Data Center Maps: Find Out Where the Data Centers Near You Are Located

Under the AI Boom, Data Centers Are Quietly "Invading" Your Community

When Oregon resident Isabelle Reksopuro first heard that Google was consuming public land in her state to build data centers, she didn't know who to believe. "There's so much misinformation about data centers," she said. "Google denied taking over that land." She eventually pieced together the full story: the city of The Dalles, near the Washington state border, had transferred public land to Google for data center construction — legally distinct from "direct seizure," but for local residents, the outcome was no different.

The Dalles didn't become Google's target by accident. This small city sits along the Columbia River Gorge, adjacent to large hydroelectric facilities managed by the Bonneville Power Administration. Electricity prices have long remained among the lowest in the nation, at roughly 3–4 cents per kilowatt-hour — well below the national average. Google established its first major data center there as early as 2006 and has continued expanding ever since. In fact, data center site selection typically follows several core principles: low electricity costs (which account for 40%–60% of total operating expenses), cool climates (to reduce cooling energy consumption), abundant water sources (for evaporative cooling systems), low natural disaster risk, and proximity to fiber optic backbone network nodes. The Pacific Northwest nearly perfectly satisfies all of these conditions, which is the fundamental reason why Oregon and Washington have become data center hotspots.

This is not an isolated case. The AI boom is driving tech giants to expand data centers at an unprecedented pace, and these behemoths — occupying tens of thousands of square meters — are often built right in ordinary residents' "backyards."

Data Center Map Tools: Giving Information Back to the Public

For a long time, ordinary people had little way of knowing how many data centers were operating or being planned near their homes. Some organizations and researchers have begun building interactive data center maps that allow anyone to view the distribution of data centers in their area online.

Currently influential data center mapping tools include Data Center Map (datacentermap.com), Baxtel, and tracking platforms developed by various academic institutions and nonprofit organizations. These tools typically integrate multiple data sources: commercial real estate databases, building permit records, environmental impact assessment documents, utility large-customer connection applications, satellite imagery change detection, and local government land-use change announcements. Some tools even use machine learning algorithms to automatically identify data center building characteristics from satellite images — such as large rooftop cooling equipment, backup generators, and high-voltage substations. The limitations of these maps include data update delays, and projects in early planning stages often haven't yet entered public record systems. Nevertheless, they have significantly lowered the barrier for public access to information and have become essential tools for community organizations and investigative journalists.

The value of these maps extends far beyond satisfying curiosity. When residents discover that a large data center is being planned nearby, they can:

• Participate more effectively in public decision-making
• Monitor environmental impact assessment reports
• Oversee critical approval processes such as land-use changes

Information transparency is the first step in community advocacy.

The Triple Cost Behind Data Center Expansion

Public Land Quietly Transferred

Data center site selection favors regions with cheap electricity and cool climates. The Dalles, with its abundant hydroelectric resources and low electricity prices, became Google's ideal target. But once public land is transferred to a corporation, the community permanently loses that resource — land that could have been used for parks, schools, or other public facilities becomes a fenced-off server warehouse.

Energy Consumption Is Spiraling Out of Control

Training and running AI models requires massive computing power, directly driving up data center electricity consumption. Traditional data centers primarily handled cloud storage and web services with relatively stable loads, but the emergence of large language models (LLMs) has fundamentally altered the computing demand curve. For example, a single training run of a GPT-4-class model can consume tens of millions of kilowatt-hours of electricity — equivalent to the annual electricity usage of thousands of American households. The International Energy Agency (IEA) noted in its 2024 report that global data center electricity consumption has exceeded 460 terawatt-hours, accounting for approximately 1.5%–2% of total global electricity consumption, and is projected to surpass 800 terawatt-hours by 2026. Meanwhile, NVIDIA's GPU chip power consumption has been climbing with each generation — the latest B200 chip has a thermal design power of 1,000 watts per card, and a rack fully loaded with such chips can reach 120 kilowatts, more than ten times that of a traditional server rack. This leap in energy density means data centers need not only more power but entirely new cooling architectures.

This places enormous pressure on local power grids and may also undermine regional carbon reduction goals. Additionally, data center cooling systems consume vast amounts of water, making conflicts even more acute in already water-scarce regions. Data center cooling methods are broadly divided into air cooling and water cooling, with evaporative cooling being one of the most common solutions for large data centers. It works by allowing water to carry away heat as it evaporates — highly efficient but astonishingly water-intensive. By some estimates, a typical large data center can consume millions of liters of water per day, equivalent to the daily water usage of a small town. Google disclosed in its 2023 environmental report that its global data centers consumed approximately 5.6 billion gallons (about 21.2 billion liters) of water that year. In the arid western United States, competition between data centers, agricultural irrigation, and residential water use is intensifying. Residents in places like Mesa, Arizona and Loudoun County, Virginia have protested data centers competing for water resources. Some companies are beginning to explore alternatives such as liquid cooling and immersion cooling, but large-scale commercial deployment still faces challenges related to cost and technological maturity.

Tech Companies Deliberately Create Information Black Boxes

The "misinformation" problem Reksopuro mentioned is no accident. Many tech companies use shell companies or limited liability companies (LLCs) during the land acquisition phase, deliberately concealing the true buyer's identity. This strategy aims to prevent land prices from being driven up once the buyer's identity is revealed, while also avoiding premature community opposition. For example, Google has purchased land in multiple locations under names like "Sharka LLC," and Microsoft and Meta have engaged in similar practices. In the United States, LLC registration information often doesn't require disclosure of the actual controlling party, making it very difficult for community residents and local media to identify the true project driver in early stages.

Community residents often only learn what's happening after construction begins, when the window for participating in decision-making has long since closed. By the time building permit applications or environmental impact assessment notices are published, land transactions are typically already complete, and the community's bargaining leverage has been drastically reduced. In recent years, some states have begun pushing legislation requiring large infrastructure projects to disclose ultimate beneficial owner information during the land transaction phase, but progress has been slow — the tech industry's lobbying power is one of the main obstacles. This information asymmetry is one of the most underestimated issues in data center expansion.

What Communities Can Do

Facing rapid data center expansion, more and more communities are taking proactive action:

• Make good use of map tools and public records: Regularly check on local data center planning developments
• Speak up at critical junctures: Land-use change hearings and environmental assessment public comment periods are windows for residents to express their views
• Push for information disclosure: Demand that local governments require tech companies to disclose data center energy consumption, water usage, and emissions data
• Negotiate community compensation: Don't just accept empty promises of "job creation" — push for substantive infrastructure investment and tax revenue return

Regarding the last point, it's worth understanding in depth: Tech companies typically negotiate significant tax abatements or tax increment financing (TIF) incentives with local governments during site selection, promising "job creation" and "economic multiplier effects" in return. But the operational characteristics of data centers mean that direct employment numbers are extremely limited — a multi-billion-dollar hyperscale data center may require only 50–200 people for day-to-day operations. By comparison, a manufacturing plant with equivalent investment might create thousands of jobs. Loudoun County, Virginia is one of the world's most data center-dense regions, and while data centers contribute significant property tax revenue, residents widely complain about noise pollution, traffic congestion, and declining quality of life. Some communities have begun requiring Community Benefits Agreements to be included in tax incentive deals, explicitly specifying the amounts that companies must invest in local infrastructure construction, education funds, or environmental compensation to ensure that economic benefits truly reach local residents.

Computing Power Expansion Cannot Bypass Public Oversight

Data centers are the bedrock of the digital economy, and no one denies their importance. But importance doesn't mean carte blanche. The emergence of mapping tools sends a positive signal: technology can be used not only to build data centers but also to hold the process accountable.

In the AI era, computing power is productivity — but the expansion of computing power must proceed within a framework of transparency and fairness. That data center in your backyard should be able to withstand everyone's scrutiny.