NVIDIA Corporation (NVDA. US) is backing Cassava to raise $700 million to build Africa's largest AI computing power network.

date
14:59 18/11/2025
avatar
GMT Eight
Nvidia-backed Cassava Technologies plans to raise up to $700 million to build and upgrade a batch of data centers across Africa equipped with artificial intelligence chips.
NVIDIA Corporation (NVDA.US)'s investment in Cassava Technologies aims to raise up to 7 billion dollars to build and upgrade a series of data centers across Africa equipped with artificial intelligence chips, with the goal of enabling companies, non-profit organizations, and the world's poorest populations to access AI services at lower costs. Cassava Technologies CEO Hardy Pengwah confirmed in an interview that the company, founded by Zimbabwean billionaire Strive Masiyiwa, plans to raise funds through debt and equity, but did not disclose specific financing details. The company aims to deploy a total of 12,000 NVIDIA Corporation GPUs in data centers across four regions of Africa, starting from South Africa and later expanding to North, East, and West Africa to enhance AI technology. Currently, Cassava is collaborating with the Rockefeller Foundation to provide computational resources to non-profit organizations, aiming to significantly lower the costs of agriculture, healthcare, and education services. Pengwah stated: "The goal of this partnership is for us and the Rockefeller Foundation to jointly subsidize computational resources, empowering the African artificial intelligence startup ecosystem and encouraging these startups to use local African data to develop applications that address real issues in Africa." Leading global technology companies have committed billions of dollars to build data centers to strengthen AI systems and promote widespread applications. However, Africa currently accounts for less than 1% of global data center capacity. Despite this, AI is rapidly gaining popularity on the continent, which has the fastest-growing and youngest population in the world. Due to limited access to banking and other services, African people are leveraging technologies like AI to bridge gaps. Pengwah emphasized that Cassava's initiatives aim to bring "game-changing" benefits to local organizations. The more successful startups or non-profit organizations are, the more discounts they can receive on AI technology through a points-based system. "It's a tiered discount," he said. Rockefeller Foundation President Rajiv Shah stated in a statement: "The latest advances in science and technology should benefit everyone, not just a few lucky ones. This means giving African innovators the tools they need to shape Africa's future." According to Pengwah, about 1,100 African startups in critical social sectors such as agriculture and healthcare are being considered for support. One of these is the non-profit organization Digital Green, which provides advice to self-sufficient and small-scale farmers to improve yields through identifying and treating diseases and adopting better farming practices. Its funding sources include tech companies' social responsibility departments, such as Alphabet Inc. Class C (GOOGL.US), Microsoft Corporation (MSFT.US), and Meta (META.US). Reducing costs Digital Green, established 19 years ago, initially served agricultural communities through on-site educational video broadcasts. Today, it also uses AI to answer farmers' queries on WhatsApp through text, images, or videos. The organization has successfully reduced the cost of agricultural extension services to farmers from $35 per person per year to 3.5 cents, enabling them to reach 8 million people in Africa and Southeast Asia. "Previously, farmers would receive a video every 1-2 weeks, now they ask multiple questions every week," said Digital Green founder Rikin Gandhi. Gandhi stated that if Cassava could reduce AI costs as planned, the non-profit organization could further expand its reach. "These local AI power centers are crucial for the African continent, otherwise you would have to spend more money on accessing servers and GPU clouds in North America, which would increase latency and costs," he said. "Local centers can reduce latency and save costs."