AI数据中心或推高美国电力行业排放
据彭博获得的塞拉俱乐部分析，计划中的美国 AI 数据中心可能在2025年至2030年间增加6.18亿吨电力行业碳排放，增幅达42%，并可能抵消电力行业15年的减排成果。该估算比该组织一年半前的预测高出2至4倍，但罗迪姆集团的最高排放情景认为，美国电力行业2040年的排放水平仍可能比2025年低24%。最终结果取决于项目实际建设数量、供电来源以及可再生能源部署。
塞拉俱乐部的估算凸显 AI 数据中心能源需求和排放的潜在规模，而罗迪姆集团的对比情景显示，可再生能源和项目建设的不确定性会显著影响预测结果。
Serving tech enthusiasts for over 25 years. TechSpot means tech analysis and advice you can trust.
In a nutshell: Data centers have already drawn widespread resistance over their impact on electricity prices and water supplies, and estimates of how their energy use might affect carbon emissions have recently grown. Still, projections vary widely, owing to uncertainty over renewables and over how many data centers will actually be built in the next several years.
In a report obtained by Bloomberg, the environmental nonprofit Sierra Club estimates that planned US AI data centers would add 618 million metric tons to the power sector's carbon emissions between 2025 and 2030, a 42% increase. The figure would effectively reverse 15 years of greenhouse gas reductions from electricity generation.
Adding that much CO2 would bring the sector's 2030 emissions to 2.1 billion metric tons. The latest estimate of the AI buildout's effect is between two and four times what the group would have projected a year and a half ago.
#Sierra Club#Rhodium Group#Sightline Climate#Amazon#Environmental Defense Fund