About the tracker

Africa is in the middle of an AI infrastructure announcement cycle. Hyperscalers, colocation providers and "AI factory" ventures are promising hundreds of megawatts of new capacity, often in places where the grid is already strained. Announcements are easy to find. What has actually been built, how much power it draws and who owns it are much harder to pin down.

The Africa Compute Tracker is an attempt to keep that record straight. Each facility is a plain text file in a public repository. Every figure carries a source, a trust tier, a retrieval date and a confidence level. Contested assertions are kept apart from sourced facts, and the maintainer's own judgement is labelled as such.

Who maintains it

The tracker is built and maintained by Tawanda Moyo, a software engineer and technical writer in Harare working on AI safety, with a focus on compute governance.

Get in touch

• Found an error? Report a correction. Corrections are handled first.
• Know of a facility that should be here? Suggest it with a source.
• Using the data, or want to collaborate? Open a conversation or reach out via tawanda.dev.

What it is not

This is not a complete census. It covers hyperscale, large colocation and purpose-built AI facilities, and it favours fewer, well-sourced records over many thin ones. A facility's absence means it has not been recorded yet. See the methodology for scope and sourcing rules, and the changelog for what has changed.