AI News AI资讯 23h ago Updated 16h ago 更新于 16小时前 42

Microsoft’s carbon emissions went up 25 percent last year 微软去年碳排放量上升25%

Microsoft’s carbon emissions rose by 25% in 2025, reaching 34 million metric tons, primarily due to datacenter expansion and the cessation of purchasing non-additional renewable energy certificates. The company acknowledges that sustainability solutions are not scaling quickly enough to offset the resource demands driven by AI infrastructure, jeopardizing its 2030 carbon-negative goal. Industry-wide trends show similar spikes, with Google reporting a 25% increase in supply chain emissions and Am 微软2025年碳排放激增25%至3400万吨,主要受数据中心基础设施扩张及停止购买非额外可再生能源证书影响。 微软承认AI基础设施导致能源、水和材料需求激增,而可持续性解决方案的扩展速度未能跟上这一需求。 谷歌供应链排放同样飙升25%,亚马逊增加16%,凸显大型科技公司在AI驱动下的共同气候挑战。 微软重申2030年实现碳负排放的目标,但近期报告连续显示其面临达成该目标的挫折。 亚马逊披露2025年数据中心耗水量达25亿加仑,引发关于AI资源消耗规模的进一步关注。

65
Hot 热度
60
Quality 质量
55
Impact 影响力

Analysis 深度分析

TL;DR

  • Microsoft’s carbon emissions rose by 25% in 2025, reaching 34 million metric tons, primarily due to datacenter expansion and the cessation of purchasing non-additional renewable energy certificates.
  • The company acknowledges that sustainability solutions are not scaling quickly enough to offset the resource demands driven by AI infrastructure, jeopardizing its 2030 carbon-negative goal.
  • Industry-wide trends show similar spikes, with Google reporting a 25% increase in supply chain emissions and Amazon noting a 16% rise, highlighting a sectoral challenge in balancing AI growth with environmental targets.

Why It Matters

This development signals a critical bottleneck in the AI industry’s ability to sustain rapid infrastructure growth while meeting strict environmental commitments. For practitioners and stakeholders, it underscores the urgent need for scalable green technologies and transparent accounting methods, as current sustainability efforts are failing to keep pace with the energy and resource intensity of modern AI models.

Technical Details

  • Emission Metrics: Microsoft recorded 34 million metric tons of carbon emissions in 2025, a 25% year-over-year increase, attributed largely to datacenter infrastructure expansion.
  • Policy Shifts: The emission spike correlates with Microsoft’s decision to stop purchasing "non-additional, unbundled renewable energy certificates," forcing a reliance on actual grid energy or additional renewable investments.
  • Resource Consumption: AI infrastructure is identified as the primary driver for increased demand in energy, water, and land, with specific metrics showing significant water usage comparisons between major cloud providers like Amazon and Microsoft.

Industry Insight

  • Strategic Realignment: Companies must accelerate investment in direct renewable energy procurement and on-site generation rather than relying on market-based instruments like unbundled RECs to meet net-zero targets.
  • Infrastructure Planning: Future AI datacenter deployments require integrated sustainability planning from the outset, focusing on energy-efficient hardware and cooling technologies to mitigate the environmental footprint of increased compute power.
  • Transparency and Accountability: Stakeholders should expect stricter scrutiny on sustainability reports, as discrepancies between growth and environmental goals become more pronounced, potentially impacting corporate reputation and regulatory compliance.

TL;DR

  • 微软2025年碳排放激增25%至3400万吨,主要受数据中心基础设施扩张及停止购买非额外可再生能源证书影响。
  • 微软承认AI基础设施导致能源、水和材料需求激增,而可持续性解决方案的扩展速度未能跟上这一需求。
  • 谷歌供应链排放同样飙升25%,亚马逊增加16%,凸显大型科技公司在AI驱动下的共同气候挑战。
  • 微软重申2030年实现碳负排放的目标,但近期报告连续显示其面临达成该目标的挫折。
  • 亚马逊披露2025年数据中心耗水量达25亿加仑,引发关于AI资源消耗规模的进一步关注。

为什么值得看

本文揭示了AI基础设施的高速扩张与可持续发展目标之间的严重脱节,为科技行业提供了关于能源和水资源真实消耗的警示。对于AI从业者和政策制定者而言,这强调了在追求算力增长的同时,必须加速绿色技术和可持续解决方案的创新与规模化应用。

技术解析

  • 碳排放数据:微软2025年碳排放总量为3400万公吨(不含特定干预措施),较前一年增长25%。这一增长主要由数据中心基础设施的物理扩张驱动。
  • 政策变更影响:微软于去年二月停止购买“非额外、无捆绑的可再生能源证书”,这一策略调整直接导致了报告期内碳排放数据的上升,反映了公司对“额外性”标准的严格遵循。
  • 资源消耗对比:亚马逊报告2025年数据中心用水量为25亿加仑,虽声称低于微软,但具体对比数据未完全披露;谷歌则报告供应链排放增长25%,亚马逊增长16%,表明不同巨头在间接排放管理上存在差异。
  • 可持续性瓶颈:报告明确指出,尽管AI驱动了对能源、水和材料的巨大需求,但现有的可持续性解决方案(如清洁能源部署、能效提升技术)在规模化速度上存在滞后。

行业启示

  • ESG合规与透明度升级:科技公司需重新评估其碳中和路径,从简单的购买碳信用转向实质性的基础设施绿色改造和供应链深度脱碳,以应对日益严格的监管和公众 scrutiny。
  • AI能效技术成为核心竞争力:随着算力需求指数级增长,硬件能效比、液冷技术、可再生能源直接供电等绿色AI基础设施将成为企业竞争的关键壁垒和成本控制中心。
  • 水资源管理纳入战略重点:数据中心的水资源消耗问题日益凸显,企业需将水足迹管理纳入整体可持续发展战略,特别是在水资源紧张地区的数据中心选址和运营优化。

Disclaimer: The above content is generated by AI and is for reference only. 免责声明:以上内容由 AI 生成,仅供参考。

Research 科学研究 Policy 政策