The burgeoning demand for artificial intelligence (AI) workloads presents a dual challenge and opportunity for Nigeria’s data centres. As AI capabilities expand, so too does their energy appetite. However, AI also offers sophisticated solutions for optimising energy consumption and accelerating decarbonisation efforts. This dynamic interplay requires data centre operators in Nigeria to not only manage the growing power needs but also to strategically integrate renewable energy sources into their operations.
The global market for generative AI is on a trajectory for significant growth, with projections indicating it could reach US$1.3 trillion by 2032. Furthermore, AI’s potential to boost the global economy is substantial, estimated by PwC to contribute up to US$15.7 trillion by 2030, positioning it as a pivotal driver of productivity this decade. In Nigeria, where the digital economy is a cornerstone of national development, the increasing adoption of AI technologies will inevitably lead to a considerable portion of the installed data centre capacity being dedicated to these advanced workloads.
Schneider Electric highlights that the relationship between AI and energy is no longer a simple one-way street. Data centres are tasked with providing the robust power infrastructure necessary to support these intensive AI computations. Simultaneously, AI itself possesses the inherent capability to optimise energy usage patterns and significantly advance decarbonisation goals within these facilities.
The Power Demands of AI Workloads
AI training racks, in particular, are known for their substantial energy requirements, often drawing between 100 and 140 kilowatts each. This creates high-density, often unpredictable, power loads that can strain existing infrastructure.
In Nigeria, where consistent and reliable grid supply can be a challenge, a straightforward increase in raw power capacity is not a sustainable or viable solution. Instead, data centres must embrace intelligent energy management strategies. The implementation of predictive algorithms is now enabling operators to forecast potential energy spikes and adjust operations dynamically. This ability to smooth out load variability is crucial for protecting and stabilising the national grid. Furthermore, smart scheduling allows for energy-intensive AI tasks to be strategically executed during periods when renewable energy sources are abundant or during off-peak hours, thereby reducing operational costs and mitigating strain on the grid.
Addressing Cooling Challenges with Advanced Technology
Cooling systems have historically been among the most energy-intensive components of data centre operations. With the advent of high-density AI workloads, the criticality and energy demands of these cooling systems have intensified. To meet these evolving needs, Schneider Electric has introduced its Liquid Cooling Portfolio, developed in collaboration with Motivair. This portfolio features advanced cooling engineering specifically designed for AI-driven environments. The closed-loop systems within this technology minimise water consumption, a vital consideration for operational continuity in regions that may face resource constraints.
Ajibola Akindele, Country President of Schneider Electric Anglophone Africa, emphasised the importance of building for AI in Nigeria with a strong focus on resilience and scalability. He stated, “By integrating liquid cooling and collaborating on reference designs, we can ensure that local data centres are not only ready for the next generation of processors but are also contributing to a greener, more efficient ecosystem.”
Enhancing Resilience and Sustainability
Strengthening the resilience of data centres can be achieved by integrating the main grid supply with on-site renewable generation and battery storage solutions. Schneider Electric’s “Looming Power Crunch” report underscores the critical need for close collaboration between data centres and utility providers to effectively address grid constraints. This collaboration can facilitate the alignment of new data centre developments with microgrid strategies, bridging the gap between sustainability objectives and operational reliability.
Reference designs that have been co-engineered by Schneider Electric and NVIDIA showcase the significant advantages of a unified architecture. In such systems, power, cooling, and digital management functions operate in seamless harmony. The combination of liquid cooling and advanced power management capabilities can effectively support racks operating at power levels up to 142 kilowatts, all while maintaining optimal energy efficiency. Schneider Electric’s White Paper 212, titled “Bending the Energy Curve,” further illustrates that even modest improvements in Power Usage Effectiveness (PUE) can collectively lead to a substantial reduction in the overall energy growth of the data centre industry, potentially by as much as 17 per cent.
As Nigeria continues to expand its digital infrastructure, AI-ready data centres are poised to play a pivotal role as partners to the national grid. They will contribute to balancing demand through flexible and adaptive operational strategies. The future of digital infrastructure will be defined by a circular economy approach, prioritising efficiency, resource reuse, and adaptability. This paradigm shift will ensure that the energy required to power AI is managed intelligently, and that AI itself becomes a tool for optimising energy consumption.
Schneider Electric affirms its commitment to empowering AI-ready data centres to achieve “sustainable intelligence at scale” through ongoing innovation and co-engineering efforts. This represents the vanguard of digital infrastructure development, where technology and energy evolve in tandem, driven by intelligence and a commitment to sustainability.
