What is the Microsoft Fairwater data center? The Microsoft Fairwater data center is a 315-acre, 1.2 million square foot AI compute facility in Mount Pleasant, Wisconsin that Microsoft CEO Satya Nadella calls "the world's most powerful AI datacenter." Going live ahead of schedule in April 2026, Fairwater connects hundreds of thousands of Nvidia GB200 and GB300 GPUs into a single seamless cluster, delivering 10 times more performance than today's fastest supercomputer. For CRE investors, this $7.3 billion Wisconsin investment signals a decisive shift in data center geography, construction scale, and infrastructure requirements. For a broader overview of AI's impact on commercial real estate, see our complete guide on AI tools for real estate investors.
Key Takeaways
- Microsoft's Fairwater data center occupies 315 acres and 1.2 million square feet across three buildings in Mount Pleasant, Wisconsin, with approved expansion plans that could bring the total campus to 9 million square feet and $13 billion in assessed value.
- The facility's $7.3 billion total investment ($3.3 billion initial plus $4 billion for a second campus) is creating a new technology hub in southeastern Wisconsin, with 3,000 peak construction jobs and 800 permanent positions.
- Fairwater uses closed-loop liquid cooling for over 90% of its compute capacity, confirming that liquid cooling is now the baseline requirement for AI data center CRE, not an optional upgrade.
- Multiple identical Fairwater data centers are under construction across the United States and Europe, signaling a replicable campus model that will drive demand for large-acreage parcels in secondary and tertiary markets.
- The facility's 250 MW solar project in Portage County illustrates the growing link between AI data centers and adjacent renewable energy CRE developments.
Inside the Fairwater Facility: CRE Fundamentals
Microsoft's Fairwater campus delivers unprecedented scale for a single AI compute facility. The three-building complex required 46.6 miles of deep foundation piles, 26.5 million pounds of structural steel, 120 miles of medium-voltage underground cable, and 72.6 miles of mechanical piping during construction. The facility's storage systems alone span the equivalent of five football fields. According to DataCenterDynamics, recent approvals for 15 additional buildings could bring the total campus to nearly 9 million square feet, with a total taxable value exceeding $13 billion.
For CRE investors evaluating data center assets, these construction specifications reveal important benchmarks:
- Land requirements: At 315 acres for a single campus, hyperscale AI data centers now require acreage comparable to large industrial distribution centers or regional shopping malls. This pushes site selection toward suburban and exurban locations with available land and power, exactly the trajectory we analyzed in our coverage of how energy constraints are reshaping CRE site selection.
- Structural intensity: The 26.5 million pounds of structural steel for 1.2 million square feet translates to approximately 22 pounds of steel per square foot, roughly double the structural intensity of a Class A office building and triple that of a standard warehouse. AI data centers are among the most material-intensive commercial buildings ever constructed.
- Underground infrastructure: The 120 miles of underground cable and 72.6 miles of mechanical piping mean that the subsurface infrastructure investment may approach or exceed the above-ground building cost. This has direct implications for site remediation, environmental assessment, and long-term asset value.
Liquid Cooling Becomes the Standard
Fairwater confirms what CRE data center investors have anticipated: liquid cooling is now the default infrastructure requirement for AI compute facilities. The facility uses a closed-loop liquid cooling system for over 90% of its compute capacity, recirculating water continuously to minimize waste while managing the extreme heat output of hundreds of thousands of Nvidia GPUs.
According to Tom's Hardware, AI data centers require 2 to 3 times more electricity than traditional facilities. The liquid cooling infrastructure at Fairwater addresses this by dramatically improving thermal efficiency compared to legacy air-cooled systems.
For CRE investors, the liquid cooling mandate creates a clear dividing line in the data center market:
- New builds: Any data center constructed for AI workloads in 2026 and beyond must include liquid cooling infrastructure from day one. This adds approximately 15% to 25% to construction costs but is non-negotiable for hyperscaler tenants.
- Retrofit opportunity: Existing air-cooled data centers face costly retrofits estimated at $60,000 to $195,000 per rack to support AI workloads, as we covered in our analysis of Nvidia's Vera Rubin cooling requirements. Facilities that cannot retrofit will be limited to non-AI workloads, creating a two-tier market.
- Water infrastructure: Closed-loop systems reduce but do not eliminate water requirements. Site selection increasingly depends on access to water sources and local regulations governing industrial water use.
Wisconsin as an Emerging Data Center Market
The choice of Mount Pleasant, Wisconsin for what Microsoft calls the world's most powerful AI data center is significant for CRE geographic strategy. Wisconsin is not traditionally ranked among the top data center markets (Northern Virginia, Dallas-Fort Worth, Phoenix, and Columbus typically dominate). Microsoft's $7.3 billion commitment signals that hyperscalers are actively expanding beyond established hubs. With 92% of corporate occupiers having initiated AI programs, the physical infrastructure needed to support these programs is driving CRE demand in unexpected markets.
Several factors make southeastern Wisconsin attractive for AI data center CRE:
- Power availability: Wisconsin's grid has available capacity that constrained markets like Northern Virginia and the Bay Area lack. The state's mix of natural gas, nuclear, and growing renewable sources provides the reliable baseload power AI data centers require.
- Land cost arbitrage: Industrial land in southeastern Wisconsin is priced at a significant discount to Northern Virginia's Loudoun County or the Dallas-Fort Worth corridor, where hyperscaler competition has driven prices sharply higher since 2023.
- Labor market: The 3,000 construction workers deployed at peak and the planned 800 permanent staff represent substantial local economic impact. Wisconsin's manufacturing heritage provides a skilled trades workforce suited to data center construction and operations. Microsoft also launched the state's first Datacenter Academy with Gateway Technical College to train over 1,000 students in the next five years.
- Proximity to Chicago: Mount Pleasant sits within the Chicago metro area's broader infrastructure network, providing access to fiber routes, talent, and supporting services without the congestion and cost of a major metro location.
CRE investors should monitor land transactions in the Mount Pleasant and Racine County area. Microsoft's commitment typically catalyzes supporting development including employee housing, retail, and hospitality properties. If you are ready to evaluate emerging data center markets for your portfolio, The AI Consulting Network specializes in identifying AI infrastructure CRE opportunities before they reach mainstream investor awareness.
The Replicable Campus Model
Perhaps the most important CRE signal from Fairwater is Microsoft President Brad Smith's confirmation that multiple identical Fairwater data centers are under construction across the United States and Europe. This replicable campus model means Microsoft has developed a standardized design that can be deployed across multiple locations, creating predictable demand for similarly specified sites. Microsoft has already committed $6.2 billion to a similar facility in Narvik, Norway, powered by hydroelectricity.
For CRE investors, the replicable model implies:
- Consistent site requirements: Each Fairwater clone will need approximately 300+ acres with access to 250 MW or more of power capacity, reliable water supply, and fiber connectivity. Investors who can assemble land parcels meeting these specifications in power-available secondary markets are positioned for hyperscaler acquisition interest.
- Construction pipeline visibility: Standardized designs accelerate construction timelines and create recurring demand for the same materials and trades across sites, benefiting CRE-adjacent construction and logistics properties.
- Geographic diversification: Microsoft is deliberately distributing AI compute capacity across the country rather than concentrating in a few markets. This diversification reduces single-market risk for data center CRE investors and creates opportunities in markets that would not traditionally attract hyperscaler investment.
The AI in real estate market is projected to reach $1.3 trillion by 2030 at a 33.9% CAGR, and AI data center infrastructure represents the single largest capital allocation driving that growth. CRE sales volume is forecast to increase 15% to 20% in 2026 (Source: CBRE Research), with data center transactions accounting for a rapidly growing share of total volume. For personalized guidance on evaluating AI data center CRE opportunities, connect with The AI Consulting Network.
Frequently Asked Questions
Q: How large is the Microsoft Fairwater data center in Wisconsin?
A: The Fairwater campus currently occupies 315 acres and includes 1.2 million square feet of space across three buildings, housing hundreds of thousands of Nvidia GB200 and GB300 GPUs. With approved expansion plans including 15 additional buildings, the total campus could reach 9 million square feet with over $13 billion in assessed value. Microsoft has committed $7.3 billion to the Wisconsin site, including a second facility of similar scale.
Q: Why did Microsoft choose Wisconsin for its most powerful AI data center?
A: Wisconsin offers available power capacity, lower land costs compared to established data center markets like Northern Virginia and Dallas, a skilled trades workforce from the state's manufacturing heritage, and proximity to Chicago's fiber and infrastructure network. These factors make it attractive for hyperscale AI facilities that require 300+ acres and hundreds of megawatts of power.
Q: What does the Fairwater data center mean for CRE investors?
A: Fairwater signals three trends for CRE investors. First, AI data centers are expanding beyond traditional hub markets into secondary locations with available power and land. Second, liquid cooling is now the baseline infrastructure requirement, creating a two-tier market between AI-capable and legacy facilities. Third, Microsoft's replicable campus model means standardized 300+ acre site requirements will drive land demand in multiple markets simultaneously.
Q: How does liquid cooling affect data center CRE valuations?
A: Liquid-cooled facilities command premium valuations because they can support AI workloads that generate 40 to 60 kW per rack, compared to 8 to 15 kW for traditional air-cooled facilities. Existing air-cooled data centers face $60,000 to $195,000 per rack retrofit costs or risk being limited to lower-value non-AI tenants. New builds with integrated liquid cooling from day one are the highest-value data center CRE assets in 2026.