Summary
NVIDIA says Egypt’s AI ecosystem is expanding through new data-centre investment, startup activity and developer training. It also reports that AI factories are being announced or brought online across Africa, bringing accelerated computing closer to local developers and enterprises.
NVIDIA says Egypt’s artificial-intelligence ecosystem is moving from developer enablement towards larger-scale commercial and institutional deployment, supported by new data-centre investment, startup activity and technical training.
At an event held at the Grand Egyptian Museum, the company highlighted Egyptian startups and enterprises working in areas including healthcare, cybersecurity, robotics, financial services and Arabic-language speech technology. NVIDIA said the learner base of its Deep Learning Institute in Egypt grew more than tenfold in one year.
Egypt combines AI training with infrastructure investment
In June, Egypt’s National Telecommunications Regulatory Authority licensed Hassan Allam Data Centers to build and operate data centres in the country. Under that licence, Hassan Allam Utilities and investment firm A15 agreed to develop a new data centre with an estimated investment of $400 million, according to NVIDIA.
The company also described activity around Egypt’s startup and investment ecosystem. Members of NVIDIA’s Inception programme include Aidera, which is developing enterprise intelligence software; Intella, which is working on Arabic speech intelligence; Marses Robotics, which is developing autonomous robotics and industrial automation; Proteinea, which combines AI protein design with laboratory experiments; and Stakpak, which is building an open-source autonomous developer-operations agent.
Other companies named by NVIDIA include Paymob and Thndr, which are developing technology-enabled financial services. NVIDIA’s venture-capital alliance has also expanded in Egypt through the addition of A15 and M Empire.
A separate initiative involving Cassava Technologies and Vodafone Egypt is intended to provide businesses and government organisations with locally hosted AI infrastructure through GPU-as-a-service. GPU-as-a-service allows organisations to access accelerated computing remotely rather than purchasing and operating their own hardware. Hosting that infrastructure in-country can also place computing closer to where data is generated and processed.
AI factories are spreading across Africa
NVIDIA uses the term “AI factory” for computing infrastructure designed to support the development and operation of AI systems, including model training, fine-tuning and inference. These facilities typically combine large numbers of accelerated processors with data-centre power, cooling, storage and networking systems.
The company said four AI factories had been announced or brought online across Africa within a year, with another 656 megawatts of capacity in the pipeline. The projects described by NVIDIA span South Africa, Egypt, Nigeria, Kenya and Morocco.
Cassava Technologies, which NVIDIA identifies as Africa’s first NVIDIA Cloud Partner, is expanding accelerated-computing access through an AI factory in South Africa and planned deployments in Egypt, Kenya, Nigeria and Morocco. NVIDIA said the investment could reach $720 million. The Egypt initiative with Vodafone is intended to support locally hosted AI application development and deployment.
In South Africa, AI cloud provider Stratos Lab is partnering with ECOBLOX and Digital Parks Africa on an AI cloud powered by more than 50 NVIDIA HGX B300 systems. NVIDIA says the deployment will deliver more than 7 exaflops of performance for model training, inference and agentic-AI workloads.
In Morocco, Nexus Core Systems announced a planned AI factory outside Casablanca in collaboration with the country’s Ministry of Digital Transition, Ministry of Investment and investment agency AMDIE. NVIDIA described the project as involving an initial investment of $1.2 billion and 500 megawatts of planned capacity, using NVIDIA Blackwell accelerated computing and renewable power from TAQA Morocco.
Developer growth provides a base for adoption
NVIDIA said it set a target in 2024 of training 100,000 developers across Africa through its Deep Learning Institute within three years. The company reports that more than 85,000 developers have been trained so far, and that Nigeria is now the institute’s second-largest market globally.
The regional ecosystem also includes applications adapted to local scientific and language needs. Tunisian company InstaDeep developed drug-discovery, protein-design and logistics-optimisation applications before being acquired by BioNTech for about $680 million, according to NVIDIA. South African language-model company Lelapa AI has built InkubaLM and the Vulavula speech service for languages including isiZulu and Sesotho.
Together, the projects show the two parts of an AI infrastructure transition: locally available computing and organisations capable of using it. The figures combine operating facilities with announced and planned projects, while the investment amounts are estimates or stated project commitments rather than a single measure of deployed capacity.