By Eamonn Ryan
Why ‘pay-as-you-grow’ is becoming the data centre design philosophy
(Left to right): Moderator Dion Chang, Marc Matthews, director of engineering and head of PMO at
Open Access Data Centres (OADC); Kevin Subramani, director of industry for mission critical facilities and
aviation at Atana; Wojtek Piorko, managing director for Africa at Vertiv.
All images by © RACA Journal
As organisations across Africa explore artificial intelligence (AI) deployments, one of the biggest challenges facing the data centre sector is determining how much infrastructure to build – and when. Overbuilding risks massive capital expenditure on unused capacity, while underbuilding can leave operators unable to meet rapidly growing demand.
These issues were explored in depth during a panel discussion at the seminar AI Unlocked: What ‘AI-ready’ means for 2026 decisions, featuring:
• Marc Matthews, director of engineering and head of PMO at Open Access Data Centres (OADC)
• Wojtek Piorko, managing director for Africa at Vertiv
• Kevin Subramani, director of industry for mission critical facilities and aviation at Atana
The discussion highlighted the growing importance of scalable infrastructure, flexible design and realistic planning horizons in an AI-driven data economy.
THE DANGER OF BUILDING TOO BIG
According to Matthews, one of the most common mistakes is building infrastructure far larger than the market requires. He illustrated this with a case study from Nigeria. OADC initially planned a 24MW data centre based partly on competitor activity and expectations of future demand.
“On paper the project looked fantastic. But once we started procurement and the numbers came back, our eyes started watering,” Matthews said.
When the full cost profile became clear, the company realised the project had been significantly over-capitalised. After consulting with customers and partners, OADC adopted a much different strategy. “We spoke to one of our tier-one clients and asked where they thought the Nigerian market was going. Their view was that they wouldn’t go beyond one megawatt at this stage,” Matthews explained.
The result was a shift to a ‘pay-as-you-grow’ model, in which capacity is expanded in phases as demand materialises rather than being built all at once.
DESIGNING FOR CHANGE, NOT CERTAINTY
Subramani argued that the rapid evolution of AI technologies makes traditional long-term infrastructure planning increasingly difficult. “The risk with overbuilding or underbuilding is that you’re planning for the wrong future,” he said.
AI workloads, chip technologies and compute requirements are changing at such speed that designing a facility for a fixed future state is becoming unrealistic.
Instead of attempting to ‘future-proof’ data centres, Subramani suggested designers should focus on ‘future-enabling’ them. “This means building facilities that can adapt as workloads change, as technologies evolve, and as new requirements emerge. You invest what you need now and scale later.”
This typically involves modular designs, flexible power infrastructure and cooling systems capable of supporting higher rack densities over time.
WHERE DATA CENTRES ARE OFTEN MISDESIGNED
Piorko added that early-stage designs frequently include unnecessary levels of redundancy and capacity. “Power infrastructure is often overdesigned. You’ll see requests for additional UPS systems, more generators or higher tiers of redundancy than the workload actually requires,” he said.
At the same time, other critical elements are sometimes under-designed – connectivity is one of them. “With the volume of data and compute today, connectivity becomes absolutely critical. But it’s sometimes overlooked in the early planning stages,” Piorko said.
Cooling infrastructure is another area where designs can fall short. As compute densities increase – particularly with AI workloads – traditional air-cooling systems may struggle to keep up. “Liquid cooling is coming whether we like it or not. That means there will be water in the white space, and facilities must be designed for that from the beginning,” Piorko said.
MANAGING TECHNOLOGICAL OBSOLESCENCE
Another challenge discussed by the panel was how to design data centres that remain relevant as technologies evolve. Matthews noted that major infrastructure components, such as power systems, are changing rapidly. For example, large-scale uninterruptible power supply (UPS) systems that were considered cutting-edge a decade ago may now be insufficient for modern AI workloads.
To mitigate this risk, OADC prioritises investment in long-life infrastructure such as busbars and cabling rather than installing large numbers of expensive power units upfront. “We might design the facility with space for five UPS systems but install only one initially. When the demand grows, we simply add the additional units,” he said.
This approach allows operators to avoid premature obsolescence while maintaining the ability to scale.
AI READINESS: A REALISTIC 12–18-MONTH ROADMAP
When asked what a realistic AI-readiness roadmap looks like, the panel agreed that planning horizons should remain relatively short. Subramani suggested that 12 to 18 months is the most practical timeframe. “You can’t realistically plan ten years ahead in this industry anymore. The technology and the demand profile will be completely different.”
Within that period, organisations should focus on three key areas:
• Workload planning – understanding the types of AI applications that will run on the infrastructure.
• Deployment strategy – deciding whether workloads will be hosted on-premises, in colocation facilities, or in hyperscale clouds.
• Scalability – ensuring infrastructure can evolve as demand grows.
Piorko added that companies often overlook an even more fundamental question: why they want AI in the first place.
“Everyone wants to have something AI. But the first question should be: what do you need it for? Is it revenue growth, automation, optimisation?” he said. Only once those objectives are clear should companies start designing the infrastructure required to support them.
THE HUMAN FACTOR IN AI-READY FACILITIES
While much of the discussion focused on infrastructure, Matthews emphasised that people remain just as important as technology. Operating AI-ready data centres requires new skill sets that combine traditional facilities engineering with advanced IT expertise.
“For us, one of the biggest challenges is the human interface. You need technicians who understand both critical infrastructure and IT systems,” he said.
As a result, OADC is investing heavily in recruitment and training. “AI won’t replace people in this sector. If anything, we’ll need more highly skilled people to operate these facilities,” Matthews said.
DESIGNING FOR CLIMATE RESILIENCE
The panel also explored the impact of extreme weather on data centre design. In coastal regions such as Nigeria, Matthews said operators must consider risks such as storm surges, ocean swells and heavy rainfall.
In some cases, this leads to innovative design solutions. For example, one OADC facility has no ground-level operations floor, with the first usable level elevated above the surrounding terrain to mitigate flooding risks.
Subramani added that modern data centre planning increasingly includes detailed climate modelling. “Each site is unique,” he said. “You analyse rainfall patterns, flooding risks and other climatic factors during the design stage to ensure the facility remains resilient.”
Looking ahead, the panel agreed that global data consumption will continue rising rapidly, particularly as AI adoption accelerates. However, that does not necessarily mean the world will see exponentially more data centres.
Instead, Matthews suggested that facilities will become more powerful and more efficient, handling greater workloads within smaller footprints. Edge computing may also play a role, distributing smaller data centres closer to users.
Piorko added that while infrastructure will continue expanding, future investment will increasingly focus on applications that deliver real economic value. “At the moment there is a lot of hype around AI. But eventually the investment will become more focused on where it creates real societal and economic benefit.”
For Africa’s rapidly growing digital economy, that balance – between infrastructure expansion, efficiency, and meaningful AI deployment – may define the next phase of data centre development.
