Peru’s AI Data Centre Push Is Becoming a Land Assembly Story
Peru’s next major development cycle may not begin with housing towers or logistics parks. It may begin with power, water, fibre, and very large land parcels positioned just beyond Lima’s established industrial edge. In an interview with BNamericas, DMR Inmobiliaria general manager Fernando Mascaró outlined a proposed US$500 million artificial intelligence data centre project on 180 hectares near Lima, alongside a wider portfolio of eight sites being prepared for hyperscale demand.
For developers and planners, the signal is clear. Data centres are no longer only a technology infrastructure issue. They are becoming a land use, servicing, and regional growth issue. The sites that matter are not simply vacant parcels with road access. They are properties with redundant energy supply, water security, fibre connectivity, permitting pathways, and enough scale to accommodate future phases.
The most important detail is the size requirement. Mascaró noted that Lima’s Lurín district, already recognized as a new industrial area, is not suited to true AI-scale facilities because available single-owner parcels tend to be in the range of seven to 10 hectares. That may support edge facilities, colocation, or smaller data storage uses, but it does not solve the land assembly problem for hyperscale AI infrastructure.
This distinction matters for urban strategy. Industrial land markets often treat large serviced parcels as interchangeable. AI data centres expose how limited that thinking is. A hyperscale site needs continuous electricity, cooling capacity, water availability, security buffers, technical redundancy, and room for expansion. Once those filters are applied, the number of viable properties shrinks quickly, and land value begins to move toward corridors rather than conventional industrial nodes.
Mascaró described the preferred locations as roughly 50 kilometres from Lima, beyond Lurín, with overhead wholesale power trunk lines, nearby wells, river access, a desalination plant, and proximity to major generating assets. That combination turns peripheral land into strategic infrastructure land. For municipalities, it raises immediate questions about zoning, environmental approvals, water allocation, grid capacity, road access, and fiscal benefit. For private owners, it creates a new class of buyer with different feasibility metrics than warehousing, manufacturing, or residential subdivision.
The next premium land corridor will not be defined by frontage alone. It will be defined by power, water, fibre, and entitlement certainty.
The Peru case also shows how public sector demand can shape private land development. According to the BNamericas interview, the US$500 million project is linked to a private US operator and the US government, with the land prepared by an investment fund and the facility structured around long-term use. That type of arrangement changes the risk profile. It can support patient capital, but it also places heavy importance on government approvals, intergovernmental coordination, and local confidence in long-term operating conditions.
For Peru, the opportunity is to position itself as a regional AI infrastructure node before the market fully matures. Cheap energy, fibre access, Lima’s economic concentration, and available land outside the tightest industrial districts are meaningful advantages. But they are not enough on their own. Hyperscale investment will follow jurisdictions that can translate raw land into entitled, serviced, politically durable development platforms.
Developers, planners, and infrastructure investors should watch how Peru handles approvals for this first major project. If the process is clear and repeatable, the country could create a new development asset class around data centre-ready land. If it stalls, the lesson will be just as important: in the AI economy, land potential only becomes value when power, water, policy, and trust are aligned.
Source: BNamericas
