Southeast Asia is undergoing one of the most substantial infrastructure expansions in its history. Hyperscalers, cloud providers, and sovereign AI initiatives are investing heavily in data centres at an unprecedented rate. The region’s data centre market is projected to reach US$30.47 billion by 2030, expanding at a compound annual growth rate of 14.24 percent.
This represents a remarkable demonstration of regional ambition.
However, a critical challenge looms: adequate power supply.
The Numbers Don’t Add Up
Power generation across Southeast Asia is increasing by less than seven percent each year, while data centre demand is rising by more than twice that rate. This widening gap cannot resolve itself automatically, and the longer it remains, the tighter the ceiling it imposes on the AI ambitions of governments and hyperscalers throughout the region.
This is a current constraint already influencing investment decisions, deal timelines, and site selection discussions across Singapore, Malaysia, Indonesia, the Philippines, and Vietnam.
Consider Malaysia, where data centre electricity consumption reached 8.5 TWh in 2024 and is projected to climb to 68 TWh by 2030—a eightfold increase over six years. At that level, data centres would consume up to 30 percent of the nation’s total electricity. Indonesia faces a comparable surge, with demand expected to rise from 6.7 TWh to 26 TWh, even as the country aims to triple its data centre capacity to roughly 800 MW. In the Philippines, consumption is forecast to grow from 1.1 TWh today to 20 TWh by 2030.
These are not isolated cases; they represent three of the region’s fastest‑growing data centre markets, each confronting the same structural challenge.
Why renewables alone cannot solve this
The intuitive solution is to expand solar and wind capacity. Southeast Asia possesses abundant solar resources and is increasingly deploying renewable energy. However, even with this growth, there is a limiting ceiling.
Our research indicates that intermittent renewables, especially solar, can realistically satisfy about 30 percent of the projected 2030 data centre demand. The remainder will require firm, dispatchable baseload power, as data centres cannot operate on intermittency and must have continuous 24/7 availability, irrespective of weather conditions.
We term this the “solar plateau” — the stage at which further solar installations no longer enhance reliability but merely add capacity that remains unusable during darkness. Batteries provide marginal assistance, yet at the scale required for Southeast Asia, storage alone is not a viable solution within this decade.
The uncomfortable reality is that roughly 70 percent of the ASEAN electricity grid still relies on coal and natural gas. If the AI buildout proceeds at its current pace without a clean baseload solution, data centres will increasingly depend on fossil fuels—a scenario that no hyperscaler, sovereign wealth fund, or national government wishes to be associated with.
Where This Leads
At Earth VC, we have been closely monitoring this dynamic, not merely as observers but as investors. We have backed Aalo Atomics, a U.S. developer building small modular reactors (SMRs) for data centre applications. Additionally, we invested in Blykalla, the Swedish SMR developer whose advanced lead‑cooled reactor attracted the attention of Vietnam’s Prime Minister, marking the first direct engagement between a head of government and an SMR developer in Southeast Asia.
We made these investments because the data pointed unequivocally in one direction: only a technology capable of delivering firm, zero‑carbon, scalable baseload power at the necessary speed and density can meet data centre requirements, and that technology is nuclear — specifically, the next generation of SMRs.
SMRs can generate up to 300 megawatts per unit, achieve a capacity factor of roughly 90 percent, and are designed for factory production and modular deployment. Unlike the massive, decade‑long nuclear projects of the past, SMRs represent infrastructure‑grade power plants built to deliver clean baseload power on a timeline of years rather than decades.
A Region Without Operational Reactors
Southeast Asia currently has no operating commercial nuclear reactors.
A region that accounts for one of the world’s fastest‑growing data centre markets and is already straining its power grid to meet existing demand has no nuclear capacity at all. The decision window to change this is narrowing; SMR projects that aim to be commercially viable by 2030 require capital allocation decisions and regulatory groundwork today. Some markets are taking steps forward. Singapore is exploring a blended‑finance model for SMR deployment that could serve as a regional template. Vietnam has signaled strong intent at the head‑of‑government level. Thailand and Indonesia are engaged in policy discussions that would have seemed unthinkable five years ago.
The question is no longer whether nuclear has a role in Southeast Asia’s energy future; the arithmetic settles that debate. The real question is which markets move fast enough to have clean baseload power in place when their data centre pipelines demand it, and which ones discover too late that they built infrastructure without securing the necessary power.
What Southeast Asia’s AI Ambitions Actually Require
Southeast Asian governments have made bold commitments to AI sovereignty, digital infrastructure, and economic competitiveness. These commitments are only as credible as the underlying power infrastructure.
A data centre lacking reliable power is not an AI hub; it is merely an expensive warehouse.
The supply wall is real. The window to address it is narrow, and the technology to solve it at scale, at speed, and without carbon emissions already exists today.
—
Also Read
- Federal Health Minister Emphasizes Local Governance as Key to Solving Basic Health Challenges
- IESCO Terminates Load Management Programme, Guarantees Continuous Power Supply
- NATO Summit and Russian Attacks Frame Ukraine’s Critical Air Defense Challenges]
- PML-N Veteran Advocates Dialogue to Defuse Azad Kashmir Crisis


