Data centre renaissance
coming to Africa

“Data Centres are cool again,” Cisco senior vice president Tom Gillis declared during the opening session at the Cisco Live conference in Amsterdam on Tuesday. And it was not a pun about the massive cooling efforts required.

It is also no coincidence that his view carried an echo of South Africa’s state of the nation address last week, when President Cyril Ramaphosa declared that these vast facilities would be the next big driver of growth in South Africa.

A few weeks earlier, the United States government had announced plans for $500-billion investment in data centre infrastructure. Some of the biggest technology companies in the world lined up to endorse the strategy.

Much of this investment and growth is a result of the rapid acceleration of artificial intelligence, and the expectation that every business would be obliged to embrace AI in the coming years. However, the parallel growth of AI use and data centre roll-out poses massive dangers to privacy and security.

These issues coalesced at Cisco Live Europe, attended by a record 18,000 delegates, in keynotes, sessions and talks that revolved as much around AI as data centres.

Cisco EVP and chief product officer Jeetu Patel said in his keynote address that AI would create “a whole new area of risk.”

“We are now at a point where, in 2025, it’s considered to be pretty likely that we will have artificial general intelligence become a reality, which means that the cognitive capabilities of AI actually match those of humans. The composition of our workforce is changing, and so is the composition of our infrastructure.

“AI will require data centres that can scale faster than ever before. But this brings a whole new area of risk. Security at the infrastructure level has to be rethought entirely. Attackers will exploit AI-driven models, and if we don’t embed security at the architecture level, we will be left vulnerable.”

Speaking to Gadget after they keynote, Patel reinforced the scale of transformation AI will demand. “AI workloads don’t just mean more compute power – they mean a rethink of network design, cooling solutions, and power management. Every company is looking to harness AI, but they’re underestimating the infrastructure challenge that comes with it.”

Data centres are the backbone of AI, cloud computing, and large-scale enterprise applications. Yet, as AI adoption grows, the infrastructure to support it must evolve at an unprecedented pace.

Cisco president for Europe Middle East and Africa (EMEA) Oliver Tuszik told Gadget: “Every company is becoming a digital company, and that means every company is becoming an AI company. That transformation is going to be driven by data centre capacity, but we have to ask: do we have enough, and are they built the right way?”

In response, Gillis said: “AI isn’t just driving demand for more data centres – it’s changing the very design of these centres. The classic cloud computing model is being redefined, and hyperscale data centres have to evolve to keep up with AI’s processing demands. The networking backbone needs to be rebuilt, and that’s the challenge we’re facing right now.

“If you look at GPU-based workloads today, they are dramatically outpacing traditional CPU workloads. This means data centres need to handle exponentially higher power loads, which is why we are seeing innovations in AI-ready infrastructure like liquid cooling and advanced silicon.”

One of the biggest obstacles is energy supply.

“AI-ready data centres consume 10 times more power than traditional ones,” Gillis said. “That’s why we are investing in liquid cooling and AI-driven power optimisation solutions that reduce consumption at scale. It’s not just about efficiency – it’s about survival.”

Chintan Patel, CTO of Cisco EMEA, agreed: “Data centres are back in vogue. We’re seeing an incredible increase in both interest and investment at an organisational and national level. Data is now a strategic asset, and having control over its processing and storage is a competitive differentiator for businesses and countries alike.”

He outlined some of the critical challenges facing next-generation data centres: “The platform you put into your data centre has to be ready not only for the workloads of today but for what’s coming. AI is just the beginning – scalability, security, and energy efficiency will define the leaders in this space. Space, power, cooling, and even cabling are all major considerations, as are supply chain constraints for infrastructure components.

“AI workloads require a rethink of data centre power supply strategies. Can we leverage solar and alternative energy sources to sustain growth? And how do we optimise operations to ensure these massive compute clusters are running efficiently? These are the questions shaping investment decisions today.”

Mary de Wysocki, Cisco chief sustainability officer, said in an exclusive interview: “We cannot expand AI infrastructure without considering the energy footprint. Our goal should be to decarbonise data centres while increasing computational capacity. That’s why we’re investing heavily in liquid cooling, renewable-powered data centres, and circular economy approaches.”

She stressed that Africa has an opportunity to lead in green data centre infrastructure.

“With the vast solar and wind energy potential on the continent, Africa can build a next-generation data centre economy that is sustainable from the start. Unlike older markets, where retrofitting is necessary, Africa can leapfrog directly to renewable-powered AI infrastructure.”

Historically, businesses and government agencies on the continent have relied on data centres located in Europe and the US, leading to high latency and performance limitations. However, the rapid expansion of local data centres in South Africa, Kenya, Nigeria, and Egypt is changing that dynamic. With Microsoft, Amazon Web Services (AWS), Google, Oracle and Huawei all investing heavily in African infrastructure, along with regional players like Teraco, African Data Centres and BCX, there is a new race to establish regional AI-ready data centres.

De Wysocki reinforced the importance of partnerships in sustainable AI data centre rollouts: “Public and private sector collaboration is critical. Governments must create incentives for green data centre investment, while enterprises must commit to sustainable best practices. AI needs to be supported by responsible infrastructure.”

Beyond energy concerns, the security challenges of AI-driven data centres are as urgent, said Cisco VP of product, AI software and platform, DJ Sampath.

“We’re seeing attackers leverage AI itself to create more sophisticated threats,” he said. “We’re moving into an era where AI is attacking AI. If your data centre isn’t secured at the infrastructure level, your entire AI stack is vulnerable.”

To counter this, Cisco has launched AI-powered security frameworks that integrate directly into networking and cloud systems. Said Jeetu Patel: “The future of security is AI-driven, real-time, and self-healing. If we continue to rely on traditional security models, we will not be able to keep up with the speed at which AI-driven threats evolve.

“We have to assume AI-driven cyber threats will be the new normal. Data centre security cannot be an afterthought – it must be embedded at the design level. That means zero-trust architectures, continuous monitoring, and automated response to anomalies.”

As South Africa moves towards becoming a major data centre hub, the interplay between AI, security, sustainability, and power availability will define its digital economy. President Ramaphosa’s vision of data centres driving South Africa’s growth was seen in Amsterdam as not mere political rhetoric – but as an economic necessity.

Patel warned: “Every company is now a digital company. Every company will become an AI company. The infrastructure to support that transition has to be built now. If we get it right, the future of work – and the future of Africa’s digital economy – will be unstoppable.”