Excitement is building in Kigali, Rwanda, as the Global AI Summit on Africa kicks off today, offering fresh opportunities for economic growth and technological advancement. A major highlight is Cassava Technologies’ recent announcement about establishing Africa’s first “artificial intelligence factory” in collaboration with Nvidia, a leading AI chipmaker founded by Zimbabwean telecoms billionaire Strive Masiyiwa.
Starting in June, Nvidia’s powerful supercomputers, equipped with advanced graphics processing units (GPUs) essential for AI applications, will be installed in Cassava’s data centers in South Africa. These systems will eventually expand to facilities in Egypt, Kenya, Morocco, and Nigeria.
Africa has lagged in the global AI race largely due to limited access to computational resources. According to Zindi, a community of 80,000 AI enthusiasts from 52 African countries, only 5% of AI professionals across the continent have the necessary computing power for research and innovation.
Access to GPUs is expected to revolutionize AI development in Africa. Alex Tsado, founder of Alliance4AI, a non-profit dedicated to promoting emerging technologies across African institutions, explains that without these resources, developing AI solutions takes significantly longer compared to regions like the US or Europe. “Having access to GPUs not only speeds up development but also motivates data collection, talent training, and further innovation,” he says.
Although Cassava has not disclosed the full details of its deployment strategy, it has highlighted that Nvidia’s GPU-powered supercomputers will accelerate AI model training. The data centers will provide secure environments for businesses and researchers to build, train, scale, and deploy AI technologies effectively.
Strive Masiyiwa emphasized the transformative potential of this initiative: “Our AI factory is designed to support innovation at scale, giving African businesses, startups, and researchers access to world-class AI infrastructure. Now, they no longer have to look beyond Africa to turn their bold ideas into groundbreaking realities.”
Democratizing AI
One of the major obstacles to AI adoption in Africa has been the high cost of hardware relative to income levels. Some of Nvidia’s most powerful GPUs can cost up to $40,000. According to AI4D, an African AI development program, purchasing a GPU in Kenya can cost as much as 75% of the country’s GDP per capita—making it 31 times more expensive than in Germany when adjusted for relative income. This financial barrier has forced many African businesses and researchers to rely heavily on foreign cloud services, which are not only costly but also suffer from slower data transmission due to geographical distances.
For Zindi, a network of 80,000 AI builders across Africa, the development of local data centers powered by Nvidia’s GPU-equipped supercomputers is seen as a transformative shift. Celina Lee, CEO and co-founder of Zindi, describes it as a “game-changer.” She explains, “Having Africa-based AI infrastructure will provide more affordable computing resources, faster access to AI tools, and lower latency, making it easier to develop and deploy AI models effectively.”
The introduction of Nvidia’s advanced technology—used by global tech giants like OpenAI, Tesla, Meta, and Google—will act as a great equalizer, democratizing AI development. Lee adds, “It will strengthen local AI ecosystems, drive innovation across sectors like agriculture, healthcare, and financial services, and enable the creation of homegrown solutions tailored to Africa’s unique challenges.”
Additionally, this advancement could help improve Africa’s representation in global AI datasets. Currently, African languages, dialects, and diverse demographic data are underrepresented, leading to biases in AI systems, such as facial recognition technology that struggles with darker skin tones. “With greater access to AI infrastructure, African practitioners will be able to train models on local datasets that better reflect the continent’s diversity in demographics, economies, and environments,” says Lee.
However, there are still concerns about how Cassava Technologies will address infrastructure challenges, such as unreliable power grids that have deterred similar projects in the past. Furthermore, despite improved access to GPUs, there are ongoing difficulties for African end-users in running AI models efficiently, primarily due to the prevalence of lower-grade smartphones and limited internet access, notes Lee.
Despite these challenges, both Lee and Tsado view Cassava’s initiative as a significant step forward. “It’s a very welcome boost that Africa’s big tech companies are finally engaging with African AI communities and investing in GPUs,” says Tsado. “Strive Masiyiwa’s announcement is the first major public commitment, and it could inspire others to follow suit.”