**Arm CEO警告：美国AI数据中心电力消耗或将飙升至总电网的20%至25%**
Arm公司的首席执行官雷内·哈斯近日发出警示，随着人工智能（AI）技术的不断发展，如果不提高能效，AI数据中心的电力消耗将成为一个严峻的环境挑战。据哈斯透露，目前美国AI数据中心的耗电量在全美电网中占比仅为4%，但到2030年，这一比例可能激增到20%至25%。
哈斯特别提到了ChatGPT等大语言模型对电力的大量需求，指出这类流行应用在推动人工智能进步的同时，也加剧了能源消耗问题。随着全球对AI技术的依赖日益增强，数据中心的能耗问题已成为业界和政策制定者关注的焦点。哈斯的警告为未来数据中心的可持续发展敲响了警钟，暗示需要在技术创新和能源效率之间找到平衡。
这一预测引发关注，因为高能耗不仅会增加碳排放，推高运营成本，还可能对全球电力供应构成压力。因此，AI行业和数据中心运营商将面临如何在保持技术进步的同时，降低能耗并实现绿色转型的紧迫任务。随着绿色能源和节能技术的发展，行业将需要寻找更高效的数据处理解决方案，以应对AI带来的能源挑战。
英语如下：
News Title: “ARM CEO warns: US AI data center power consumption could soar to 25% by 2030”
Keywords: AI energy consumption rise, data center warning, ChatGPT impact
News Content:
**ARM CEO warns that US AI data centers’ power usage could surge to 20-25% of the grid**
ARM’s CEO, René Haas, has recently issued a cautionary note, stating that without increased energy efficiency, the power consumption of artificial intelligence (AI) data centers could pose a significant environmental challenge. Haas revealed that these centers currently account for only 4% of the US power grid, but by 2030, this figure might escalate to 20% to 25%.
Haas specifically mentioned the high power demands of large language models like ChatGPT, highlighting that while such popular applications drive AI advancements, they also contribute to escalating energy consumption. As global reliance on AI technologies grows, the energy consumption of data centers has become a focal point for both the industry and policymakers. Haas’ warning underscores the need for a balance between technological innovation and energy efficiency in the sustainable development of future data centers.
This projection raises concerns due to the potential increase in carbon emissions, higher operational costs, and strain on global power supply. Consequently, the AI industry and data center operators face the pressing task of reducing energy consumption while maintaining technological progress and transitioning towards greener operations. With the development of green energy and energy-saving technologies, the sector will need to seek more efficient data processing solutions to address the energy challenges posed by AI.
【来源】https://www.cls.cn/detail/1642509
Views: 2