System.Xml.XmlElement

A datacenter with 15 megawatts of IT capacity is estimated to use about 80-130 million gallons of water each year.

据估计，拥有 15 兆瓦 IT 容量的数据中心每年大约消耗 80-1.3 亿加仑的水。

The current AI boom has the world's tech giants building big data centers across the world.

当前的人工智能热潮促使世界科技巨头在世界各地建立大数据中心。

We all know they need a lot of energy.

我们都知道他们需要大量的能量。

But what about water?The two are very closely tied together.

但水呢？两者紧密相连。

This channel has a not-so-secret obsession with water.

这个频道对水有着不那么秘密的痴迷。

In this video, we look at the data center's massive water footprint.

在本视频中，我们将了解数据中心的巨大水足迹。

Some small enough to fit into a closet.

有些小到足以放入壁橱。

Or massive custom-built facilities that span entire football fields.

或者跨越整个足球场的大型定制设施。

Generally we classify them by floor space, the number of servers inside them, or how much power they consume during operations.

通常，我们根据占地面积、内部服务器数量或运行期间消耗的电量对它们进行分类。

The biggest facilities are referred to as "hyper scale" - having about 5,000 servers and being 10,000 square meters large.

最大的设施被称为“超大规模”——拥有约 5,000 台服务器，面积达 10,000 平方米。

Facilities built by global-scale operators like Google or Meta are often used for their global-scale applications - Gmail or Facebook or something like that.

由 Google 或 Meta 等全球规模运营商构建的设施通常用于其全球规模的应用程序 - Gmail 或 Facebook 或类似的东西。

But data centers can also be leased out to customers.

但数据中心也可以出租给客户。

For security, latency, and reliability reasons, cloud providers need to build these cloud data centers in a range of zones.

出于安全、延迟和可靠性的原因，云提供商需要在一系列区域中构建这些云数据中心。

These hyperscale facilities are so large because of energy efficiency scaling and economics.

由于能源效率扩展和经济性，这些超大规模设施规模如此之大。

We measure a data center's energy efficiency through power usage effectiveness or PUE.

我们通过电力使用效率或 PUE 来衡量数据中心的能源效率。

PUE is calculated by dividing the total energy delivered by the total energy going to the ICT equipment.

PUE 的计算方法是用输送的总能量除以 ICT 设备的总能量。

So the most efficient possible efficiency rating would be 1.0 - meaning that 100% of the energy going in is used on ICT.

因此，最有效的效率等级为 1.0，这意味着 100% 的能源都用于 ICT。

The larger a data center is, the lower their PUE tends to be.

数据中心越大，其 PUE 往往越低。

Google and Microsoft claim that their hyperscale data centers have a PUE of 1.2 or 1.1.

谷歌和微软声称他们的超大规模数据中心的 PUE 为 1.2 或 1.1。

A closet data center on the other hand might have 2.5 or so.

另一方面，一个封闭的数据中心可能有 2.5 个左右。

This gap is almost entirely attributable to cooling.

这种差距几乎完全归因于冷却。

Hyperscale data centers can afford more efficient cooling systems, modeling airflows through the aisles or even employ liquid cooling systems.

超大规模数据中心可以提供更高效的冷却系统，对通过通道的气流进行建模，甚至采用液体冷却系统。

So let us talk cooling.

那么我们来谈谈冷却。

Even if a data center isn't working at its full capacity - and they rarely are - it is still withdrawing 60-100% of its maximum power.

即使数据中心没有满负荷运行（这种情况很少），它仍然会消耗其最大功率的 60-100%。

That is a whole lot of heat.

那是大量的热量。

To prevent long term damage, electronic equipment have to be kept cool.

为了防止长期损坏，电子设备必须保持凉爽。

Hard disk drives need to kept at temperatures of about 45 degrees Celsius or 113 degrees Fahrenheit.

硬盘驱动器需要保持在大约 45 摄氏度或 113 华氏度的温度。

Solid-state compute and storage chips have higher limits - 85 degrees Celsius.

固态计算和存储芯片有更高的限制 - 85 摄氏度。

Maintaining that stable temperature - and humidity - is the job of the cooling system.

保持稳定的温度和湿度是冷却系统的工作。

There are two types of cooling systems - air and liquid cooling systems.

有两种类型的冷却系统 - 空气冷却系统和液体冷却系统。

Most data centers use the former as part of a raised-floor system.

大多数数据中心使用前者作为高架地板系统的一部分。

The racks are elevated about 2-4 feet above the ground.

机架高出地面约 2-4 英尺。

The underfloor area might have some cables running through it, but it is mostly cold air.

地板下区域可能有一些电缆穿过，但主要是冷空气。

Cold air comes out of the data center's Computer Room Air Conditioner or CRAC.

冷空气来自数据中心的机房空调或 CRAC。

The air is shepherded into cold aisles, sucking heat from the servers.

空气被引导到冷通道，从服务器吸收热量。

This hot air rises and is then re-collected to send to fluid heat exchangers for transfer out of the room.

这些热空气上升，然后被重新收集，发送到流体热交换器，以便传送出房间。

There are a few systems but the one most commonly used has two fluid loops - a process and condenser loop.

有几种系统，但最常用的系统有两个流体回路 - 过程回路和冷凝器回路。

In the process loop, heat is taken off the data room floor by a fluid refrigerant - usually a mix of water and glycol.

在过程回路中，数据室地板上的热量通过流体制冷剂（通常是水和乙二醇的混合物）带走。

The heat is then transferred to the water-based condenser loop for final transfer to the outside.

然后热量被传递到水基冷凝器回路，最终传递到外部。

This two-loop system is reminiscent of those for nuclear power plants and prevents contamination between the inside and outside.

这种双回路系统让人想起核电站的系统，可以防止内部和外部之间的污染。

It also grants greater operating flexibility and efficiency at the cost of being more expensive.

它还提供了更大的运营灵活性和效率，但代价是成本更高。

At the end, we have a set of cooling towers to cool the water in the final loop.

最后，我们有一组冷却塔来冷却最终循环中的水。

Inside the tower, the hot water flows down and some of it evaporates - releasing energy.

在塔内，热水流下，其中一些蒸发，释放能量。

The cooled water returns into the system for reuse.

冷却后的水返回系统重复使用。

The evaporated water leaves the tower as steam.

蒸发的水以蒸汽形式离开塔。

About 1% of the water evaporates for every 10 degrees Fahrenheit of cooling (~5.6 degrees Celsius) - though this depends on ambient temperature and humidity conditions.

每冷却 10 华氏度（约 5.6 摄氏度），就会有约 1% 的水蒸发 - 尽管这取决于环境温度和湿度条件。

How much water is used each day depends on the size of the CRAC it must support.

每天使用多少水取决于其必须支持的 CRAC 的大小。

Regardless, evaporated water must be replaced with new, make-up water.

无论如何，蒸发的水必须更换为新的补充水。

The second major way that data centers use water is an indirect one but also far larger: Energy.

数据中心用水的第二种主要方式是间接的，但规模也更大：能源。

In 2022, data centers made up about 1 to 1.3% of the world's electricity consumption.

到2022年，数据中心约占全球电力消耗的1%至1.3%。

Water is withdrawn to generate certain types of power.

抽取水来产生某些类型的电力。

Anything thermoelectric like coal, natural gas or nuclear, where we are boiling water to spin a turbine needs water.

任何热电产品，如煤炭、天然气或核能，我们用沸水来旋转涡轮机都需要水。

In 2021, 73% of the US's power came via thermoelectric means - per the US Energy Information Administration.

根据美国能源情报署的数据，到 2021 年，美国 73% 的电力来自热电方式。

That has been trending down recently as the mix has moved away from coal and towards natural gas.

随着能源结构从煤炭转向天然气，这一比例最近呈下降趋势。

Per a study by the Berkeley National Laboratory, water usage via energy generation can be as much as 2-3 times larger than what is directly consumed by the cooling systems.

根据伯克利国家实验室的一项研究，通过发电产生的用水量可能比冷却系统直接消耗的水量多 2-3 倍。

Evaporative cooling towers might consume thousands of gallons of water.

蒸发冷却塔可能消耗数千加仑的水。

But without them, data centers must resort to air conditioning systems which use even more power and thus water, indirectly.

但如果没有它们，数据中心必须求助于空调系统，这会间接消耗更多的电力和水。

Companies like Apple are working to have their data centers use more renewable energy, or run programs to offset their use.

苹果等公司正在努力让其数据中心使用更多的可再生能源，或运行计划来抵消其使用。

But for the most part, a data center pulls power from its local grid.

但在大多数情况下，数据中心从本地电网获取电力。

In 2022, Google said that they used about 4.3 billion gallons of water for cooling.

谷歌表示，2022 年他们使用了约 43 亿加仑的水进行冷却。

They say that 25% of the water they used is reclaimed wastewater, or seawater.

他们说他们使用的水有 25% 是再生废水或海水。

Digital Realty is a large real estate investment trust that specializes in data centers, with over 300 facilities around the world.

Digital Realty 是一家专注于数据中心的大型房地产投资信托公司，在全球拥有 300 多个设施。

They report their water sources with a great deal of helpful breakdowns.

他们在报告水源时提供了大量有用的详细信息。

About 36% of their water comes from municipal non-drinkable sources.

大约 36% 的水来自市政非饮用水源。

They also report that roughly half of their 2022 water consumption were in areas experiencing some form of water strain.

他们还报告称，2022 年用水量的大约一半来自出现某种形式水紧张的地区。

I laud them for reporting that stat.

我赞扬他们报告了这一统计数据。

AWS does not mention how much of the water they use comes from potable sources, but their 2022 Sustainability Report does say that 20 of their global data centers use recycled water for cooling.

AWS 没有提及他们使用的水有多少来自饮用水源，但他们的 2022 年可持续发展报告确实表示，他们的全球数据中心有 20 个使用回收水进行冷却。

Microsoft's 2022 sustainability report discusses their commitments to getting to water positive, which includes sponsoring projects that replenish more water than they consume - reaching a million people with water access.

Microsoft 的 2022 年可持续发展报告讨论了他们对实现水积极目标的承诺，其中包括赞助补充的水量超过其消耗量的项目，从而使 100 万人能够获得水。

Which is always good.

这总是好的。

Of course, this does means that for Google, Digital Realty, and probably Microsoft and AWS, the majority of the water directly withdrawn and used for cooling is drinkable water from the local water supply.

当然，这确实意味着对于谷歌、Digital Realty，或许还有微软和AWS来说，直接抽取并用于冷却的大部分水都是来自当地供水系统的饮用水。

Some 40-50% of the global population lives in areas suffering water scarcity, and many data centers need to be located near large population centers.

全球约 40-50% 的人口生活在缺水地区，许多数据中心需要位于大型人口中心附近。

Which means adding a new and competing demand for water.

这意味着增加新的、竞争性的水需求。

SemiAnalysis reports that one of the leading states in the US for data center buildouts is Arizona.

SemiAnalysis 报告称，美国数据中心建设的领先州之一是亚利桑那州。

Builders include Microsoft, PayPal, Meta, and more.

构建者包括 Microsoft、PayPal、Meta 等。

This is probably because it is so sunny - the state grew its solar capacity by 20% in 2023.

这可能是因为这里阳光明媚——该州在 2023 年将太阳能发电量增加了 20%。

Land is also cheap, the government is rather business-friendly, and there is relatively low risk of natural disasters like earthquakes.

土地也很便宜，政府对企业相当友好，地震等自然灾害的风险相对较低。

But the area also experiences drought-like conditions from time to time and is highly dependent on the Colorado River for its water.

但该地区也不时经历类似干旱的情况，并且高度依赖科罗拉多河的水源。

The concern seems to be top of mind for companies like Meta.

对于像 Meta 这样的公司来说，这个担忧似乎是首要考虑的问题。

For their part, Meta announced that they will fund projects to restore water in the Colorado River and Salt River basins and source its water using long-term storage credits so no water is taken from the municipal area.

就 Meta 而言，他们将资助科罗拉多河和盐河流域恢复水源的项目，并使用长期储存信用来获取水源，这样就不会从市政区域取水。

And it will also leverage direct free cooling to cut water usage by 60%.

它还将利用直接自然冷却将用水量减少 60%。

What's that?

那是什么？

Okay.

好的。

So is there any other way to cool down a data center without running a huge HVAC system or evaporating thousands of gallons of water each day?

那么，有没有其他方法可以在不运行大型 HVAC 系统或每天蒸发数千加仑水的情况下冷却数据中心呢？

Probably the most widely adopted water-free cooling system is "free cooling", or utilizing nature to cool things down.

最广泛采用的无水冷却系统可能是“自然冷却”，即利用自然来冷却物体。

The simplest implementation of free cooling is to leave the windows open: Direct free cooling.

自然冷却最简单的实施方式是保持窗户打开：直接自然冷却。

But outside air is rarely of high enough quality to be indoors - it has smoke, dust, gases, what have you.

但室外空气的质量很少比室内高——它有烟雾、灰尘、气体等等。

It will ruin the electronics.

它会毁掉电子设备。

So we need dehumidification, air filtration, and cleaners.

所以我们需要除湿、空气过滤和清洁剂。

So certain areas that would require the use of dehumidifiers or filters will partially defray the cost benefits of direct free cooling.

因此，某些需要使用除湿器或过滤器的区域将部分抵消直接自然冷却的成本效益。

So total energy savings are dependent on location.

因此，总能源节省取决于位置。

But generally, free cooling works.

但一般来说，自然冷却是有效的。

Studies done of data centers in various locations around Europe found that direct free-cooling cuts some energy consumption no matter where the center is.

对欧洲各地数据中心进行的研究发现，无论数据中心位于何处，直接自然冷却都可以减少一定的能源消耗。

Average savings of about 5.4-7.9%.

平均节省约5.4-7.9%。

Studies done in Australia found that direct free cooling can save up to 60% in the southern capital cities.

澳大利亚的研究发现，在南部首府城市，直接自然冷却可节省高达 60% 的费用。

A good thing since 60% of the country's energy comes from burning fossil fuels.

这是一件好事，因为该国 60% 的能源来自燃烧化石燃料。

The potentially significant advantages of a cold climate are why several Nordic governments are chasing data center foreign investment ...

寒冷气候的潜在显着优势是几个北欧国家政府追逐数据中心外国投资的原因......

and in recent years, getting it too.

近年来，也得到了它。

This when Microsoft sealed a data center into some metal Tupperware and dumped it underwater for a while.

当时微软将一个数据中心密封在一些金属特百惠中，并将其倾倒在水下一段时间。

It works, but I doubt many people will be comfortable with putting $30,000 H100s under the water.

它确实有效，但我怀疑很多人会愿意将价值 30,000 美元的 H100 放入水下。

But there is something similar that might make a lot of sense: Waterside free cooling.

但有一个类似的东西可能很有意义：水边自然冷却。

Data centers located near cold water seas can use that water to cool their systems.

位于冷水海附近的数据中心可以使用这些水来冷却其系统。

Google's Hamina data center in Finland is an example.

谷歌位于芬兰的哈米纳数据中心就是一个例子。

The data center is a converted paper mill factory that takes in fresh seawater from the ocean using the paper mill's existing pipes.

该数据中心是由造纸厂改建而成，利用造纸厂现有的管道从海洋中获取新鲜海水。

Thanks to the two-loop cooling system, the seawater does not mix with the fluid used to directly cool the datacenter.

由于采用了双回路冷却系统，海水不会与用于直接冷却数据中心的流体混合。

After cooling the process loop fluid, it is then mixed with seawater again for cooling down before being sent back out into the sea.

冷却工艺回路流体后，再次与海水混合进行冷却，然后送回大海。

Another idea would be to recover the heat and reuse it for something else: Heat recapture.

另一个想法是回收热量并将其重新用于其他用途：热量回收。

Recaptured heat can be used many ways: Desalinating water, pre-heating water in thermoelectric plants, direct power generation.

回收的热量可以通过多种方式使用：海水淡化、热电厂中的预热水、直接发电。

Or, we can simply pipe it to people's houses for general heating or to produce their hot water.

或者，我们可以简单地将其通过管道输送到人们的房屋，用于一般供暖或生产热水。

This is certainly possible.

这当然是可能的。

Captured heat from air cooled data centers comes out hot enough - about 35 to 45 degrees Celsius.

从风冷数据中心捕获的热量散发出来时温度足够高 - 大约 35 到 45 摄氏度。

And space/water heating is the single largest end use of energy in a home, accounting for about 6% of America's total energy consumption.

空间供暖/水供暖是家庭中最大的能源最终用途，约占美国总能源消耗的 6%。

The issue is that we cannot efficiently move heat as far as we can move electricity.

问题是我们无法像传输电力那样有效地传输热量。

The demand sources - i.e. the households - need to be relatively close to the data center.

需求源（即家庭）需要相对靠近数据中心。

A second issue concerns infrastructure.

第二个问题涉及基础设施。

For instance, district heating infrastructure.

例如，区域供热基础设施。

Not every city will have the network of insulated pipes needed to bring heat to homes, and few are willing to pay to have it built despite the financial benefits of doing so.

并非每个城市都会拥有为家庭提供供暖所需的隔热管道网络，而且很少有人愿意花钱建造它，尽管这样做可以带来经济效益。

Anyway, if data centers do do heat recapture - and they should, despite the payback issues - then they will likely also adopt liquid cooling solutions too.

无论如何，如果数据中心确实进行热量回收——而且他们应该这样做，尽管存在回报问题——那么他们也可能会采用液体冷却解决方案。

Liquid is far better at capturing and transferring heat than air is.

液体在捕获和传递热量方面比空气好得多。

That cuts down on overall energy consumption and also improves processor performance to boot.

这减少了整体能耗，并提高了处理器的启动性能。

There are a number of ways to do it, but it could involve directly bringing cold liquid to the chips.

有很多方法可以做到这一点，但可能涉及直接将冷液体引入芯片。

I am sure there is some Linus Tech Tips video about such a setup.

我确信有一些关于此类设置的 Linus 技术提示视频。

Finally, there is a point to be made that we do not have to cool the data center down all that much.

最后，有一点需要指出的是，我们不必对数据中心进行太多冷却。

The American Society of Heating, Refrigerating and Air-Conditioning Engineers or ASHRAE maintains a set of data center cooling standards.

美国供暖、制冷和空调工程师协会 (ASHRAE) 维护着一套数据中心冷却标准。

They recommend a temperature range between 15 to 32 degrees Celsius.

他们建议温度范围在 15 至 32 摄氏度之间。

Some operators run even cooler, assuming that electronics perform better in colder conditions.

一些操作员的运行温度甚至更低，假设电子设备在较冷的条件下性能更好。

Raising a data center's temperature by even 1-2 degrees has significant financial benefits.

即使将数据中心的温度提高 1-2 度，也会带来显着的经济效益。

So it is worth testing the boundaries.

因此值得测试边界。

Yet there is not a lot of information on this and what has been observed seems to be rather contradictory.

然而，这方面的信息并不多，而且观察到的情况似乎相当矛盾。

Google has run a few tests showing that their data centers are operable at even higher temperatures.

谷歌进行了一些测试，表明他们的数据中心可以在更高的温度下运行。

Yet at the same time, there are some Google tests which seem to show that hard drives perform worse at cooler temperatures.

但与此同时，谷歌的一些测试似乎表明硬盘驱动器在较低温度下性能较差。

So who knows?

那么谁知道呢？

One thing is for sure: The ongoing AI boom will have these massive data centers ramping up faster and using more energy.

有一点是肯定的：持续的人工智能热潮将使这些大型数据中心的发展速度更快并消耗更多的能源。

Microsoft's 2022 sustainability report mentions that their 2022 water consumption leapt 34% from 2021.

微软2022年可持续发展报告提到，他们2022年的用水量比2021年跃升了34%。

Considering there was only a 13% increase from 2020 to 2021, it appears that this jump is due to ChatGPT and other generative AI products.

考虑到 2020 年至 2021 年仅增长了 13%，这一增长似乎是由于 ChatGPT 和其他生成式 AI 产品所致。

Bob Blue - love this name - CEO of the utility Dominion Energy, serves the Northern Virginia data center cluster and said in an earnings call:

鲍勃·布鲁（Bob Blue）——喜欢这个名字——Dominion Energy 公用事业公司的首席执行官，为北弗吉尼亚州的数据中心集群提供服务，并在财报电话会议中表示：

> For some context, historically, a single data center typically had a demand of 30 megawatts or greater.

> 在某些情况下，从历史上看，单个数据中心的电力需求通常为 30 兆瓦或更高。

However, we're now receiving individual requests for demand of 60 megawatts to 90 megawatts or greater, and it hasn't stopped there.

然而，我们现在收到了 60 兆瓦至 90 兆瓦或更大需求的个人请求，而且这还没有停止。

> We get regular requests to support larger data center campuses that include multiple buildings and require total capacity ranging from 300 megawatts to as many as several gigawatts.

> 我们经常收到支持大型数据中心园区的请求，这些数据中心园区包括多个建筑物，并且需要总容量从 300 兆瓦到多达几千兆瓦不等。

SemiAnalysis did an excellent job analyzing this massive coming demand.

SemiAnalysis 出色地分析了这一巨大的未来需求。

They believe that AI will propel data center share to 4.5% of global energy generation by 2030.

他们相信，到 2030 年，人工智能将推动数据中心在全球能源发电中的份额达到 4.5%。

Water and power consumption are correlated.

水和电力消耗是相关的。

Just imagine water consumption growing that much too.

想象一下用水量也增长这么多。

Nvidia's coming GPU products like the B100 and beyond are getting even more power hungry.

Nvidia 即将推出的 B100 及更高版本的 GPU 产品变得更加耗电。

Training a leading edge model will run these GPUs to the max, which means more power and water consumption.

训练前沿模型将最大限度地运行这些 GPU，这意味着更多的电力和水消耗。

At the same time, deploying them will require immense scale which ...

与此同时，部署它们将需要巨大的规模......

again means more power and water consumption.

再次意味着更多的电力和水消耗。

So no matter how you slice it, it seems like the future of compute and AI will require far more electricity - and water.

因此，无论你如何划分，计算和人工智能的未来似乎都将需要更多的电力和水。

First at the semiconductor fabrication stage.

首先是在半导体制造阶段。

Here in Taiwan, TSMC alone is responsible for 6% of all the energy that Taiwan will consume.

在台湾，仅台积电就消耗了台湾能源消耗总量的 6%。

By 2025, that is estimated to grow to 12.5% - thanks in part to new EUV lithography machines.

到 2025 年，预计这一数字将增长至 12.5%——部分归功于新的 EUV 光刻机。

Then after the chips are fabbed, we put them into these data centers and run them to full tilt.

芯片制造出来后，我们将它们放入这些数据中心并全力运行。

They then generate heat that we need even more energy to move out of the room.

然后它们会产生热量，我们需要更多的能量才能离开房间。

If things develop as they seem to be developing, future data centers will need to rapidly adopt a combination of free-cooling, waste heat recovery, and renewable energy like solar.

如果事情像看上去的那样发展，未来的数据中心将需要迅速采用自然冷却、废热回收和太阳能等可再生能源的组合。

The faster this transition happens, the better.

这种转变发生得越快越好。

It is both a question of sustainability and money.

这既是可持续性问题，也是金钱问题。

Thanks for reading The Asianometry Newsletter!

感谢您阅读《亚洲测量学通讯》！

This post is public so feel free to share it.

这篇文章是公开的，所以请随意分享。