MDDI 演讲稿 · 2026-03-13

杨莉明部长在 AirTrunk 区域总部开幕仪式上的主旨演讲实录

Transcript of Minister Josephine Teo's Keynote Speech at AirTrunk Regional Headquarters Opening Ceremony

Josephine Teo · 数码发展及新闻部长 · AirTrunk 区域总部开幕仪式

要点

  • 新加坡数字经济占 GDP 比重:2023 年约 17%,2024 年增至 18.6%,已与制造业齐平、超过金融服务。
  • 在「小国论坛」(Forum of Small States)的多次交流中,新加坡总结出数字增长四要点:基础设施、能力(数字 + 企业 + 劳动力 + 研究 + 治理)、全社会动员、合理的法规与国际合作。
  • 新加坡数据中心容量约 1.4 GW,并在裕廊岛预留了高达 700 MW 的成长空间,专门用于支持即将爆发的 AI 工作负载。
  • 下一阶段不是一味扩容,而是朝「更绿色」走——既看数据中心本身(能耗),也看算法(绿色计算),并通过《数字连通蓝图》打通海底电缆网络。
  • AirTrunk 在新加坡设立区域总部,是新加坡数据中心扩张期的一部分。Josephine 表示新加坡愿意把本地形成的运营标准(如热带高温下的运行、海底电缆修复时效)输出为国际标准。
  • 面向未来,新加坡力图把 AI、量子计算、研究社区(微软、Google DeepMind、Quantinuum 在新加坡设立研发能力)连成一张网,支持像 AirTrunk 这样的公司一起解决人类与地球的难题。

完整译文(中文)

MDDI 英文原文译文 · 翻译日期:2026-05-02

早安,AirTrunk 创始人兼首席执行官 Robin Khuda 先生、黑石集团(Blackstone)主席 Gautam Banerjee 先生、各位同事与朋友:

很高兴在这个美丽的地点出席 AirTrunk 区域总部的正式开幕。除了感谢各位邀请我出席这一重要时刻,我也要感谢这片好风景。

我刚到部里时,有一件事让我不太满意。当时这个部叫做「通讯及新闻部」(MCI)。

我问同事:「我们做了这么多事来发展数字经济,那我们必须先知道数字经济到底有多大吧?」因为如果连衡量都做不到,就难以判断我们的努力是否真的有所成就。

他们解释说——衡量数字经济没有国际通行的标准方法。

于是我说:「好,那等国际机构搞清楚之前,我们当然要自己先动手,设计一套属于我们的方法论。」

2023 年,我们整理出第一份《新加坡数字经济报告》。

当时基于我们扎实的方法论,我们评估数字经济约占新加坡 GDP 的 17%。

当我们发布 2024 年的更新版报告时,这一比重已增长到 18.6%。这个体量让我们几乎与新加坡制造业齐平,并超过了金融服务。

新加坡是个小国,但我们尽到自己的一份。在联合国,「小国论坛」(Forum of Small States)有约 108 个成员——人口在 1000 万以下的国家。

每次我和论坛成员见面,他们都很好奇我们是怎么发展数字经济的,想知道我们采用了什么策略。

我们认为有四件事必须做好。

第一,跟所有事一样——从基础设施起步。你必须打好数字基础设施的底子。

第二,你需要数字能力、企业能力、劳动力能力。你还需要研究社区的能力——非常重要的还有「治理能力」。在新领域里,没有治理,就创造不出有利于增长的环境。

第三,在「数字」这种横跨各领域的事情上,政府不能单打独斗。你必须动员全社会,去做数字包容与 AI 采用。所有这些情境里,你都得把所有力量调动起来,并把它们对齐到最佳状态。

第四,你必须考虑法规与立法。你需要操守与理性的态度去引领这些进展。在这个领域,必须与国际伙伴合作。

我也告诉论坛成员,我们做的事并没有什么秘密——「Singapore Digital Gateway」一直公开,我们也很坦诚地谈这些事。

他们共同提到的一个挑战,是在那些「极其困难」的地点建设数字基础设施——这正是 AirTrunk 大显身手的地方。

通过你们的工作,你们正在满足一种看起来「永不饱和」的需求。这是个适合做这门生意的好时代。当然不是只有上行机会——还有大量需要克服的问题。在许多正在建设数据中心的地方,本地社区非常担心能耗,担心它会与居民通过电网获取的电力争夺。

他们同样担心水资源消耗,担心这对自己的需求(比如农业,可能是这些社区非常重要的生命线)意味着什么。

这是一个不仅需要细致管理、也存在大量创新空间的领域。

就新加坡而言,我们看到机会,也看到挑战。

AirTrunk 与我们一同走过了 10 年——所以你们知道,从数据中心密度看,我们大约有 1.4 GW,并在裕廊岛预留了 700 MW 的成长空间。

我们已经划定了这块区域,它是我们支持未来 AI 活动起飞的一个重要构成。

但我们做的不只是扩容。我们必须思考如何更聪明地运营,所以需要一张「更绿色数据中心」的路线图——更省电,或者使用更绿的方案。

这就引出一个问题:要绿的只是数据中心吗?还是包括算法的写法?有没有办法让算法本身更绿?「绿色计算」是另一个我们想推进的方向。同时也别忘了,数据中心之间需要彼此通信——我们的《数字连通蓝图》(Digital Connectivity Blueprint)就在尝试构建一张海底电缆网络,把「连通性」真正落到实处。这就是我们看到的挑战与机会。

AirTrunk 落户新加坡,我们由衷欢迎,理由很多——尤其是我们珍视你们作为运营商、开发商、投资者的洞察。

我们希望通过比以往更深入地听取你们的看法,能开发出更好的方案——这些方案不只服务于新加坡的利益,也服务于更广泛的全球利益。

新加坡还希望以哪些方式支持 AirTrunk?我想有几个层面。在国内,你们的进入将是我们数据中心格局扩张期的一部分;你们可以借力一个非常活力、富有创新的金融服务体系。我注意到你们通过绿色贷款融到了相当可观的资金——而我们也清楚,伴随建设进程,资本胃口非常大。资金的获取并非小事,是一个必须在国际层面正面回应的问题。

我们非常希望,那些在新加坡发展出来的想法,能够成为世界上其他地方的标准。

比如,我们引入了「在热带气候下运行数据中心」的标准——能不能在比此前认为的更高一点的温度下运行?但我们不想停在这里。

我们也希望在一些重要事项上推动标准——例如海底电缆能否在合理时间内被修复。打造一个支持产业增长的环境,是新加坡非常想做的事。

第三,随着 AI 在我们经济中份额变大,能源消耗只会更高。所以这是个非小的问题:「AI 能不能更资源高效?」如果我们能在这个问题上下点功夫——不只是在落地层面,更要在研究层面,把最聪明的头脑集合起来回应这些对人类与地球而言至关重要的问题——我相信我们能做出更大的贡献。

新加坡的研究社区是我们鼓励像 AirTrunk 这样的公司充分接入的资源。

尽管微软多年前就在新加坡有业务,但他们直到去年才在这里设立研究机构。Google DeepMind 也决定在这里组建团队——而我可以告诉你们,他们要解的不是小问题,而是真正想攻克人类面临的最艰难问题。

几天前,我刚见过 Quantinuum——一家做量子计算机的公司。

他们正在建立一个平台,开发能够解决「经典计算机难以求解」问题的算法——其中很多与优化(optimisation)有关。

他们告诉我一件非常令人兴奋的事。我们有一个非常繁忙的港口,而港口优化其实非常难。

排列组合多到,如果某艘船延误,经典计算机要花几千年才能算出最优解。概念上,你可以让量子计算机先把可选方案缩小,再用经典计算机上的 AI 完成后续的优化求解。

我分享这件事,只是因为我看到——新加坡在生长出来的这张「网络」,能以多少种方式支持你们的扩张计划与作为一家公司的梦想。

我希望新加坡作为一个环境,不只给你们这片好风景,而是真正成为一个好背景——支撑你们成为你们能成为的最伟大公司。

再次祝贺。谢谢。

英文原文

MDDI 官网原始记录 · 抓取日期:2026-05-02

Good morning, Mr. Robin Khuda, Founder and CEO of AirTrunk, Mr. Gautam Banerjee, Chairman of Blackstone, Colleagues and friends.

I'm delighted to join you for the official opening of AirTrunk’s regional headquarters at this beautiful location. Apart from thanking you for inviting me to this important occasion, I would also like to thank you for the view too.

When I first came to my Ministry, there was one aspect that I was unhappy about. At the time, the Ministry was called the Ministry of Communications and Information (MCI).

I asked my colleagues: “For all the work that we do to grow the digital economy, we must first know the size of the digital economy?” Because if we can't measure it, it is hard to know whether our efforts have amounted to anything.

They explained there was no internationally standardised way of measuring the digital economy.

So, I said, “Okay, whilst international bodies figure this out, we certainly should take a stab at it and design our own methodology.”

In 2023, we put together the Singapore Digital Economy Report.

At that time, based on our robust methodology, we assessed the digital economy to be around 17% of Singapore’s GDP.

When we published the updated report for 2024, it had grown to 18.6%. This size puts us at almost on par with manufacturing in Singapore and surpasses financial services.

Singapore is a small country, but we do our part. At the United Nations, the Forum of Small States has about 108 members – comprising of countries with populations below 10 million.

Each time I meet with Forum members, they are curious about how we grew our digital economy, and they want to know what strategies we pursue.

There are four things we think we need to do well.

First, as in everything, you start with infrastructure. You need to have a good foundation in digital infrastructure.

Second, you need digital, enterprise, and workforce capabilities. You also need capabilities in the research community, and very importantly, you need governing capabilities. In new domains, without governance, you cannot create the right conditions for growth.

Third, in cross-cutting areas like digital, the government cannot do it alone. You must rally the whole of society for digital inclusion and AI adoption. In all of these instances, you have to bring the forces to bear and align them in the best way you can.

Fourth, you must consider regulations and legislation. You need ethos and a rational approach to shepherd developments. In this domain, you must cooperate with international partners.

I share with the members of the Forum that there is nothing very secretive about what we do. There is the Singapore Digital Gateway, and we talk quite openly about these things.

One common challenge they shared is building digital infrastructure in locations that have been incredibly difficult, and this is where AirTrunk comes in.

Through the work that you do, you are meeting a demand that appears to be still insatiable. That's a very good time to be in this sort of business. It's not the case that there are only upside opportunities, because there are a lot of issues to try and overcome. In many places where data centres are being built, local communities are very concerned about the energy consumption, competing with what citizens will be able to secure through the grid.

They are equally concerned about water consumption and what it means for their own needs, such as agriculture, which could be a very important lifeline for these communities.

It is an area that demands not just careful management but also has room for innovation.

Where Singapore is concerned, we see both challenges and opportunities.

AirTrunk has been with us on this journey for a decade, so you know that in terms of our density of data centres, we have about 1.4 gigawatts, and we've created headroom for up to 700 megawatts on Jurong Island.

We have designated this area, and it's a very important part of how we hope to support AI activities as they take off.

But we are not just doing capacity expansion. We've got to think about ways to operate more smartly, so we need a roadmap towards greener data centres, either using less energy or using greener options.

This raises the question, is it just the data centres or is it the way algorithms are written? Is there a way to make them greener? Green compute is another area that we'd like to work on. Keeping in mind that the data centres need to be able to talk to each other, our Digital Connectivity Blueprint attempts to look at ways in which we also secure a network of subsea cables that can turn connectivity into a reality. So, these are the challenges and opportunities that we see.

AirTrunk’s presence in Singapore is very much welcomed for a variety of reasons, not least because we appreciate your insights as an operator, developer, and investor.

We hope that by being able to hear from you in a much deeper way than we could before, we can develop better options that don't necessarily only serve Singapore's interests, but also the wider global interests.

How else does Singapore hope to support AirTrunk? I think in several ways. Domestically, your presence here will be part of an expansionary phase in our data centre landscape. You can tap into a very vibrant, innovative financial services system. I saw that you raised a very significant amount through green loans, and we know that the appetite for capital is very significant with the ongoing build-out. Access to funding is non-trivial and something that has to be addressed head-on internationally.

We are very interested to see ideas that we develop in Singapore, potentially becoming standards elsewhere in the world.

For example, we introduced standards for operating data centres in a tropical climate – can you operate them at a slightly higher temperature than what was initially thought possible? But we don't want to stop there.

We want to also look at standards when it comes to important things like getting subsea cables repaired in good time. Being able to create an environment that supports industry growth is something that Singapore is very keen to do.

Third, as AI becomes a bigger part of our economy, energy consumption will only increase. So, it is a non-trivial question for us to ask: “Can AI be made more resource efficient?” And I think if we can take a stab at this problem, not only in terms of the implementation, but at the research level, bringing the best minds to bear to tackle these important questions for humanity and Mother Earth – I think that could help contribute in a much bigger way.

The research community in Singapore is certainly one that we would encourage companies like AirTrunk to tap into.

Although Microsoft has been in Singapore for years, it was only last year that they set up a research facility here. Google DeepMind also decided to build a team here, and I can share with you that they're not looking for simple problems to solve. They really want to tackle the most difficult problems for humankind.

Just a few days ago, I met Quantinuum, a company that makes quantum computers.

They are building this platform to develop algorithms that solve problems that classical computers find very difficult to solve, and a lot of it will have to do with optimisation.

They were telling me about something really exciting. We have a very busy port, and port optimisation is actually very difficult to accomplish.

There are so many permutations that classical computers would take thousands of years to compute the most optimal solution – if, let's say a particular vessel was going to be delayed. Conceptually, you could get the quantum computer to narrow down the options and then apply AI using classical computers to complete the optimisation problem solving.

I share this only because I see that there are so many ways in which this network that is growing in Singapore can support your expansion plans and your dreams as a company.

I hope that Singapore as an environment, provides you, not only with this great view, but really a great backdrop for you to become the greatest company that you can be.

Congratulations once again. Thank you.