Nvidia used a reception at the Grand Egyptian Museum on September 21 to mark what it calls the shift from potential to production for Egypt's AI ecosystem, pointing to four AI factories announced or brought online across Africa in the past year and another 656 megawatts of capacity in the pipeline. The event featured a keynote from Paolo Guglielmini, Nvidia's vice president for EMEA, and a session from regional AI adoption lead Ahmed Mostafa. In Egypt specifically, the Nvidia Deep Learning Institute learner base grew more than tenfold in a single year.
The infrastructure story is moving fastest. In June, Egypt's National Telecommunications Regulatory Authority licensed Hassan Allam Data Centers to build and operate facilities in the country, with Hassan Allam Utilities and investment firm A15 agreeing to develop a new data center — an estimated $400 million commitment. That license effectively opens Egypt's market to onshore, at-scale accelerated compute for the first time.
Africa is home to roughly 18% of the world's population but less than 1% of global data center capacity, which has forced African developers to lean on cloud compute hosted outside the continent for large-scale training. The four new AI factories are the first serious attempt to reset that ratio. They give local developers and enterprises somewhere to train, fine-tune and deploy models without shipping data across borders.
Key facts
- 01Four AI factories have been announced or brought online across Africa within a year, with 656 megawatts of additional capacity in the pipeline.
- 02Nexus Core Systems is building a $1.2 billion, 500MW AI factory outside Casablanca on Nvidia Blackwell with renewable power from TAQA Morocco.
- 03Cassava Technologies' AI factory rollout across South Africa, Egypt, Kenya, Nigeria and Morocco could reach $720 million in investment.
- 04Nvidia has trained over 85,000 African developers toward a 100,000 target set in 2024, with Nigeria now its second-largest DLI market globally.
- 05McKinsey projects African data center demand rising from 0.4GW today to 1.5–2.2GW by 2030, requiring $10–20 billion in construction.
Cassava Technologies, Africa's first Nvidia Cloud Partner, anchors much of the buildout. Its AI factory in South Africa is already live, with planned deployments across Egypt, Kenya, Nigeria and Morocco in an investment that could reach $720 million. In Egypt, Cassava and Vodafone Egypt are launching a GPU-as-a-service AI factory aimed at businesses and government organizations that need locally hosted infrastructure with in-country data residency.
South African neocloud Stratos Lab is partnering with ECOBLOX and Digital Parks Africa to launch an AI cloud powered by more than 50 Nvidia HGX B300 systems, delivering over 7 exaflops of performance. The pitch is lower latency and reduced cost for training, inference and agentic workloads run locally rather than routed to Europe or the US.
The largest single commitment is in Morocco. At GITEX Africa in Marrakech in April, Nexus Core Systems announced a collaboration with Morocco's Ministry of Digital Transition, its Ministry of Investment and the investment agency AMDIE to build the Nexus AI factory outside Casablanca — a $1.2 billion initial investment with 500 megawatts planned, Nvidia Blackwell accelerated computing and renewable power sourced from TAQA Morocco.
The demand side backs up the spend. Nvidia set a target in 2024 of training 100,000 African developers through its Deep Learning Institute within three years, and has already trained more than 85,000, putting the continent among its fastest-growing developer regions. Nigeria is now the institute's second-largest market globally — a data point that reorders assumptions about where the next tranche of AI engineers will come from.
Egyptian startups in the Nvidia Inception program span the industry mix Nvidia wants to point to: Aidera in enterprise intelligence, Intella in Arabic speech, Marses Robotics in industrial automation, Proteinea in AI-driven drug design, Stakpak in autonomous DevOps agents, and Paymob and Thndr in financial services. The Nvidia VC Alliance has added A15 and M Empire in Egypt to link local investors into its global startup network, alongside programming with RiseUp, Plug and Play, Flat6Labs and Algebra Ventures.
The continent's language diversity is a distinct opportunity. Africa accounts for roughly a third of the world's languages, most of them invisible to frontier models trained largely on English and Mandarin corpora. Lelapa AI has built InkubaLM and the Vulavula speech service for South African languages including isiZulu and Sesotho — the kind of work that only gets economically viable when compute sits on the same continent as the data. InstaDeep, which joined Nvidia Inception in 2017 as a small Tunis team, was acquired by BioNTech for about $680 million, offering a proof point for African AI exits.
The gap between ambition and delivery is still large. McKinsey pegs African data center demand at roughly 0.4 gigawatts today, rising to between 1.5 and 2.2 gigawatts by 2030 — a jump that would require between $10 billion and $20 billion in construction. Power, grid reliability and financing all remain constraints, and 656 megawatts in the pipeline is a fraction of what 2030 demand will need. Whether Cassava, Nexus, Stratos and their peers can execute on their announced numbers on schedule is the open question.
For Nvidia, Africa is a hedge and a market-expansion play at once. Every megawatt of Blackwell shipped to Casablanca, Johannesburg or Cairo is a megawatt that doesn't depend on already-saturated US and European buildouts, and every developer trained through the Deep Learning Institute is a future customer locked into the CUDA stack rather than a cheaper alternative. The competitive frame here is not just AMD or Intel — it's whether Chinese vendors move to fill African infrastructure demand first. On current pace, Nvidia is moving to make sure they don't.
Working on something we should cover, or seeing a story we missed? Send leads, documents, or feedback to hello@aichatdaily.com. For sensitive tips, see our secure tips page for Signal and PGP options.
Spotted an error? Email hello@aichatdaily.com with the URL and the issue, or read our full corrections policy.
