawatts
AMD, Cisco and HUMAIN said they intend to expand the deployment substantially, with plans to scale total AI infrastructure capacity to as much as 250 megawatts. That figure represents a major increase in compute density and underscores the scale of investment being funnelled into sovereign AI capabilities in Saudi Arabia, where power availability and land have been positioned as competitive advantages for hosting energy-intensive data centres.
The expansion reflects a broader pattern across the Gulf, where governments are racing to build domestic compute capacity rather than relying solely on infrastructure hosted abroad. Sovereign AI capacity has become a strategic priority for Gulf states seeking to control data residency, reduce dependence on foreign cloud providers, and attract AI-driven investment and talent. Saudi Arabia’s HUMAIN, launched with government backing, has emerged as the primary vehicle for these ambitions, striking partnerships with major US technology firms to secure access to advanced chips and networking equipment.
The United Arab Emirates has pursued a parallel strategy, with entities such as G42 and other state-linked technology groups building out AI compute capacity and striking their own agreements with US chipmakers and cloud providers. The competition between Saudi Arabia and the UAE to establish themselves as the region’s leading AI infrastructure hub has intensified over the past two years, with both countries courting the same pool of global technology partners, including Nvidia, AMD, Microsoft and Cisco.
For the UAE and other GCC states watching the Saudi deployment, the AMD-Cisco-HUMAIN partnership offers a template for how foreign chipmakers and networking vendors are structuring deals with regional operators: supplying advanced hardware and technical expertise while local entities manage deployment, energy provisioning and regulatory alignment. Analysts have noted that this partnership model is likely to become increasingly common as Gulf governments seek to build sovereign AI capacity without developing semiconductor manufacturing or networking technology domestically.
The scale of the planned expansion, and the involvement of established US technology suppliers, is likely to reinforce Saudi Arabia’s positioning in a regional AI infrastructure race that has significant implications for data sovereignty, technology partnerships and investment flows across the Gulf Cooperation Council.
