In Loyang, computers sit in liquid.
The equipment appears in Singapore's own account of its green data-centre push: servers cooled by immersion at a facility run by ST Telemedia Global Data Centres. It is a clean image for a messy bargain. The city-state wants more computing, more AI, more cloud capacity, more cybersecurity, more fintech, more medical data, more automation in port and factory operations. The machines want electricity. They want land. They want cold.
That is where a guide to Singapore's ICT market should begin: not with software logos, but with heat.
Mordor Intelligence estimates that Singapore's information-and-communications-technology market will reach US$79.24 billion in 2026, up from US$69.77 billion in 2025, and rise to US$149.68 billion by 2031. The report's projected compound annual growth rate, 13.57 percent, is the sort of figure that can make a market look smooth. Singapore's digital economy is anything but smooth. It is a concentrated system of state planning, private cloud spending, regulated finance, ageing hospitals, scarce engineers, and server farms pressing against the limits of a small island.
ICT, in this market, means more than laptops and telecom lines. It includes hardware, software, IT services, cybersecurity, communication services, and the infrastructure underneath it all: data centres, colocation halls, servers, storage, switching gear. The plumbing. The bill.
The State as Buyer
Singapore has spent decades turning public administration into a technology buyer. The National Computerisation Programme started in the 1980s. Singpass became part of everyday identity. PayNow became part of everyday payment. By the time Smart Nation 2.0 was published in 2024, the government could point to a country where nearly all public services could be completed online, 99 percent of resident households were connected to the internet, and 5G standalone coverage had passed 95 percent outdoors.
The state did not leave the market to discover demand by itself. It created demand, certified it, subsidised parts of it, and regulated the rest.
That matters because Singapore's ICT market is shaped less like a loose consumer internet boom than like a public-private operating system. Agencies push digital identity, cybersecurity, e-payments, smart-city systems, SME grants, and AI adoption. Banks, telcos, hospitals, manufacturers, logistics firms, and cloud vendors then build around those standards. Official data put the digital economy at 17.7 percent of GDP in 2023, up from 13.8 percent in 2018. The information-and-communications sector alone generated S$36.3 billion in nominal value added that year.
Big numbers. Tight space.
Clouds With Borders
The cloud in Singapore is local and regional at the same time. AWS, Google Cloud, Microsoft Azure, Alibaba Cloud, telcos, colocation providers, systems integrators, and managed-service firms serve domestic customers from a place that also functions as a command centre for Southeast Asia.
AWS announced in 2024 that it would put an additional S$12 billion into Singapore cloud infrastructure through 2028, after investing S$11.5 billion in its Singapore region through 2023. The company said the new spending would support an estimated 12,300 full-time-equivalent jobs a year across its local supply chain. The number is broad. It includes construction, facilities, engineering, telecommunications. In Singapore, cloud expansion is not just code. It is concrete and cable tray.
Mordor's segmentation shows why the shift is still uneven. On-premise systems accounted for 55.78 percent of the market in 2025, while cloud deployments are projected to grow faster, at 17.15 percent annually. The contradiction is the point. Banks, government-linked entities, healthcare institutions, and large multinationals may buy cloud tools, but many still keep sensitive workloads close. They want elasticity. They also want auditability, latency control, and a defensible answer when regulators ask where data sat and who touched it.
Hybrid architecture is not a compromise in Singapore. It is the normal shape of caution.
The Small-Firm Gap
The public story says Singapore's SMEs have digitalised. In one sense, they have. Smart Nation figures say 95 percent of SMEs use digital technology. The more useful question is what kind.
IMDA's 2024 digital-economy report found that AI adoption among SMEs reached 4.2 percent in 2023. Among non-SMEs, it was 44 percent. That gap explains much of the next phase of spending. A small logistics company may already use e-invoicing, a workforce app, and a cloud accounting system. That does not mean it has clean data, an internal AI champion, a cybersecurity budget, or enough managerial slack to rebuild operations around predictive tools.
Government grants soften the edge. SMEs using digital tools through the Productivity Solutions Grant reported average cost savings of 48 percent per solution from 2018 to 2023, according to IMDA. Mordor expects SMEs to grow faster than large enterprises in ICT spending, at a 14.88 percent CAGR through 2031. The cloud subscription model helps. So do pre-approved software lists and consultants.
Still, there is a stubborn difference between buying software and changing work. A grant can lower the invoice. It cannot make the owner of a twenty-person firm trust an algorithm with inventory, hiring, credit, or customer service.
Banks, Hospitals, Factories
Finance remains the largest vertical in Mordor's forecast, accounting for 21.78 percent of the ICT market in 2025. That is not surprising in a city that sells trust as an export product. The rise of digital banks and the pressure of real-time payments have pushed incumbents toward API architectures, fraud analytics, compliance automation, and zero-trust security.
Healthcare is the faster mover in the forecast, with Mordor projecting a 17.93 percent CAGR for healthcare and life sciences through 2031. The reasons are less glamorous than the software brochures. Singapore is ageing. Clinicians are scarce. Chronic disease management consumes capacity. Public hospitals and healthtech agencies need systems that can support tele-consults, imaging analysis, data-sharing, patient triage, and operational scheduling without letting sensitive medical records leak into the open.
Manufacturing follows a different logic. The question is whether AI can detect a defect before a human inspector sees it, predict a machine failure before a line stops, or make a semiconductor supplier's schedule less brittle. The 2026 update to Singapore's National AI Strategy names advanced manufacturing, financial services, connectivity, and healthcare as national AI mission sectors. Together, the government said, they contributed about 40 percent of GDP in 2025.
This is industrial policy with a dashboard.
The Grid Says No
The hard constraint is power.
Singapore paused new data-centre development in 2019 while it reviewed how much additional capacity the island could absorb. When the door reopened, it did so with conditions. A 2022 pilot allocated about 60 megawatts and required applicants to meet green criteria, including stronger efficiency standards. The Green Data Centre Roadmap now aims to provide at least 300 megawatts of additional data-centre capacity in the near term, with more possible through green energy deployments. A related government account put the potential addition at 300 megawatts, plus another 200 megawatts reserved for operators using green energy options, on top of more than 70 data centres and about 1.4 gigawatts of existing capacity.
Those are not abstract figures. Data centres are unusually honest buildings. They convert strategy into heat.
The new standards push operators toward hotter data halls, efficient IT equipment, liquid cooling, immersion cooling, renewable procurement, and better software discipline. A Singapore Standard launched in 2025 aims to cut data-centre IT energy consumption by at least 30 percent through more efficient equipment and operation at higher temperatures. The market will not be allowed to grow in the old way. That is policy. It is also physics.
The Security Bill
The more Singapore digitises, the larger its attack surface becomes. The Cyber Security Agency of Singapore wrote in its 2024/2025 landscape report that advanced persistent-threat activity had continued to grow in scale and sophistication, naming UNC3886 as one group targeting high-value strategic targets, including critical infrastructure. It also noted ransomware and infected infrastructure as local concerns, often involving old malware strains.
Old malware is an ugly phrase. It means the frontier is not always breached by a brilliant new exploit. Sometimes the door was patched years ago and left open anyway.
Mordor cites a shortage of 2,800 to 4,400 cybersecurity professionals, a drag on rollout timelines and a wage problem for firms that cannot outbid banks, cloud providers, and government contractors. The market response is predictable: managed detection, identity tools, secure-access service edge, AI anomaly detection, compliance automation. The human shortage becomes a product category.
Security spending in Singapore is often described as resilience. A less polite word is maintenance. The national digital machine has to be kept clean enough to run.
What the Forecast Cannot Smooth
Market reports impose order. They have to. Infrastructure has a share, software has a CAGR, cloud has a curve, healthcare has a faster line than finance. Mordor's model says software will grow at 16.35 percent a year through 2031. It says large enterprises still dominate spending, with 66.78 percent of the market in 2025. It says the market concentration is medium.
The lived system is rougher.
A minister can announce AI missions. A vendor can open an accelerator. A bank can deploy a fraud model. A polytechnic can train more cloud engineers. None of that erases the grid queue, the shortage of security staff, the conservative instincts of regulated institutions, or the fact that many SMEs still see AI as too expensive, too vague, or simply irrelevant to the scale of their businesses. IMDA found that 57.5 percent of SMEs not using AI said they saw no need because their businesses were small.
That sentence may matter more than another billion-dollar infrastructure pledge.
Singapore's ICT market is growing because the country has made digital capability part of its basic economic machinery. It is also growing because the machinery keeps demanding more: more compute, more governance, more cooling, more skilled workers, more trust. At an Asia Tech x Singapore panel cited in the data-centre roadmap coverage, Green Software Foundation chair Sanjay Podder put the matter more plainly: "Good software programming is something we have lost track of."
The forecast ends at US$149.68 billion. The power meter does not.
