Nigeria is positioning itself to lead Africa’s next great technological leap — one powered by the hardware that fuels artificial intelligence (AI): GPUs (graphics processing units). Behind this bold ambition is Alex Tsado, a former Nvidia executive who helped deploy the first AI GPUs to global cloud providers, and who now wants to ensure Africa isn’t left behind in the AI race.
For the past six years, Tsado and his teams at Alliance4AI, a global non-profit pushing adoption of AI in Africa, and Udutech, an African AI infrastructure and implementation company, have been quietly laying the foundation for what they call Africa’s “AI backbone.” Their mission is simple: to make high-performance computing available to African innovators at a fraction of the global cost, and to build the ecosystem that ensures AI development happens on the continent, by Africans, for Africans.
“We believe Africa should not repeat its mistakes from the age of oil, when it didn’t build refineries,” Tsado told . “In the age of AI, GPUs are the refineries; they process data to unlock knowledge and automate progress.”
Building Africa’s first GPU hub
In August 2025, Udutech launched the Africa GPU Hub, a platform offering GPUs for rent at prices optimised for African markets, less than $1 an hour. Located in Lagos, the hub connects local servers with GPU clusters across Africa and beyond, giving developers access to the computational power needed to train and deploy AI models without prohibitive costs.
The model mirrors cloud computing’s early growth phase, but with a uniquely African twist. Udutech’s hub allows developers, startups, and research institutions to access AI-grade GPUs remotely, eliminating the need for expensive imports or high-maintenance physical installations.
Like telecom base stations or power grids, GPUs are capital-intensive infrastructure. A company that buys a few GPUs can create AI solutions that serve millions of users and recover costs within a year. But maintaining them is expensive. Operators spend 20–30% of their cost annually on electricity, cooling, and staffing. Udutech’s shared model spreads those costs across multiple users, enabling economies of scale.
Powering Nigeria’s AI “Primes”
Udutech’s commercial mission is to nurture at least 10 Nigerian “AI Primes” — companies built locally that can grow to serve over a million users each. To complement this, Alliance4AI, Tsado’s non-profit arm, has mobilised about $8 million worth of GPUs through philanthropic partnerships. These resources provide free or subsidised access to university students and innovation hubs across Africa, helping them build competitive AI solutions and attract investment.
The demand, Tsado said, is already overwhelming. “We’re receiving requests for more GPUs from the Nigerian AI startups we support. That’s why we’re raising a new round, to meet that demand and accelerate their growth.”
Globally, GPUs are dominated by Nvidia, the U.S. company whose chips underpin nearly every modern AI model. The barriers to entry are immense; companies like Google have spent over $5 billion trying to build alternatives. For now, Africa’s challenge is not to manufacture GPUs but to build the capacity to deploy, maintain, and eventually design them.
To that end, Udutech has partnered with Chipmango, an emerging chip-design startup training Nigerian engineers in semiconductor design. They hope that within three to five years, Udutech could buy locally designed AI chips, optimised for African environments, a major step toward self-reliance in computing infrastructure.
Nigeria’s growing AI infrastructure ecosystem
Udutech’s rise comes as Nigeria’s AI infrastructure market shows early signs of takeoff. According to Grand View Research, the Nigerian GPU-as-a-Service market generated about $18.7 million in 2024, projected to nearly double to $35 million by 2030.
Cassava Technologies, a digital infrastructure and technology company founded by billionaire Strive Masiyiwa, in partnership with Nvidia, is building Africa’s first “AI Factory”, deploying advanced GPU clusters across the continent, including Nigeria, to support data-driven applications in healthcare, agriculture, and finance.
Telecom operators are also joining the race. MTN Nigeria recently launched the Sifiso Dabengwa Data Centre, the largest Tier III facility in West Africa, designed to power AI and cloud workloads with locally priced services. Airtel Africa has signed a multi-million-dollar partnership with Xtelify to deploy AI-driven network optimisation software across its 14 markets. These moves signal a broader continental shift. The GSMA reports that operators are investing heavily in digital infrastructure to support AI-scale data loads, from 5G expansion to edge computing. Collectively, these investments are transforming Africa’s capacity to build, train, and deploy machine learning models within its own borders.
The long road ahead
For all the momentum, Nigeria still faces immense challenges. Manufacturing GPUs locally remains a distant goal; the capital, supply chain, and fabrication infrastructure required are massive and concentrated in a few global hubs. The epicentres of global GPU manufacturing are Taiwan (TSMC), South Korea (Samsung), Shenzhen/Guangdong in China (board assembly and integration), and—rapidly emerging—certain US states (AI and advanced modules).
Most Nigerian “hardware manufacturers” today are assembling or importing computers, not fabricating chips.
But Tsado believes progress starts with ownership of the ecosystem, not necessarily the chips themselves, but the right to compute.
If the oil refineries of the 20th century defined Nigeria’s industrial era, the GPU hubs of the 21st may well define its digital one.
