What is the projected scale of global data center capacity expansion and what are the average shell and core construction costs?
Answered at 10:19 in The Dawn of American Ingenuity: Building Smarter for a Brighter Tomorrow · Real Estate Intelligence Daily — PropTech, Finance & Commercial Market Insights
What is the projected scale of global data center capacity expansion and what are the average shell and core construction costs?
JLL forecasts nearly 100 gigawatts of global data center capacity additions between 2026 and 2030, driven by surging demand for AI infrastructure and cloud services. Shell and core construction costs are projected to average $11.3 million per megawatt in 2026, with annual modular and micro data center system sales potentially reaching $48 billion by 2030.
This expansion reflects a structural shift in real estate demand. As discussed in the episode, data centers have become mission-critical infrastructure, competing directly with traditional office and industrial properties for capital and development capacity. The scale of this buildout—100 gigawatts over five years—is not incremental; it represents a fundamental reallocation of construction resources and lending appetite toward technology infrastructure.
Construction costs matter because they determine project economics and feasibility. At $11.3 million per megawatt for shell and core, a 50-megawatt facility requires roughly $565 million in construction alone, before power systems, cooling infrastructure, and security. These capital requirements filter through lenders' underwriting standards, tightening approval criteria and raising the bar for operator credit quality and lease certainty—themes explored in depth across the broader commercial real estate market pressures covered in this podcast.
The emergence of modular and micro data center systems—forecast to generate $48 billion in annual sales by 2030—signals a diversification within the sector itself. Rather than building one monolithic facility, operators are increasingly deploying smaller, distributed nodes closer to compute demand. This reduces latency for AI workloads and spreads risk, but it also fragments the real estate footprint, requiring builders and developers to rethink site selection, zoning, and supply chain logistics.
Why these numbers matter right now
The timing is critical. Commercial real estate lending is thawing—the Mortgage Bankers Association forecasts $796 billion in commercial and multifamily lending in 2026, up 14.8% from 2025. Yet the 10-year Treasury remains above 4%, meaning capital is flowing but at a higher cost than it was five years ago. Lenders are demanding tighter documentation, verified assumptions, and defensible exit strategies. Data center developers, backed by tech companies and hyperscalers with strong credit ratings and long-term contracts, are winning a disproportionate share of that capital because their revenue models are more resilient and their tenants' needs are non-discretionary.
Full answer and detailsRead the whole answer, key facts and every other question from this episode
In short
JLL forecasts nearly 100 gigawatts of global data center capacity additions between 2026 and 2030, driven by surging demand for AI infrastructure and cloud services.
Guest · The Dawn of American Ingenuity: Building Smarter for a Brighter Tomorrow · Real Estate Intelligence Daily — PropTech, Finance & Commercial Market Insights · answer at 10:19
Key takeaways
- Shell and core construction costs are projected to average $11.3 million per megawatt in 2026, with annual modular and micro data center system sales potentially reaching $48 billion by 2030.
- This expansion reflects a structural shift in real estate demand.
- Construction costs matter because they determine project economics and feasibility.
- The emergence of modular and micro data center systems—forecast to generate $48 billion in annual sales by 2030—signals a diversification within the sector itself.
The full answer · about 1 min
JLL forecasts nearly 100 gigawatts of global data center capacity additions between 2026 and 2030, driven by surging demand for AI infrastructure and cloud services. Shell and core construction costs are projected to average $11.3 million per megawatt in 2026, with annual modular and micro data center system sales potentially reaching $48 billion by 2030.
This expansion reflects a structural shift in real estate demand. As discussed in the episode, data centers have become mission-critical infrastructure, competing directly with traditional office and industrial properties for capital and development capacity. The scale of this buildout—100 gigawatts over five years—is not incremental; it represents a fundamental reallocation of construction resources and lending appetite toward technology infrastructure.
Construction costs matter because they determine project economics and feasibility. At $11.3 million per megawatt for shell and core, a 50-megawatt facility requires roughly $565 million in construction alone, before power systems, cooling infrastructure, and security. These capital requirements filter through lenders' underwriting standards, tightening approval criteria and raising the bar for operator credit quality and lease certainty—themes explored in depth across the broader commercial real estate market pressures covered in this podcast.
The emergence of modular and micro data center systems—forecast to generate $48 billion in annual sales by 2030—signals a diversification within the sector itself. Rather than building one monolithic facility, operators are increasingly deploying smaller, distributed nodes closer to compute demand. This reduces latency for AI workloads and spreads risk, but it also fragments the real estate footprint, requiring builders and developers to rethink site selection, zoning, and supply chain logistics.
Why these numbers matter right now
The timing is critical. Commercial real estate lending is thawing—the Mortgage Bankers Association forecasts $796 billion in commercial and multifamily lending in 2026, up 14.8% from 2025. Yet the 10-year Treasury remains above 4%, meaning capital is flowing but at a higher cost than it was five years ago. Lenders are demanding tighter documentation, verified assumptions, and defensible exit strategies. Data center developers, backed by tech companies and hyperscalers with strong credit ratings and long-term contracts, are winning a disproportionate share of that capital because their revenue models are more resilient and their tenants' needs are non-discretionary.
More answers from this episode
- 1:03What does the $875 billion commercial mortgage maturity wall in 2026 represent and how is it distributed across property sectors?
- 2:03How has the shift from cheap lending to elevated Treasury rates changed what lenders prioritize in commercial real estate underwriting?
- 5:54What are the three specialized AI agents that Atom Intelligence deployed and how do they differ from traditional property analytics dashboards?
- 5:54How much US property data does Atom Intelligence cover and what architectural principles determine its trustworthiness?
- 8:48What shift has occurred in construction robotics deployment from pilot phase to production-level adoption?
- 9:48What labor and cycle time improvements do construction robotics demonstrate on specific scopes?
- 12:48What was the aggregate change in US commercial property insurance rates and how did shared versus single carrier programs compare in mid-2026?
- 13:20What was the insurance industry's underwriting gain and policyholder surplus position in the first half of 2026?
- 15:33What is the role of Faye, Juniper Square's AI oversight agent, in fund administration and what is its scope of operations?
- 17:30How much investor equity and financial statement assets does Juniper Square's platform expose to AI-powered review?