There’s a sense of anticipation in Kigali, Rwanda, leading up to the Global AI Summit on Africa, which begins today and is set to bring economic prospects and innovation.
This excitement is fueled by the recent revelation from Cassava Technologies, a technology company established by Zimbabwean telecom magnate Strive Masiyiwa, regarding the construction of Africa’s first “artificial intelligence factory” in collaboration with prominent AI chip manufacturer Nvidia.
The deployment of Nvidia’s supercomputers, which utilize graphic processing units (GPUs)—the chips commonly used to power AI—is expected to start at Cassava’s data centers in South Africa as early as June, with plans for additional rollout to the company’s other facilities in Egypt, Kenya, Morocco, and Nigeria.
Africa has lagged in the global AI surge, partly due to insufficient computing power. Research from Zindi, a network of 80,000 AI developers across 52 African nations, indicated that merely 5% of Africa’s AI experts have access to the computational resources necessary for research and innovation.
Improved access to GPUs will help “drive the AI revolution” throughout the continent, says Alex Tsado, founder and director of Alliance4AI, a nonprofit organization that assists African institutions in adopting new technologies. There are currently no significant GPU clusters in Africa, which is hindering innovators: “Without access to GPUs, it requires substantially more hours, if not days, to create the same solutions (or AI systems) compared to someone in a location where such access is readily available (like the US or Europe),” he informed CNN.
“People will begin to see more incentives to gather data because they can now process it; there will be greater motivation to train talent, thanks to the availability of GPUs… (It) triggers the ecosystem to enhance all the progress that has already been made.”
Although Cassava has not yet revealed specifics of its deployment strategy, it announced in a press release that Nvidia’s GPU-based supercomputers would facilitate quicker AI model training and that the data centers will guarantee businesses and researchers have the AI computing capabilities necessary to create, train, scale, and implement AI in a secure setting.
“Our AI factory offers the infrastructure needed for this innovation to expand, equipping African businesses, startups, and researchers with access to state-of-the-art AI infrastructure to transform their ambitious ideas into tangible breakthroughs—without needing to look beyond Africa,” Masiyiwa stated in the press release.
Democratizing AI
The high costs of hardware in relation to income levels in Africa have been a significant obstacle to AI adoption. Some of Nvidia’s most advanced GPUs can cost up to $40,000. According to AI4D, an initiative for African AI development, the price to acquire a GPU in Kenya equates to 75% of the GDP per capita, making it 31 times more expensive than in Germany, when adjusted for relative cost. This situation has led to a heavy dependence on international cloud services, including Amazon Web Services, Google Cloud, and Microsoft Azure—but these services are also costly, as transmitting data between different geographical areas increases expenses and time.
For Zindi’s network of AI developers, the establishment of local data centers featuring Nvidia GPU-powered supercomputers will be a “game-changer,” states Celina Lee, its CEO and co-founder.
“By setting up AI infrastructure based in Africa, ‘Zindians’ will benefit from more budget-friendly computing resources, quicker access to AI tools, and reduced latency, which will simplify the efficient development and deployment of AI models,” she explains.
Utilizing Nvidia’s cutting-edge technology—the same employed by OpenAI, Tesla, Meta, Google, and other tech leaders—will serve as a significant equalizer, she asserts: “This would democratize AI development, fortify local AI ecosystems, and stimulate innovation across sectors like agriculture, healthcare, and financial services, enabling more indigenous solutions suited to Africa’s distinct challenges.”
Additionally, it could help enhance Africa’s visibility in global AI training datasets. Currently, African languages and dialects are inadequately represented due to insufficient data, and there are often biases in training datasets, such as facial recognition systems that perform poorly with darker skin tones.
“With greater access to AI infrastructure, African practitioners would be able to train models with local datasets that more accurately reflect diverse demographics, economic systems, and environments,” Lee notes.
Tsado points out that there remain questions regarding how Cassava plans to tackle infrastructure challenges while constructing the factories—such as Africa’s unreliable power grids, which have discouraged such initiatives in the past. Furthermore, beyond access to GPUs, there are ongoing challenges for African end users executing AI models efficiently due to their reliance on lower-quality smartphones with limited access to internet applications, Lee adds.
Nonetheless, both agree that Cassava’s initiative represents significant progress. “This is a highly welcomed advancement indicating that Africa’s major tech is finally engaging with the African AI communities and investing in GPUs,” Tsado remarks. “Strive (Masiyiwa)’s announcement marks the first significant public commitment.”
In 2021, Google revealed a $1 billion investment in Africa, which aims to improve connectivity, foster innovation, and support local entrepreneurs. Three years later, the company asserts that this investment is effective. As per a recent report commissioned by the tech giant, in 2023, its services—including Search, Maps, and YouTube—generated $16 billion in economic impact across the continent.
Alex Okosi, the managing director for Sub-Saharan Africa and originally from Nigeria, discussed his vision for the company with CNN, highlighting transformative infrastructure initiatives such as the Equiano subsea internet cable that links Africa to Europe, as well as the newly announced Umoja cable route connecting Kenya to Australia. He also mentioned the expansion of Google Translate to include more African languages and the significance of AI.
How crucial is the African market for Google? What prompted the $1 billion investment?
Alex Okosi: The African market possesses remarkable potential. Consider this: in the upcoming four or five years, over 40% of the global youth population will be based in Africa. This indicates that the workforce of both today and the future will emerge from this continent. We view Africa not simply as a business opportunity but also as a chance for the entire world to benefit from the vast potential and talent that Africa can provide.
There are undeniable challenges that we need to address. The digital gap on this continent still exists, which is why it was essential for us to ensure that our investments in Africa are comprehensive and impactful across every significant value chain to help the continent realize its full capabilities.
What significance does investing in infrastructure hold, and how crucial are these projects for businesses and individuals in Africa?
AO: One of the primary challenges we face here is the persistent digital divide. Data costs remain high, and connectivity poses difficulties. By investing in infrastructure like Equiano and Umoja, we aim to reduce these costs and allow businesses to utilize this digital highway more efficiently.
However, the reduction in data expenses will not happen instantly. We still need to collaborate with last-mile operators to ensure they can harness this increased bandwidth, leading to lower costs for consumers. Our goal is to avoid leaving anyone behind, which requires us to lower data costs.
In what way might expanding Google Translate to include more African languages promote business communication and enhance social inclusion across the continent?
AO: The capabilities of AI have allowed us to add more languages online at a much quicker pace. I am thrilled about the introduction of 13 new languages to Google Translate, including my native Igbo. This development opens up opportunities for over 300 million speakers to access the internet in a language that feels natural to them.
From an inclusion perspective, this means a larger community will be able to engage with digital tools to advance their businesses and careers. For someone from a rural area who speaks a specific language, they can now partake in this digital transformation we frequently discuss as part of the continent’s development.
In 2018, Google established its first African AI research center in Ghana. What opportunities do you believe these AI research centers will create for the continent?
AO: Our AI research centers will provide numerous advantages for the continent, and we are already observing some of these benefits. For instance, one major issue impacting farmers in Africa is pest outbreaks. We have developed an AI solution that allows us to forecast potential devastating crop infestations a week in advance. Additionally, in the context of flooding and climate change, 23 African countries are now part of our flood forecasting hub, which provides them with information regarding when potential floods could occur, enabling people to plan accordingly.
Discuss the risks associated with AI and how its adoption could impact job trends across the nation.
AO: Certainly, there are several challenges that AI poses: there’s the risk of misinformation and cybersecurity issues if misused. As Google, we emphasize the importance of being bold and responsible. This involves collaborating with communities, academia, and government bodies to ensure we have effective frameworks for AI development that truly benefit our citizens and communities. The key focus is on implementing AI in a responsible manner while reducing the risks linked to it.
Where do you envision Africa in a decade?
AO: Looking ahead ten years, I anticipate that a significantly larger number of individuals will possess digital skills, enabling them to harness technology and AI to transform the continent, generating increased job prospects, innovation, new businesses, and economic growth. I believe AI will be instrumental in addressing many societal issues, such as food security, healthcare improvements, and enhancing productivity in the agricultural sector.
My hope is that as we progress into the future, we will witness a rise in African startups, like Moniepoint (a fintech company from Nigeria), reaching unicorn status (private companies valued over $1 billion) across the continent. I genuinely believe that this is the direction we are headed. Africa has immense potential, and I’m thrilled to be among those who are witnessing it unfold firsthand.