A handful of Global South companies now run their own AI compute instead of renting it abroad - though the biggest builders are corporate-backed rather than startups. In Africa, Cassava Technologies (Econet group) is installing the continent's first AI factory: 3,000 Nvidia GPUs in South Africa, scaling to 12,000 across five countries. In India, Yotta (Hiranandani group) runs Shakti Cloud, more than 16,000 Nvidia H100 GPUs. Genuine startups are further along on the models than on the metal underneath them.
The distinction the question blurs: building compute (owning GPUs and data centres) is not the same as building models. Cassava and Yotta own the metal; startups like Lelapa AI (a five-language African model) and Sarvam (India's 22-language sovereign LLM) build the models that run on it.
Why own rather than rent is the real story. Sovereign AI is only sovereign if the compute is yours: rent foreign cloud and your models, your data, and your roadmap sit on someone else's hardware, in someone else's jurisdiction. Owning the GPUs carries far higher upfront capex, but it is what keeps the intelligence layer - and the leverage - on-continent. That capex-versus-rent choice matters more than any single GPU count.
Beyond Africa and India there is a third pole: in the Gulf, Abu Dhabi's TII builds and open-sources the Falcon models, and G42 operates at hyperscale. India's Krutrim is also standing up a domestic Nvidia GB200 cluster.