MDDI 演讲稿 · 2026-02-24
陈杰豪高级政务部长在新电信数码基建公司与英伟达应用 AI 卓越中心启动仪式上的演讲
Speech by SMS Tan Kiat How at Launch of Singtel Digital InfraCo - NVIDIA Centre of Excellence for Applied AI
要点
- • 黄循财总理在 2026 预算案中明确：新加坡的 AI 优势不在于做最大的前沿模型，而在于「有效、负责、快速」地落地 AI。
- • 新加坡 AI 卓越中心（CoE）已超过 60 个，分两类——行业级 CoE（解决整个价值链共性问题，如制造业的预测性维护）+ 公司级 CoE。
- • SIT × NVIDIA AI 中心（SNAIC）已支持 70 家公司部署 50 个 AI 方案，覆盖制造、医疗、金融、交通；如与陈笃生医院合作的「自动视力检测」系统。
- • 两个数据中心治理工具：①《绿色数据中心路线图》（2024 年 5 月）+「IT 能源效率标准」（2025 年 6 月）；②今年内 MDDI 将向国会提交《数字基础设施法》（DIA），为所有数据中心设定 PUE 等基线能效要求，并要求重大云服务商与数据中心管理安全风险、报告事故。
- • 新加坡数字基础设施每 15 年大跳一次：30 年前 Singapore ONE、15 年前下一代国家宽带网络、5 年前 5G SA。下一波是 AI 普及的世界——CoE AAI 将借助新电信的网络（尤其是边缘网络）加速这一进程。
完整译文（中文）
MDDI 英文原文译文 · 翻译日期：2026-05-02
新电信数码基建公司（Singtel Digital InfraCo）首席执行官 Bill Chang 先生、
英伟达（NVIDIA）解决方案架构与工程副总裁 Marc Hamilton 先生、
各位嘉宾、朋友与同仁：
早安。我很高兴今天和大家一起，出席新电信—英伟达「应用 AI 卓越中心」（COE AAI）的启动仪式。
Bill、Marc——祝贺你们两个团队让这个中心落地。这是一项重要的合作，我很高兴能在这里见证它的启用。
在 2026 年财政预算案中，黄循财总理勾勒了政府的愿景——让新加坡把 AI 当作战略优势来掌握。
正如总理所言：新加坡的优势不在于打造最大的前沿模型，而在于「有效、负责、快速」地落地 AI。
我们专注的是——AI 如何被用来解决真实世界的问题，为我们的企业与人民创造价值。
这一努力中很重要的一环，是培育一个能在整个经济体中开发、部署 AI 方案的创新者与工程师生态。
自 2023 年底我们启动更新版《国家 AI 战略》以来，新加坡已经设立了 60 多个「AI 卓越中心」（COEs）。
其中一些是「行业级 CoE」——把研究机构、产业、政府聚到一起，为整条价值链上的共性挑战开发 AI 方案。例如「制造业行业级 AI CoE」，正在做该行业的预测性维护与质量保证。
另一些是「公司级 CoE」——单一企业建立内部 AI 能力，改进自身工作流程，给客户提供更好的服务。
「SIT × NVIDIA AI 中心」（SNAIC）是这个生态里令人欢迎的新成员。该中心帮助大小企业测试、开发并落地 AI 方案。
迄今，SNAIC 已支持 70 家企业部署 50 个 AI 方案，覆盖制造、医疗、金融、交通等行业，并产生真实业务影响。
比如，SNAIC 与陈笃生医院合作开发协同 AI 工具，帮助临床医生更快、更准地提供照护。
其中一项是「自动视力检测」（Automated Visual Acuity Detection）：用 AI 驱动的语音识别与姿态监测自动完成视力检查——相比传统方法所需的大量临床医生介入，这是一个重要进步。
这些 AI 工具减轻了医生的工作量，缩短了病人的等待时间，让视力检测更快，并在病人康复旅程中提供更好的支持。
现实是——这些以 AI 为驱动的数字方案，无法靠 AI 自己开发与落地。我们需要建模者、工程师、开发者、数据科学家以及更多角色，去运行先进的计算平台、参与应用 AI 的研发，并构建解决方案。
新加坡致力于在国家 AI 推进中培育一个深厚的科技与 AI 人才池。MDDI 将在即将到来的预算 / 拨款总目（COS）国会辩论中分享更多计划。
我很高兴新电信与英伟达是这一努力的有力支持者。
由 IMDA「TechSkills Accelerator」（TeSA）支持的 SNAIC AI 项目，将基于 NVIDIA Deep Learning Institute 课程，为应届生与中途职业人士提供两个月的密集再培训。他们还将通过四个月的产业实操项目获得实战技能。
我们也在通过新加坡人工智能机构（AI Singapore）的「AI 学徒计划」（AI Apprenticeship Programme）以及与新电信等领先科技公司的合作，扩大 AI 实务人才池。完成项目后，这些实务人才会在生态中的多家公司担任 AI 与机器学习工程师等岗位，许多人最终在学徒结束后留下成为正式员工。
另一个关键的支撑要素，是承载这些平台与方案的数字基础设施。今早我聚焦谈两个——数据中心（DCs）以及云服务平台。
新加坡是区域级数据中心枢纽，DC 密度在本地区数一数二。
这些数据中心是数字经济的基础。
它们承载支撑新加坡 AI 雄心所需的算力。
但它们也是密集的资源消耗者，尤其是电与水。
今天，新加坡已承载超过 1.4 GW 的 DC 容量，而随着 AI 使用的增加，对 DC 的需求只会增长。
因此，对新加坡这样自然资源受限的紧凑型城市国家来说，确保这种增长是可持续的至关重要。
IMDA 在 2024 年 5 月发布《绿色数据中心路线图》，铺设可持续增长路径，把新加坡的资源约束转化为机会。
我们正帮助数据中心借力绿色能源来增长，并采取具体措施提升能效。
为支持产业转向，我们启动了「DC 行业能效补助」，帮助运营商与用户把 IT 设备升级到更节能的型号。
我们也在 2025 年 6 月发布了《IT 能源效率标准》，指导 DC 运营商与用户选用节能 IT 设备，并更高效地运行这些设备。
今年稍后，MDDI 将向国会提交《数字基础设施法》（Digital Infrastructure Act, DIA）。
这项拟议立法将为所有数据中心——既有的与未来要建的——设立基线能效要求。我们将从对数据中心的「电力使用效率」（PUE）要求开始。
MDDI 与 IMDA 一直在与产业利益相关方协商，以确保拟议要求虽具雄心，但务实可行，并与国际基准一致。
我们希望新加坡的数据中心更节能、更可持续——不仅是未来几年新建的 DC，也包括既有的 DC，它们将在更新技术与 IT 设备时，逐步达到 DIA 的要求。
通过系统性地提升所有数据中心的能效与可持续性，我们能为新加坡数字经济日益增长的需求腾出更多空间——尤其是 AI 采用的推动。
除了可持续之外，DIA 也将为重大云服务商（CSPs）与数据中心制定要求，以增强其韧性与安全性。
数据中心与云服务平台支撑着许多与日常密切相关的数字服务、运营与交易——电子银行与支付、网约车、电商、数字身份认证等。
一旦数字基础设施被中断或遭网络攻击，对我们的经济与人民日常活动会产生重大影响。
彻底消除这些风险并不现实——就像所有数字系统一样，由于种种原因，中断会偶有发生。
但同样不负责任的是——当存在外溢效应（影响整体经济、社会、国家安全、新加坡国际声誉）时，政府如果完全放手，让 CSP、DC 运营商与他们的企业客户之间纯粹靠商业安排来处理。
因此，DIA 将要求重大 CSP 与数据中心采取措施管理安全风险、把中断降到最低、确保业务连续性。
这包括基于去年发布的《数据中心与云提供商咨询指南》制定的事前要求，以及事后事件报告要求。
我们准备好后会分享更多信息。
DIA 将是确保新加坡数字基础设施可持续、有韧性、网络安全的重要一步——重要的是，能够满足未来需求。
我要感谢许多就 DIA 提供意见的利益相关方与合作伙伴；当我们今年稍后正式提交立法时，恳请各位继续支持这项重要法律。
我以一个问题与一个邀请来结束今早的发言。
问题是：「在 AI 无处不在的未来世界里，数字基础设施会是什么样子？」
新加坡一贯投资数字连通基础设施，确保它面向未来，并成为我们企业的战略竞争优势之源。
30 年前，我们启动了「Singapore ONE」项目，把全 IP 核心铺到了我们的电信基础设施里——这把新加坡带上了网。
15 年前，我们大跨步推出「下一代国家宽带网络」，把光纤拉到所有家庭与企业，成为全球第一个提供全国 1 Gbps 连接的国家——把新加坡带进了宽带世界。
5 年前，我们部署了独立组网的移动 5G（SA）网络，成为最早实现全国 SA 覆盖的国家之一——把新加坡带进了「移动优先」的数字世界。
展望未来 5 到 10 年——AI 驱动的服务与系统将深度交织进我们的日常生活与活动：工作场所与家中的机器人、执行任务并支撑我们决策的智能体，以及更多今天还无法想象的创新。
但有一件事是确定的：我们的数字基础设施必须演化、刷新，跟上这一波 AI 创新。它必须持续做到面向未来、有韧性、安全、能响应新需求。
新加坡没有水晶球，但我们希望是最早把握这个未来轮廓的人之一，并能与伙伴一起在新加坡使这个未来成为现实。
在这方面，我很高兴 COE AAI 将借助新电信的基础设施，特别是它的边缘网络，加速这一空间的发展。
像「榜鹅数码区」（Punggol Digital District, PDD）这样的环境，可以作为公司大规模测试与打磨 AI 应用的试验场。
公司们在真实运营环境里实验与测试具身 AI、机器人、自主系统等技术。
为支持这一点，IMDA 与新加坡理工大学（SIT）、裕廊集团（JTC）合作，打造「片区级机器人试验场」（precinct scale robotics testbed），让产业伙伴能共同开发、试点、商业化并规模化真实世界的机器人方案。
另一个例子是松下（Panasonic）去年在 PDD 启用的旗舰创新中心，他们正试点一套「AI 驱动的人—机—设施」自动化系统，用于智能楼宇与机器人应用。
NTUC 平价集团（FairPrice Group）则开设了世界上第一家由生成式 AI 驱动的超市。
我邀请产业利益相关方与技术伙伴一起加入我们，去发现并塑造未来的数字基础设施。
结束之前，我再次向新电信与英伟达致以最诚挚的祝贺。
我期待看到这个中心在未来几年里产生的影响与成长。
非常感谢大家。
英文原文
MDDI 官网原始记录 · 抓取日期：2026-05-02
Mr Bill Chang, CEO of Singtel Digital InfraCo,
Mr Marc Hamilton, Vice President, Solutions Architecture and Engineering, NVIDIA,
Distinguished guests, friends and colleagues,
Good morning. I’m very happy to join all of you today for the launch of the Singtel–NVIDIA Centre of Excellence for Applied AI (COE AAI).
Bill, Marc – Congratulations to both your teams for bringing this Centre to life. This is an important partnership, and I am glad to be here to mark its launch.
At Budget 2026, Prime Minister Lawrence Wong set out the Government’s ambition for Singapore to harness AI as a strategic advantage.
As PM put it, Singapore’s advantage does not lie in building the largest frontier models – it lies in deploying AI effectively, responsibly, and at speed.
We are focused on how AI can be harnessed to solve real world problems and create value for our businesses and people.
An important part of this effort is to foster an ecosystem of innovators and engineers who can develop and deploy AI solutions across our economy.
Since we launched the refreshed National AI Strategy in end-2023, more than 60 AI Centres of Excellence (COEs) have been set up across Singapore.
Some of these are sectoral COEs that bring together research institutions, industry, and government to develop AI solutions for common challenges across the entire industry value chain. An example is the Sectoral AI COE for Manufacturing, which is working on predictive maintenance and quality assurance for the sector.
Others are company COEs, where individual businesses are building up their internal AI capabilities to improve their work processes, delivering better services to customers.
The SIT x NVIDIA AI Centre (SNAIC) is a welcome addition to the ecosystem. The Centre helps businesses – large and small, to testbed, develop, and implement AI solutions.
To date, the SNAIC has supported 70 companies to deploy 50 AI solutions with real business impact across industry sectors such as manufacturing, healthcare, finance, and transport.
For example, SNAIC worked with Tan Tock Seng Hospital to develop collaborative AI tools that support clinicians in delivering faster and more accurate care.
One such tool is the Automated Visual Acuity Detection that uses AI-driven speech recognition and posture monitoring to automate eye tests – this in comparison to more traditional methods that require significant clinician involvement.
These AI tools have reduced clinicians’ workload, reduced patient waiting times, enabled faster eyesight testing, and provided better support for patients in their recovery journey.
The reality is that the development and deployment of such AI-powered digital solutions in the real world cannot be done by AI. We need model makers, engineers, developers, data scientists and many more roles to run advanced computing platforms, contribute to applied AI research and development, and build solutions.
Singapore is committed to nurture a deep pool of tech and AI talent as part of our national AI push. MDDI will share further plans at the coming Budget/COS debates in Parliament.
I am glad that Singtel and NVIDIA are strong supporters of this effort.
The SNAIC AI Programme, supported under IMDA’s TechSkills Accelerator initiative, will provide fresh graduates and mid-career professionals with two months of intensive upskilling based on NVIDIA’s Deep Learning Institute curriculum. They will also gain practical skills from four months of hands-on industry projects.
We are also expanding the pool of AI practitioners through the AI Apprenticeship Programme under AI Singapore and collaborations with leading technology companies such as Singtel. After the programme, these practitioners take on roles such as AI and machine learning engineers in companies across the ecosystem. Many of them eventually join these companies on a full-time basis after their apprenticeship.
Another critical enabling element is the digital infrastructure on which these platforms and solutions are built and run on. This morning, let me touch on two of them – data centres, or DCs, and cloud services platforms.
Singapore is a regional data centre hub with one of the highest concentrations of DCs in the region.
These data centres are foundational to our digital economy.
They house the compute capacity that supports Singapore’s AI ambitions.
But they are also intensive users of resources, especially power and water.
Today, Singapore already hosts more than 1.4 gigawatts of DC capacity, and we can only expect the demand for DCs to grow with the increased use of AI.
It is therefore crucial for Singapore, a compact city-state with natural resource constraints, to ensure that this growth is sustainable.
IMDA launched the Green Data Centre Roadmap in May 2024 to chart a sustainable growth pathway and turn Singapore’s resource constraints into opportunity.
We are helping data centres tap on green energy sources for growth and take concrete steps to improve their energy efficiency.
To support industry in this shift, we launched the Energy Efficiency Grant for the DC sector to help operators and users refresh their IT equipment to more energy-efficient models.
We also released the IT Energy Efficiency standard in Jun 2025 to guide DC operators and users on selecting energy-efficient IT equipment, and operating IT equipment more efficiently.
Later this year, MDDI will table the Digital Infrastructure Act (DIA) in parliament.
The proposed new legislation seeks to establish baseline energy efficiency requirements for all data centres – existing and new ones that will be built. We will start by imposing Power Usage Effectiveness (PUE) requirements on data centres.
MDDI and IMDA have been consulting industry stakeholders to ensure that the proposed requirements – while ambitious – are practicable and consistent with international benchmarks.
We want to the DCs in Singapore to be more energy efficient and sustainable – not just for new DCs built in the coming years, but also existing DCs as they refresh their technologies and IT equipment over time to meet the DIA requirements.
By systematically raising the energy efficiency and sustainability of all our DCs, we can create more headroom to support the growing demands of Singapore’s digital economy, particularly with the push for AI adoption.
Beyond sustainability, the DIA will also establish requirements to enhance the resilience and security of major cloud service providers (CSPs) and the data centres.
Data Centres and cloud services platforms enable digital services, operations, and transactions across many essential day-to-day functions such as e-banking and payments, ride-hailing, e-commerce, and digital identity authentication.
There will be significant impact to our economy and people’s daily activities should disruptions or cyber-attacks to the digital infrastructure occur.
While it is not realistic to eliminate these risks, like all digital systems, such disruptions may happen from time to time, due to different reasons.
But at the same time, it is irresponsible for the Government to take a hands-off approach and leave it entirely to commercial arrangements between the CSPs, data centre operators and their enterprise customers, when there are spillover implications to the broader economy, society, national security and Singapore’s international reputation.
The DIA will therefore require major CSPs and data centres to implement measures to manage security risks, minimise disruptions, and ensure business continuity.
These include ex-ante requirements based on the Advisory Guidelines for Data Centres and Cloud Providers released last year, as well as ex-post incident reporting requirements.
We will share more information when ready.
The DIA will be a major step to ensure that Singapore’s digital infrastructure is sustainable, resilient and cybersecure, and importantly able to meet future demands.
I would like to thank the many stakeholders and partners that have contributed your views to the DIA, and I seek your continued support for this important legislation when we introduce it later this year.
I would end my remarks this morning by posing a question and sending out an invitation to all of you.
The question is: “How does the digital infrastructure look like in the future AI-pervasive world?”.
Singapore has consistently invested in our digital connectivity infrastructure to ensure it is future ready and a source of strategic competitive advantage for our enterprises.
30 years ago, we launched the Singapore ONE project which put in place a fully IP core across our telecommunication infrastructure. This brought Singapore online.
15 years ago, we took a major step to implement the Next Generation National Broadband Network which pulled fibre optic cables to all homes and businesses, becoming the first country to provide up to 1Gbps connection nationwide. This brought Singapore into the broadband world.
Five years ago, we deployed mobile 5G Standalone (SA) Networks, becoming one of the first countries to have nationwide SA coverage. This brought Singapore into a mobile-first digital world.
Looking into the next five to 10 years, AI-powered services and systems will be deeply interwoven into our daily lives and activities, from robots in workplaces and homes, to agents carrying out tasks and supporting our decisions, to many more innovations that we have yet to imagine.
But what is clear is that our digital infrastructure must evolve and refresh to keep pace with this wave of AI innovation. It must continue to be future-ready, resilient, secure and responsive to new demands.
While we don’t have a crystal ball in Singapore, we want to be among the first to grasp the contours of this future and be able to work with partners to enable this future here.
In this regard, I am glad that the COE AAI will be tapping on Singtel’s infrastructure, especially its edge networks to accelerate development in this space.
Environments like the Punggol Digital District (PDD) can serve as testbeds for companies to test and refine their AI applications at scale.
Companies are experimenting and testing technologies such as embodied AI, robotics, and autonomous systems in a real, operational environment.
To support this, IMDA is partnering with SIT and JTC to create a precinct scale robotics testbed that enables industry partners to co-develop, pilot, commercialise, and scale real world robotics solutions.
Another example is Panasonic that launched its flagship Innovation Hub at PDD last year. It is piloting an AI-driven Human-Robot-Facility automation system for smart building and robotics applications.
And NTUC FairPrice Group opened the world’s first generative AI-powered supermarket.
I invite industry stakeholders, technology partners to join us in discovering and shaping the future digital infrastructure.
In closing, let me once again extend my heartiest congratulations to Singtel and NVIDIA.
I look forward to seeing the Centre’s impact and growth in the years ahead.
Thank you very much.