All Content from Business Insider 13小时前
数据中心扩张引发担忧:资源消耗与社区影响
index_new5.html
../../../zaker_core/zaker_tpl_static/wap/tpl_guoji1.html

 

随着人工智能技术的飞速发展,美国各地的数据中心建设呈现爆炸式增长。然而,这种扩张也带来了显著的环境和社会问题。这些庞大的数据中心不仅消耗巨量的电力和水资源,堪比小型城市,其选址常常靠近居民区,导致噪音、振动等问题困扰当地居民。此外,数据中心对水资源的过度需求,尤其是在干旱地区,加剧了水资源短缺的担忧。尽管科技公司承诺采取措施,如优化冷却系统和实现水资源再生,但数据中心的快速扩张无疑对社区生活质量和区域资源可持续性构成了挑战。

📈 **AI驱动的数据中心快速扩张:** 随着人工智能技术的蓬勃发展,美国主要科技公司正以前所未有的速度建设数据中心,以满足日益增长的计算需求。根据Business Insider的调查,截至2024年底,美国已有超过1200个数据中心建成或获批建设,数量是2010年的四倍,这表明AI基础设施建设正成为科技投资的重点。

💧 **巨额资源消耗引发担忧:** 大型数据中心不仅需要消耗大量电力,其日均用水量可达数百万加仑,堪比一个小型城市。这种巨大的资源需求,尤其是在水资源紧张的地区,引发了对环境可持续性的严重担忧。例如,在亚利桑那州等干旱地区,数据中心对水资源的依赖加剧了水资源短缺的风险,尽管企业承诺实现水资源再生,但实际影响仍需密切关注。

🏘️ **社区生活质量受影响:** 数据中心常常选址在靠近居民区的区域,以便获取电力和水源。这导致了噪音、振动以及对周边环境的潜在影响。例如,弗吉尼亚州的一些居民反映,数据中心产生的噪音和振动严重影响了他们的生活质量,即使采取了隔音措施,问题依然存在。此外,数据中心对当地住房成本、环境以及生活品质的长期影响也成为社区居民关注的焦点。

⚖️ **经济效益与环境代价的权衡:** 虽然数据中心承诺为当地带来新的税收收入,但其对社区环境和社会带来的负担不容忽视。例如,乔治亚州和怀俄明州也出现了类似数据中心对社区影响的案例。如何在推动AI技术发展的同时,平衡经济效益与环境保护、居民生活质量,是亟待解决的难题,需要更周全的规划和监管。

🌍 **数据中心选址与水资源压力:** 调查发现,美国约40%的数据中心位于水资源压力最大的地区。科技公司如微软、亚马逊等,其用水量在近年来显著增加。虽然这些公司已制定了到2030年实现“水正效益”(Water Positive)或归还更多水资源的承诺,但在水资源本就紧张的地区,这些设施的扩张仍然是一个严峻的挑战,可能加剧区域水资源冲突。

Data centers sitting across the street from residential housing is not an uncommong scene in Virginia.

Build, baby, build. That's the mantra behind the AI boom sweeping America.

This year, alone, Amazon, Meta, Microsoft, and Google are projected to spend about $320 billion in capex, mostly for AI infrastructure, according to an analysis of financial statements by Business Insider.

At the heart of this AI infrastructure growth are data centers that house the specialized hardware and high-speed networking equipment, driving the intensive computations behind large language models. However, AI needs more.

Because AI learns by processing increasingly large amounts of data, improving it requires more computational power, which in turn necessitates more data centers.

A BI investigation found 1,240 data centers across America are already built or approved for construction by the end of 2024.
BI reporters, editors, and designers generated this map from months of extensive research.

That's four times more than in 2010. This is the most comprehensive map of data centers in the US, to date.

BI reporters and editors generated this map by pulling obscure state records, reviewing corporate disclosures, and consulting government researchers, engineering specialists, and researchers who study state and local economic incentive packages.

Watch the documentary below on the dark side of data centers. Keep reading to see satellite images of how data centers are changing the American landscape.
Northern Virginia is one of the most densely populated parts of the DC metro area and the most concentrated area of data centers in the US.
Loudon County in Northern Virginia is the most concentrated area of data centers in the US. Yellow circles indicate built or permitted data centers.

The largest data centers can require as much electricity as a small city and up to several million gallons of water a day. That's why they often appear in residential areas where these resources are readily available.

While these data centers promise to bring new tax revenue to Virginia residents, BI's Dakin Campbell reports that some worry about how the centers will affect housing costs, the environment, and their quality of life.

This neighborhood in Prince William County, Virginia, has seen multiple Amazon data centers crop up in its backyard in recent years.
Overhead view of Carlos Yanes' neighborhood in proximity to Amazon data centers.

Carlos Yanes lives within 1,200 feet (370 meters) of several Amazon data centers and said he spent almost $20,000 to replace his windows in an attempt to muffle the sound and vibrations from the facilities.

Amazon responded to local complaints by replacing its exhaust fans with taller versions, which did muffle the sound. However, Carlos and his neighbors can still feel the vibrations through their walls, BI's Robert Leslie reports.

An Amazon spokesperson told BI that it's "operating well below the ordinance levels that we are required to operate under."

Another neighborhood in the same county was surrounded by nature for decades.

In 2021, construction on Google's data centers began near Donna Gallant's home. By 2025, multiple facilities were operational.

Another plot of land near Gallant's neighborhood, intended for housing, was rezoned in 2023 to accommodate more data centers. Gallant, along with other locals, challenged the rezoning, but the lawsuit was dismissed in 2024 and again this year.

It's not just in Virginia. Data centers are appearing across the US.
Satellite images show part of Stanton Springs, Georgia in 2017 and 2023.

These data centers in Stanton Springs, Georgia, belong to Meta, the parent company of Facebook.

These data centers in Cheyenne, Wyoming, belong to Microsoft.
Satellite images show part of Cheyenne, Wyoming in 2020 and 2024.

Cheyenne is the data center hub of Wyoming. The largest data center in the state, belonging to Microsoft, will use up to 84,725 megawatt-hours a year, according to BI estimates.

Utility customers in at least 41 states are already seeing an increase in their electric and natural gas bills, or can expect to see an increase in 2026. This is due to a variety of factors, including power-hungry data centers, BI's Ellen Thomas reports.

Microsoft data centers in Goodyear, Arizona, are approved to use up to 3 million gallons of water a day.

Arizona state lawmakers have extended tax incentives to companies through 2033, BI's Campbell reports, encouraging more construction.

The growing presence of data centers in drought-prone regions of the country, such as Arizona, is raising concerns.

Historically, data centers use ample amounts of water to cool their computer chips.
Map of water scarcity regions in the US. Light green indicates areas of "high" water stress. Dark green is "extreme high" water stress.

BI's investigation found that 40% of the planned or existing data centers are in the nation's most water-stressed areas.

Microsoft consumed nearly 2.1 billion gallons of water in 2023, a 22% increase from 2022, according to its sustainability report. It has pledged to be water positive by 2030, as has Amazon.

Meta and Google have also made similar pledges.

Meta, with more than a quarter of its built or permitted data centers in high water-stressed areas per BI's tally, said it will return more water to the environment than it consumes by 2030.

Google pledges to replenish 120% of the water it consumes by 2030.

Read the original article on Business Insider

Fish AI Reader

Fish AI Reader

AI辅助创作,多种专业模板,深度分析,高质量内容生成。从观点提取到深度思考,FishAI为您提供全方位的创作支持。新版本引入自定义参数,让您的创作更加个性化和精准。

FishAI

FishAI

鱼阅,AI 时代的下一个智能信息助手,助你摆脱信息焦虑

联系邮箱 441953276@qq.com

相关标签

数据中心 人工智能 AI 资源消耗 社区影响 环境 水资源 Data Centers Artificial Intelligence AI Resource Consumption Community Impact Environment Water Resources
相关文章