توسّع الذكاء الاصطناعي في المملكة

8.4 GW of reported capacity across Saudi Arabia’s operating data-center base and three major campus plans.

The calculation: 467 MW in the national operating baseline, HUMAIN’s planned 6 GW Al-Saad campus, DataVolt’s 1.5 GW Oxagon campus and SDAIA’s 480 MW Hexagon project. This is selected mixed-stage coverage, not 8.4 GW operating or committed AI capacity.

NVIDIA. AMD and Cisco. AWS. xAI. Luma AI. Together AI. These partnerships show the breadth of the compute ecosystem taking shape. Their announced capacities can overlap with campus plans, so they are not added again.

Saudi Arabia’s AI opportunity extends from the power system to the cloud.

The national data-center plan reaches 3 GW by 2030 and 6.9 GW by 2034. Alongside it are major campus ambitions: HUMAIN’s planned 6 GW Al-Saad campus and DataVolt’s 1.5 GW Oxagon campus.

But the market is wider than those two projects. Aramco Digital, Oracle, stc/center3, Mobily and Zain are part of the infrastructure and service base, alongside HUMAIN’s technology and cloud partnerships.

This Renewable Vision report maps who is involved, what is operating, what is planned and where the energy opportunity lies. It also distinguishes campus ratings from phases and contracts, so overlapping announcements are not counted twice.

The power context matters: the earlier Renewable Vision register covers about 70 GW of renewable projects across stages. Its priced sample has capacity-weighted historical benchmarks of 1.47 US cents/kWh for solar and 1.77 for wind. These are generation benchmarks, not round-the-clock delivered tariffs.

Competitive AI exports will depend on turning that starting position into dependable power, efficient cooling, advanced compute and customers.