中国AI热潮的中心,一个出人意料的地方

内容来源:https://www.wired.com/story/the-unlikely-place-at-the-center-of-chinas-ai-boom/
内容总结:
内蒙古乌兰察布:中国AI数据中心建设的新热土
从北京乘火车向西仅两小时,便能抵达内蒙古连绵的草原与古老的火山锥地貌。这片广袤而干旱的土地,长期以来是中国畜牧业和煤炭开采的重镇,但近年来,它悄然成为中国AI数据中心建设最火热的区域。
乌兰察布,这座拥有约150万人口的内蒙古城市,自2016年以来已有近100个数据中心项目落地或开工。高盛上周发布的研究报告显示,中国企业承诺在该市建设的项目总规划容量合计达12.5吉瓦,其中超过70%的承诺是在过去一年内宣布的,使其成为亚洲增长最快的算力集群之一。作为对比,OpenAI耗资5000亿美元的“星际之门”项目建成后总容量仅为10吉瓦。
中国企业蜂拥至乌兰察布,原因多重。该市地处内蒙古高原,海拔较高,冬季漫长寒冷,数据中心制冷能耗大幅降低。同时,其距北京较近,数据可低延迟传输至人口稠密地区。但最具吸引力的因素在于成本——内蒙古的电价几乎是全国最低,这得益于风电、太阳能发电的快速增长和丰富的煤炭资源。
值得注意的是建设主体的变化。中国AI企业首次大规模投资自建基础设施,而非租用云服务商算力。据报道,DeepSeek、字节跳动、阿里巴巴和小红书均计划在乌兰察布建设大型AI数据中心。多年来,中国AI公司在物理基础设施建设上的投入远低于美国同行,尽管已开发出多款性能优异的热门AI模型。乌兰察布数据中心的蓬勃发展,标志着它们终于开始奋起直追。
然而,一个棘手问题摆在面前:水资源短缺。乌兰察布年降水量仅约350毫米,与丹佛相当。当地政府已难以为居民用水提供充足保障,更别提众多规划中的数据中心项目尚未投产。上个月,乌兰察布当地供水公司被迫每晚关闭部分水厂七小时以缓解高峰需求。气象数据显示,在建数据中心冬季需水量较小,一年中仅有两个月需要额外用水冷却,但所有新基础设施仍可能对该地区构成严峻的环境挑战。
偏远之地的新使命
内蒙古成为数据中心热土已有至少十年历史,远早于此轮AI热潮。华为于2016年在乌兰察布建成其首个数据中心,苹果于三年后跟进。2021年,该地区被国家“东数西算”工程确定为主要枢纽之一,该工程旨在西部内陆地区布局数据中心。
但此前存在一个重大短板:由于远离人口稠密的东部沿海,这些数据中心向大多数用户传输数据时延迟较高,最初基本仅用于备份存储。人工智能的兴起改变了这一局面。新加坡管理大学研究中国算力基础设施的教授安德鲁·斯托克斯表示:“2022年AI崛起后,人们意识到这些偏远数据中心可充分利用于模型训练。”一次AI模型的训练运行可能耗时数月,且无需过多实时调整,延迟问题便不再突出。
相对而言,乌兰察布并不算遥远。内蒙古距北京和中国其他大都市区,比任何西部数据中心枢纽都要近得多。2017年和2019年建成的两条专用光纤光缆,已将平均延迟降至5毫秒以下,足以支持AI推理等实时数据交换。
斯托克斯指出,与中国其他数据中心枢纽相比,乌兰察布的增长更多由商业需求驱动,而非政府主导投资。随着DeepSeek、月之暗面、智谱AI等国内AI初创企业吸引越来越多付费用户,它们发现在乌兰察布建设推理数据中心具备可行商业逻辑——距离适中,可服务用户且不致延迟过高。
可再生能源的“消纳场”
中国政府青睐在内蒙古建设数据中心,还有一个重要考量:有助于消化国内过剩的可再生能源产能。 Carnegie China主任马晓野表示,中国可再生能源利用率最低的地区,与数据中心建设最多的地区之间存在正相关关系。
在政府看来,这是一举两得的策略——既能追赶美国的数据中心建设速度,又能同时拉动可再生能源需求。因此,中国最大风机制造商之一远景科技本月宣布将在乌兰察布建设一座2吉瓦的AI数据中心,直接接入该公司自有清洁电力供应,也就不足为奇。
但专家提醒,将中国AI数据中心与可再生能源简单视为“天作之合”过于理想化。斯托克斯研究发现,乌兰察布目前仍有约37%的电力来自煤炭。由于数据中心需全天候运行,运营商历来偏好稳定性更强的化石能源。
“内蒙古长期以来就是中国的‘西弗吉尼亚’,是一个产煤大州。”马晓野说。该地区正努力用风能和太阳能替代煤炭,但转型速度与程度尚不明朗。目前,在这片偏远高原上拔地而起的数据中心仍将在一定程度上依赖煤电。不过,马晓野表示:“三年后,或许它们就能完全由可再生能源供电了。”
中文翻译:
乘坐火车从北京向西仅需两小时,你就会发现自己置身于内蒙古起伏的草原和古老的火山渣锥之间。这片广袤干旱的土地长期以来一直是中国养羊业和煤矿开采的中心,但在过去几年里,它已成为中国建设人工智能数据中心最火热的地方。
在乌兰察布——内蒙古一座约有一百五十万人口的城市——自2016年以来,已有近一百座数据中心建成或开工建设。中国公司已承诺在该市建设总预计容量达12.5吉瓦的项目,而其中超过70%的承诺仅在过去一年内宣布,使其成为亚洲增长最快的计算集群之一,这一结论出自高盛上周发布的一份研究报告。作为对比,OpenAI耗资500亿美元的“星际之门”项目建成后总容量也仅为10吉瓦。
中国企业纷纷涌入乌兰察布有多重原因。这座城市地处内蒙古高原,海拔较高,冬季漫长寒冷,这意味着那里的数据中心不需要消耗太多能源来维持冷却。它离北京也相对较近,因此数据可以以最低延迟传输到中国人口稠密的地区。但最诱人的因素与成本有关。内蒙古的电价比中国几乎任何其他地方都便宜,这既得益于风电和太阳能的强劲增长,也得益于丰富的煤炭供应。
同样有趣的是谁在建设这些数据中心。中国人工智能公司首次在自身基础设施上大举投资,而不是向云计算公司租用算力。据报道,DeepSeek正在乌兰察布建设一座大型人工智能数据中心,字节跳动、阿里巴巴和小红书也在建设各自的设施。多年来,中国人工智能公司在物理基础设施建设上的投入远低于美国同行,尽管它们开发了多款广受欢迎且能力出众的人工智能模型。乌兰察布数据中心的热潮表明,它们终于开始迎头赶上了。
只有一个问题:找到足够的水。乌兰察布的气候干燥程度与丹佛相当,每年降水量仅约14英寸。当地政府已经在努力提供足够的水来满足居民需求——而许多规划中的数据中心项目甚至尚未投入运行。上个月,乌兰察布当地供水公司被迫每晚关闭几座自来水厂七个小时,以缓解高峰用水需求。该市正在建设的数据中心在冬季用水需求较少——乌兰察布地方政府的气象数据显示,这些设施一年中仅有两个月需要额外用水用于冷却——但所有新基础设施仍然可能对该地区构成重大的环境挑战。
偏远之地
内蒙古成为数据中心热点地区至少已有十年之久,远在当下人工智能热潮之前。华为于2016年在乌兰察布建立了其第一座数据中心,苹果在三年后也跟进落户。2021年,该地区被指定为全国性政府项目“东数西算”的主要枢纽之一,该项目旨在中国的西部腹地建设数据中心。
不过,有一个重大缺陷。由于远离中国人口稠密的东部沿海地区,这些数据中心最初在向大多数用户传输数据时面临着高延迟的问题。因此,它们最初主要被降级用于备份存储——直到人工智能赋予了它们新的用途。“随着2022年人工智能的兴起,人们意识到那些偏远的数据中心其实可以很好地用于模型训练,”新加坡管理大学研究中国计算基础设施的教授安德鲁·斯托科斯表示。一次人工智能模型的训练可能需要数月时间,且不需要太多实时调整,因此延迟问题就不再那么重要。
相对而言,乌兰察布实际上也没有那么远。内蒙古距离北京和中国其他主要大都市区,比任何西部数据中心枢纽都要近得多。而且,2017年和2019年修建的两条专用光纤光缆现已将其连接,平均延迟速度降至不到五毫秒,足以支持人工智能推理等实时数据交换。
斯托科斯表示,与中国其他数据中心枢纽相比,乌兰察布的增长似乎更多由商业需求驱动,而非政府主导的投资。随着DeepSeek、月之暗面和智谱AI等中国人工智能初创企业开始在国内吸引更多付费用户,它们发现在靠近用户的地方建设推理数据中心是可行的商业方案,既能服务用户又不会产生过高的延迟。
可再生能源驱动
中国政府对于在内蒙古建设数据中心的热情还有另一个原因:这有助于消化该国过剩的可再生能源产能。 Carnegie China——总部位于新加坡的研究机构——的主任马小军表示,中国可再生能源闲置最多的地区与数据中心建设最多的地区之间存在正相关性。
在政府看来,这是一项双赢战略。中国可以在追赶美国数据中心建设步伐的同时,增加对可再生能源的需求。因此,中国最大的风力涡轮机制造商之一远景科技本月宣布将在乌兰察布建设一座2吉瓦的人工智能数据中心,并直接接入该公司自有的清洁电力供应,这并不令人意外。
但专家警告说,将中国的人工智能数据中心与可再生能源视为天作之合的故事过于简单。斯托科斯在研究中发现,乌兰察布约37%的电力仍来自煤炭。而且由于数据中心需要全天候运行,运营商传统上更青睐化石燃料的可靠性。
“内蒙古长期以来就是中国的西弗吉尼亚。这是一个产煤大区,”马小军说。该地区目前正竞相用风能和太阳能取代煤炭,但这一转型的速度以及能走多远仍不明朗。就目前而言,在中国偏远高原上涌现的数据中心仍将在一定程度上依赖煤炭。但“三年后,也许它们将完全由可再生能源供电,”马小军说。
这是杨泽宜和路易丝·松崎《中国制造》通讯的一期。在此阅读往期通讯。
评论
返回顶部
英文来源:
Travel just two hours west of Beijing by train, and you’ll find yourself surrounded by the rolling grasslands and ancient cinder cones of Inner Mongolia. This vast, arid land has long been China’s capital of sheep farming and coal mining, but over the last few years, it has become the hottest place in the country to build an AI data center.
In Ulanqab, a city in Inner Mongolia home to about 1.5 million people, nearly 100 data centers have been opened or begun construction since 2016. Chinese companies have pledged to build projects with a combined estimated capacity of 12.5 gigawatts in the city, and over 70 percent of the total commitments have been announced in just the last year, making it one of the fastest growing compute clusters in Asia, according to a research note published by Goldman Sachs last week. For comparison, OpenAI’s $500 billion Stargate Project is set to reach only 10 gigawatts of total capacity when it’s complete.
Chinese companies are flocking to Ulanqab for a number of reasons. The city sits at high elevation on the Inner Mongolian Plateau and has long, cold winters, which means data centers there don’t need to use as much energy to stay cool. It’s also relatively close to Beijing, so data can be transmitted to China’s populous regions with minimum latency. But the most enticing factor has to do with costs. Electricity is cheaper in Inner Mongolia than almost anywhere else in China, driven by both the strong growth of wind and solar energy and an abundant supply of coal.
What’s also interesting is who is building these data centers. For the first time, Chinese AI companies are making big investments in their own infrastructure, rather than renting compute from cloud companies. DeepSeek is reportedly building a massive AI data center in Ulanqab, as are ByteDance, Alibaba, and Xiaohongshu. For years, Chinese AI companies have spent far less on building physical infrastructure than their American peers, despite developing a number of popular AI models with impressive capabilities. The Ulanqab data center boom signals that now they are finally starting to catch up.
There’s just one problem: finding enough water. Ulanqab is about as dry as Denver, getting only roughly 14 inches of rain each year. The local government is already struggling to provide enough water to meet resident demand—before many of the planned data center projects are even up and running. Last month, the local water company in Ulanqab was forced to turn off several waterworks for seven hours each night to mitigate peak demand. The data centers being built in the city will need less water in the winter—weather data from the local government of Ulanqab shows they only require additional water for cooling during two months out of the year—but all of the new infrastructure could still pose a significant environmental challenge for the region.
The Boonies
Inner Mongolia has been a data center hot spot for at least a decade, long before the current AI boom. Huawei built its first one in Ulanqab in 2016, and Apple followed suit three years later. In 2021, the area was designated as one of the main hubs of a country-wide government project dubbed “Eastern Data, Western Compute,” which aims to build data centers in the Western hinterlands of China.
There was one major drawback, though. Because they are located far from China’s populous eastern coast, these data centers initially faced high latency rates when transferring data to the majority of users. As a result, they were initially largely relegated to backup storage—until AI gave them a new purpose. “With the rise of AI in 2022, there was the realization that actually, those remote data centers could be well-utilized for model training,” says Andrew Stokols, a professor at Singapore Management University who studies China’s compute infrastructure. A training run for an AI model can take months and doesn’t require much real-time tinkering, so latency is less of an issue.
Relatively speaking, Ulanqab is also not really that far away. Inner Mongolia is much closer to Beijing and other major metropolitan areas in China than any Western data center hub is. And it’s now connected by two dedicated fiber optics cables built in 2017 and 2019 that reduced average latency speeds to less than five milliseconds, fast enough to support real-time data exchanges like AI inference.
Compared to other data center hubs in China, Ulanqab’s growth seems to have been driven more by commercial demand than government-led investment, Stokols says. As Chinese AI startups like DeepSeek, Moonshot AI, and Zhiput AI have begun attracting more paying users at home, they have found a viable business case for inference data centers close enough to serve them without excessive latency.
Powered by Renewables
The Chinese government is excited about building data centers in Inner Mongolia for another reason: It could help absorb the country’s excess capacity of renewable energy. Damien Ma, director of Carnegie China, a Singapore-based research center, says there’s a positive correlation between regions in China with the most unused renewable energy and those that have built the most data centers.
In the eyes of the government, this is a win-win strategy. China can catch up to pace of data center construction in the US, while simultaneously increasing demand for renewable energy. So it should come as no surprise that Envision, one of the largest Chinese manufacturers of wind turbines, announced this month that it will build a 2 gigawatt AI data center in Ulanqab connected directly to the company’s own clean power supply.
But experts caution that looking at AI data centers and renewable energy in China as a match made in heaven is too simple of a story. Stokols found in his research that about 37 percent of electricity in Ulanqab still comes from coal. And since data centers need to run around the clock, operators have traditionally favored fossil fuels for their reliability.
“Inner Mongolia has long been the West Virginia of China. It’s a coal country,” says Ma. The region is now racing to replace coal with wind and solar, but it remains unclear how fast that transition will happen or how far it will go. For now, the data centers popping up in China’s remote highlands will have to keep relying on coal to some extent. But “in three years, maybe it will be completely powered by renewables,” Ma says.
This is an edition of Zeyi Yang and Louise Matsakis’ Made in China newsletter. Read previous newsletters here.
Comments
Back to top