Nigeria to get two new AI-ready data centres in a $1 billion investment

INFRAGORA plans to build two AI-ready data centres in Nigeria. They would support artificial intelligence, cloud services, enterprise computing, and disaster recovery. The project is designed as wholesale infrastructure, serving operators, businesses, government agencies, smaller companies, and international cloud providers.

The facilities are in the final stages of detailed assessment and design. The US Trade and Development Agency is funding a feasibility study covering technical and financial issues, including whether the sites can support AI workloads. INFRAGORA also plans an independent power plant alongside the centres.

Construction could start in the second quarter of 2027, but only after financial close. The company says the project has attracted commitments and support from regional and global institutional investors, yet it has not disclosed specific funding secured. The timeline therefore depends on completing design, financing, and other development steps.

INFRAGORA estimates the Nigerian investment at between $650 million and $1 billion. That would make it one of Nigeria’s largest planned data-centre investments. The project is also the proposed flagship of a wider AFRIDATA platform that could eventually cost $3 billion to more than $5 billion across Africa.

The Nigerian facilities are expected to begin with 80–100MW of data-centre capacity. A separate independent power plant would provide 160–200MW using gas, renewable energy, and battery storage. The extra electricity could potentially be sold to infrastructure providers, energy merchants, or the national grid.

MW describes power capacity, not the number of computer chips or AI models the centres can run. INFRAGORA has not disclosed GPU capacity, storage specifications, or network targets. The project’s eventual computing performance will therefore depend on equipment and design details that remain unpublished.

A data centre is a building designed to run computer servers and related equipment continuously. It provides computing, data storage, networking, power backup, security, and cooling. Businesses and public agencies use these facilities instead of operating all their digital systems on local machines.

AI-ready infrastructure needs specialised computing because AI applications process large datasets and require powerful chips, including GPUs. Those chips consume substantial electricity and produce heat, so strong cooling is essential. Fast networks move data between users, servers, and storage. Reliable storage keeps models, business records, and other information available.

The article says AI workloads require consistent power and cooling, but it does not disclose INFRAGORA’s GPU capacity, storage specifications, or network targets. Its planned Tier III+/IV standards are intended to provide redundancy and resilience. Tier IV can withstand individual equipment failures without interrupting service.

Data centres cannot provide dependable digital services without continuous electricity. Servers, storage systems, networking equipment, cooling, and backup systems all consume power. AI workloads intensify this demand because they use powerful processors and generate substantial heat. If electricity is interrupted or becomes too expensive, service reliability and customer prices can suffer.

INFRAGORA plans to address this challenge by developing a 160–200MW independent power plant beside the data centres. It would combine gas and renewable energy with battery storage. Ocha Energy, the company’s sister energy platform, would source electricity, while long-term power purchase agreements would help manage costs.

The plant would initially supply the centres’ planned 80–100MW capacity. Excess power could go to other providers, energy merchants, or the national grid where feasible. However, INFRAGORA has not disclosed electricity prices, the renewable share, or the plant’s development timeline, leaving important commercial questions open.

Hosting more computing infrastructure in Nigeria could improve how quickly businesses, agencies, and cloud users access digital services. Nearby servers can reduce latency, meaning data travels a shorter distance before an application responds. Local facilities could also expand access to cloud computing, AI processing, enterprise software, and disaster recovery.

For example, a Nigerian financial-services company could keep sensitive workloads in a local facility while using public cloud resources for changing demand. Government agencies and small businesses could use the same wholesale infrastructure. International cloud providers could also connect to the facilities rather than relying entirely on capacity outside Africa.

The opportunity is significant because Africa accounts for less than 2% of global data-centre capacity. INFRAGORA says 55% of African enterprises now run at least half their workloads in the cloud. Benefits will depend on reliable power, competitive prices, strong networks, and successful project delivery.

Data sovereignty means keeping important data under a country’s local control, including where it is stored and processed. Nigerian facilities could help organisations follow a sovereign-first approach by placing sensitive information locally. That may make oversight and access arrangements clearer, while reducing the need to send every workload to infrastructure abroad.

The article gives a practical model: African firms can keep sensitive data in local data centres while using public clouds for elastic workloads. A Nigerian bank or government agency might therefore store core records locally but obtain extra computing capacity from a public cloud when demand rises.

Local facilities would not end dependence on foreign technology. The project is intended to create opportunities for American suppliers of AI software, cloud infrastructure, chips, networking equipment, and cybersecurity tools. Local storage may improve sovereignty, but privacy still depends on governance, security, contracts, and responsible access controls.

Cloud computing lets organisations use remote computing, storage, and software through networks instead of buying every system themselves. A data centre groups these resources in one facility. Users request capacity when needed, and the servers process their workloads. They pay for access to shared infrastructure rather than managing all hardware directly.

AI adds a heavier processing layer. Specialised chips analyse large datasets to train or run models. Networks carry user requests and source data to the servers, while storage holds datasets, models, and results. Electricity powers every step. Cooling removes the heat produced by processors, allowing them to keep working reliably.

The article connects this chain to growing demand for AI, cloud services, and enterprise data processing. INFRAGORA plans wholesale infrastructure for these users, but it has not published GPU, storage, or network targets. Its proposed power plant and resilient design aim to support continuous operation.