The explosion of artificial intelligence (AI) is ushering in a new era for global technology. Yet behind this dazzling transformation lies a rarely discussed reality: AI consumes massive amounts of electricity and water to sustain its computational workloads. Data centers the “brains” powering AI are now grappling with soaring operational costs, creating significant environmental, economic, and policy challenges for Asian countries.

AI Heats Up the Data Center Market

Next-generation AI models such as GPT and multimodal vision-language systems demand unprecedented volumes of GPUs/TPUs, driving electricity consumption far beyond that of conventional cloud services. Alongside this, cooling systems must operate continuously, consuming millions of liters of water annually to maintain stable temperatures. This dynamic has made electricity and water costs strategic variables, directly influencing data center location decisions and the scalability of major technology firms.

Johor (Malaysia): An Emerging “Hotspot” and Its Hidden Costs

As Singapore tightens its green standards for data centers, many investors have shifted operations to Johor, Malaysia—a region adjacent to the border with strong infrastructure and international connectivity. Johor has quickly become a “promised land” for the AI wave, attracting billions in FDI and large-scale hyperscale projects. However, this boom brings mounting challenges: surging electricity demand is straining the national grid, particularly during peak hours; cooling facilities consume vast amounts of water, raising concerns over shortages for households and industries; and the Malaysian government is weighing higher electricity tariffs for data centers while simultaneously promoting renewable energy adoption to ensure sustainable growth.

Currently, Johor hosts around 580 megawatts (MW) of data center capacity, while planned capacity including early-stage projects is nearly tenfold, according to market intelligence firm DC Byte. Based on estimates from PKnergy, that level of power could supply electricity to 5.7 million households every hour. Meanwhile, although Johor accounts for the bulk of Malaysia’s planned data centers, facilities are proliferating nationwide. Kenanga Investment Bank Berhad projects that by 2035, data center electricity consumption could reach the equivalent of 20% of Malaysia’s total generation capacity.

Water: The “Hidden” Cost That Can No Longer Be Ignored

While electricity is the most visible cost, water is often underestimated yet its impact is enormous. Evaporative cooling technologies, still widely used, consume staggering amounts of freshwater, often tens of millions of liters annually per megawatt of capacity. For clusters of hundreds of MW, this scale is unsustainable without alternatives. Emerging solutions include reusing wastewater for cooling, deploying liquid or immersion cooling to reduce freshwater dependency, and adopting transparent, frequent measurement of Water Usage Effectiveness (WUE). Estimates suggest a 100 MW facility consumes roughly 4.2 million liters of water daily the equivalent usage of thousands of residents.

Singapore: The Region’s Standard Setter

Singapore currently enforces the region’s strictest regulations, targeting WUE reductions to ≤ 2.0 m³/MWh within the next decade, while requiring low PUE levels, Green Mark Platinum certification, and advanced water optimization in cooling towers. These policies have driven some projects to Johor, yet they also establish Singapore as the “anchor of green standards” for the region. By setting a high bar, Singapore compels operators and investors across Southeast Asia to upgrade technologies and governance, or risk exclusion from the global digital infrastructure value chain.

The Road Ahead

To prevent Asia’s data centers from being overwhelmed by AI’s hunger for energy and water, a coordinated strategy is essential. Operators must invest in efficient cooling, commit to transparent PUE/WUE reporting, and secure long-term renewable energy procurement. Governments should adopt performance-based permitting, provide tax incentives for green technologies, and align electricity-water planning with digital infrastructure expansion. Investors, in turn, must rigorously assess resource and climate risks before committing capital, prioritizing projects with future-proof upgrade potential.

AI is fundamentally reshaping the data center industry, turning energy and water into existential factors. Johor illustrates the dual reality of opportunity and resource constraints, while Singapore continues to serve as the region’s benchmark-setter for sustainability. For Asia, the challenge is not merely to expand capacity but to build a resilient digital infrastructure ecosystem that balances growth with environmental responsibility.

(Source: CNBC)