Glossary
sovereign compute
AI compute capacity owned, operated, or contractually controlled by a nation-state for the use of its own institutions and citizens, distinct from rented capacity on US-hyperscaler clouds.
National AI infrastructure: data centers, GPUsiliconA massively parallel processor originally designed for graphics, repurposed since the 2010s as the dominant compute substrate for both training and inference of large neural networks. Open full entry pools, and sometimes in-house model training, financed and governed by a state to ensure domestic access independent of foreign hyperscalers. The 2024 to 2026 wave includes UAE G42 (Stargate UAE), Saudi Humain (PIF-owned), IndiaAI Mission’s Common Compute Facility, EU AI Gigafactories program, Singapore’s AI Compute initiative, and Korea’s AI Compute plan.
The motivations stratify. Some are industrial-policy (build a domestic AI sector). Some are strategic-autonomy (avoid dependence on US-controlled clouds in geopolitical scenarios). Some are sectoral-security (defense, intelligence, finance want non-foreign hosting). All compete for the same constrained inputs: NVIDIA allocation, power, skilled operators, talent.
For the open-source AI sovereignty argument, sovereign-state compute is structurally distinct from individual-sovereign or decentralized alternatives. It re-shifts AI capacity from corporate concentration to state concentration, not from concentration to distribution. The editorial frame this site uses tracks both axes separately.