商务英语新闻 61 – 数据中心

商业英语新闻61期：数据中心

Business English News 61, Data Centers.

当人工智能仅限于牙刷或汽车等简单功能时，它看起来相对普通。

AI seemed relatively when it was limited to simple functions in your toothbrush or car.

但到了2022年，ChatGPT引起了巨大轰动，标志着生成式人工智能的商业化开端。

But then, in 2022, ChatGPT made an enormous splash, signaling the commercialization of generative AI.

从那时起，各种新工具如雨后春笋般迅速涌现。

Since then, there's been a rapid proliferation of new tools.

这些工具的流行促使了庞大的基础设施建设，以支持如此巨大的计算需求，正如《福布斯》报道的那样。

And the popularity of these tools has necessitated massive infrastructure to support all this computing power, as Forbes reports.

数据中心正在美国各地迅速崛起，超大规模云服务商正投入数十亿美元，在全国范围内建设数千个设施。

Data centers are rapidly sprouting across America, with hyperscalers spending billions to build thousands of facilities nationwide.

尤其是农村地区，由于企业寻求更便宜的土地和丰厚的地方税收优惠，发展势头尤为迅猛。

Rural areas in particular have seen a surge in development as companies seek cheaper land and generous local tax incentives.

许多人认为，这一趋势为传统城市中心以外的社区提供了进入快速增长的人工智能领域的契机，特别是那些希望寻找新的税收来源和高薪工作的地区。

Many see the trend as providing an entry point into the growing AI sector for communities outside traditional urban centers, especially those looking for new sources of tax revenue and high paying jobs.

就业的承诺可能看起来难以抗拒，尤其是在经济充满不确定性的背景下。

The promise of jobs may seem irresistible, particularly when there's so much economic uncertainty.

但弗吉尼亚州的研究表明，每个数据中心每投入1300万美元仅创造了一个就业岗位。

But research in Virginia has shown that data centers created just one job for every $13,000,000 of investment.

对于一个消耗惊人电量的行业来说，这种回报率相当低。

That's a pretty weak return for an industry that consumes a shocking amount of energy.

根据皮尤研究中心的概述：根据国际能源署的估计，2024年美国数据中心消耗了183太瓦时的电力。

As Pew Research outlines: US data centers consumed 183 terawatt hours of electricity in 2024, according to IEA estimates.

这相当于去年全国总用电量的4%以上，大致相当于整个巴基斯坦的年用电需求。

That works out to more than 4% of the country's total electricity consumption last year, and is roughly equivalent to the annual electricity demand of the entire nation of Pakistan.

一个典型的以AI为重点的超大规模数据中心，每年消耗的电量相当于10万户家庭的用电量。

A typical AI focused hyperscaler annually consumes as much electricity as 100,000 households.

目前正在建设的更大规模的数据中心，预计耗电量将是其20倍。

The larger ones, currently under construction, are expected to use 20 times as much.

能源消耗本身还伴随着环境影响。

Energy usage itself is accompanied by environmental impacts.

但随着电网负荷不断增加，还会产生连锁影响，例如新建发电厂和输电线路的扩展。

But with an increasing load on the grid, there are knock on effects, such as new power plants and transmission line expansion.

农村电网往往更依赖化石燃料，这对关心气候变化的人来说令人担忧。

Rural electrical grids also tend to be fossil fuel heavy, which is alarming for anyone who cares about climate change.

此外，根据CBS的报道，这一代新型数据中心的碳足迹并非唯一的问题，AI数据中心还需要大量水资源来冷却硬件。

And the carbon footprint isn't the only concern with this new generation of data centers, as CBS explains: Aside from their electricity needs, AI data centers also require a great deal of water to cool the hardware.

麻省理工学院的研究表明，这可能会加剧市政供水压力并破坏当地生态系统。

This can strain municipal water supplies and disrupt local ecosystems, research from MIT shows.

目前，还没有任何法规要求企业披露其AI工具消耗的能源或水量。

For now, there are no regulations that require corporations to disclose how much energy or water their AI tools consume.

到目前为止，超大规模运营商仍能找到足够多愿意为新建数据中心提供优惠的社区，但越来越多的地区开始视其为与魔鬼做交易。

The hyperscalers have so far found enough communities willing to provide breaks to new data centers, but an increasing number of regions see it as a deal with the devil.

随着抵制情绪的增长，这种反对跨越了政治光谱的两极。

As resistance grows, it encompasses both sides of the political divide.

如今，似乎每个人都能找到一些理由来讨厌数据中心。

Today, it seems that everyone can find something to hate about data centers.

据NPR报道：一些人不希望巨大的工业设施及其伴随的噪音和光污染改变他们社区的面貌。

As NPR reports: Some people don't want huge industrial facilities, and all the noise and light pollution that comes with them, changing the character of their community.

但人们也担心数据中心的冷却系统会耗尽当地的水资源，推高电费，并且如果这些设施依赖化石燃料发电厂供电，还会加剧气候变化。

But people are also concerned about data centers depleting local water supplies for their cooling systems, driving up electricity bills and worsening climate change if the facilities rely on fossil fuel power plants for the electricity they need.

面对如此强烈的批评，科技公司急于改善公众舆论。

Faced with such fierce criticism, tech companies are desperate to shore up public opinion.

它们一直在投入资源，探索各种降低环境影响的方法。

They've been pouring resource into different ways of reducing the environmental impact.

在滨水区域，一些数据中心开发了蒸发冷却系统，而另一些则依赖耗水量更少的闭环系统。

On the waterfront, some data centers have developed evaporative cooling systems, while others rely on closed loop systems that use less water.

公司也在推行各种政策变革，正如世界经济论坛所指出的：亚马逊已将其全球运营的电力消耗100%与可再生能源发电相匹配。

Companies are also enacting various policy changes, as the World Economic Forum notes: Amazon has matched 100% of its operations' global energy consumption with electricity generated by renewable energy systems.

微软制定了供应商行为准则，要求供应商在2030年前过渡到100%无碳电力。

Microsoft established a supplier code of conduct requiring suppliers to transition to 100% carbon free electricity by 2030.

Salesforce宣布，它正在游说制定新法规，强制企业报告AI排放数据和能效标准，作为其可持续AI政策倡议的一部分。

And Salesforce announced that it is lobbying for new regulations to compel companies to report AI emissions data and efficiency standards as part of its Sustainable AI Policy initiative.

批评者认为，这些举措充其量只是绿色洗白。

Critics see much of this as greenwashing at best.

其他人则很高兴，至少公司已经开始承认这些问题。

Other people are happy that companies are at least acknowledging the issues.

无论如何，公众的讨论似乎远远落后于技术和行业的前沿。

Either way, it seems the public conversation is two steps behind the leading edge of technology and industry.

事实上，许多监管机构和公用事业公司表示，需求增长将继续超过任何效率提升。

Indeed, many regulators and utilities say that demand growth will continue to outpace any efficiency gains.

这使世界陷入了一个相当严峻的困境，正如《福布斯》所报道的。

This sets the world up for a pretty tough predicament, as Forbes reports.

人工智能的能源足迹已不再是小众问题。

The energy footprint of AI is no longer niche.

它正成为电力系统的核心关切。

It is becoming a central concern for the power system.

人工智能对电力的需求将与工业、家庭和交通日益增长的需求产生竞争。

AI's appetite for electricity will compete with increasing needs from industry, homes, and transportation.

我认为每个人都明白正确处理这个问题的紧迫性，”北美电力可靠性公司首席执行官吉姆·罗布说。

I think everybody understands the criticality of getting this right, says Jim Robb, CEO of the North American Electric Reliability Corporation.

但我们中的许多人仍被困在过去二十年的做法中，而这些做法在未来十年根本行不通。

But a lot of us are stuck in the way we've done things for the last twenty years, which simply isn't viable for the next ten.