As investment in digital infrastructure accelerates, cost per megawatt has become one of the most widely referenced measures of data center development economics. Yet the apparent simplicity of the metric can obscure substantial differences in project scope, technical requirements and delivery strategy.
A rapidly evolving cost environment
According to JLL’s 2026 Global Data Center Outlook, average global data center construction costs increased from approximately $7.7 million per MW in 2020 to $10.7 million per MW in 2025.
JLL forecasts a further 6% increase in 2026, taking the average to approximately $11.3 million per MW.
This growth reflects sustained demand for new capacity, increasing technical complexity, constraints on skilled labor and longer procurement lead times.
However, these figures should not be interpreted as universal development costs. Benchmarking requires a consistent definition of scope, particularly when distinguishing base-building infrastructure from tenant technology and IT equipment.
How costs compare across US markets
JLL’s 2026 research identifies the following indicative construction cost ranges for major US data center markets:
| Market | Construction cost per MW |
|---|---|
| Chicago | $12M–$14M |
| Northern Virginia | $11M–$12M |
| Dallas | $10M–$11M |
| Phoenix | $10M–$11M |
| Atlanta | $10M–$11M |
These benchmarks demonstrate meaningful regional differences, but they should be treated as reference points rather than direct estimates for a particular development.
Why the same MW capacity can produce different costs
Several factors can materially influence construction costs.
Power density and cooling strategy
The growth of AI workloads is increasing demand for high-density computing infrastructure and liquid-cooling systems.
JLL estimates that liquid-cooled facilities can carry a construction cost premium of approximately 10% compared with conventional air-cooled facilities.
Turner & Townsend has also identified a 7–10% US construction cost premium for AI data centers compared with traditional facilities of similar IT capacity.
Campus scale and phasing
Large campuses may benefit from repeatable building designs, standardized equipment and coordinated procurement.
However, scale does not automatically translate into lower unit costs. Power infrastructure, enabling works, phasing constraints and accelerated delivery schedules can offset potential economies.
Power availability
Grid connection timelines, utility upgrades and alternative power strategies increasingly influence both site selection and project viability.
For some developments, the timing and cost of securing sufficient power may have a greater impact on the investment decision than relatively modest differences in building construction costs.
Procurement and delivery strategy
Early equipment reservations, contractor capacity, design standardization and the sequencing of multiple buildings can all influence final costs.
For large programs, procurement decisions made at the campus level may create opportunities that would not be available through individual building packages.
The importance of comparing like for like
A credible cost-per-MW benchmark must establish what is included.
Key considerations include whether capacity is measured as IT load or another power measure, the extent of utility infrastructure, redundancy requirements, cooling technology, building configuration and whether costs include owner expenses or technology fit-out.
Without this consistency, comparisons between projects can produce misleading conclusions.
Beyond a single benchmark
For owners and developers managing multiple facilities, the greatest value comes from connecting cost-per-MW benchmarks with detailed project data.
Tracking changes in equipment pricing, building systems, procurement packages and regional trade costs allows teams to understand not only whether a project is above or below benchmark, but why.
The Dharam perspective
As data center programs become larger and more technically complex, effective benchmarking requires a consistent, transparent cost structure. Combining detailed cost planning with historical project intelligence and digital analytics can support stronger feasibility assessments, procurement decisions and capital allocation across multi-site programs.
