Workflow
水资源短缺
icon
Search documents
AI不仅费电,还费水
Hua Er Jie Jian Wen· 2025-07-16 03:24
Core Insights - The rapid development of AI technology is leading to a significant increase in water consumption due to the expansion of data centers, which poses a serious resource challenge [1][2] - Environmental experts warn that the chemical additives used in cooling systems of data centers may contaminate groundwater, leading to long-term environmental impacts [3] Water Consumption - There are currently 1,827 active data centers in the U.S., with an additional 1,726 announced and 419 under construction, all requiring millions of gallons of water for cooling [1] - Traditional large data centers can consume up to 5 million gallons of water daily, equivalent to the daily water usage of a town with a population of 20,000 to 50,000 [2] - The demand for AI-supporting facilities is projected to grow at an annual rate of 33% by 2030, further increasing water consumption [2] Pollution Risks - The use of chemical additives in cooling systems poses a risk of contaminating soil and groundwater, with persistent chemicals like PFAS being a major concern [3] - High water extraction may lower groundwater levels, threaten local wildlife, and create competition for agricultural water [3] Regulatory Gaps - Virginia, home to the largest data center market in the U.S., lacks statewide regulations on water usage for data centers, highlighting a disconnect between regulatory needs and industry growth [4] - A proposed bill aimed at requiring environmental impact assessments for new data centers was vetoed, indicating challenges in establishing regulatory frameworks [4][5] Corporate Responses - Amazon Web Services (AWS) aims to achieve "water positive" status by 2030, focusing on sustainable water sources and reporting a 53% completion rate towards this goal [6] - Google utilizes recycled or non-potable water in over 25% of its data centers, while Microsoft has developed designs that eliminate the need for water in cooling AI workloads [6]
AI 和沉迷游戏,哪个更费水? | Knock Knock 世界
声动活泼· 2025-06-14 09:00
Group 1 - BOA's closure system, which combines a knob and a fine wire, was initially inspired by designs in medical devices [2] - The founder of BOA, Garv, created a prototype ski boot with BOA's design, leading to the idea of starting a new business [3] - BOA has become a sought-after brand in the footwear industry, with notable collaborations, including with Vans, a leading skate shoe brand [2][3] Group 2 - AI development requires significant water resources for cooling data centers, with the training of models like GPT-3 consuming approximately 700,000 liters of clean water [4] - The increasing number of servers needed for powerful AI models generates heat that is thousands of times greater than a home computer, necessitating water for cooling [4] - The establishment of new AI data centers in water-scarce regions raises concerns about exacerbating global water shortages [4] Group 3 - The World Health Organization recognized gaming addiction as a disorder in 2018, with prevalence rates between 1% and 9% among various demographics [5] - Games are designed with features that can lead to addiction, as companies implement mechanisms to keep users engaged and discourage quitting [5] - There are four common characteristics across all games that contribute to their appeal, which can lead to addictive behaviors [5]
叙利亚大马士革遭遇严重供水危机
news flash· 2025-06-06 10:04
金十数据6月6日讯,叙利亚首都大马士革眼下正面临严重供水危机,饮水和日常生活用水难以保障。许 多居民担心,随着夏季来临,水资源短缺问题将进一步加剧。今年是自1956年以来,大马士革费杰水源 地降雨量最少的一年。目前,当地水务公司正在抢修部分废弃水井,以缓解供水压力。但这些举措仍难 以从根本上解决困境。大马士革预计将迎来一个炎热而艰难的夏天。 (CCTV国际时讯) 叙利亚大马士革遭遇严重供水危机 ...