Asia is cementing its position as a global data centre powerhouse, fuelled by rapid digitalisation and hyperscaler investments. The market is projected to grow from US$13.71 billion in 2024 to US$30.47 billion by 2030, according to Telecom Review Asia, marking a strong CAGR of 14.24%. The Asia-Pacific region is the fastest-growing data centre market worldwide, with billions in planned capacity across Singapore, Malaysia, India, and beyond. Johor in Malaysia, Batam in Indonesia, and Singapore are already emerging as some of the world’s most important hubs, while Australia, Thailand, and India continue to rise in prominence as upcoming data centre hubs.
But this growth comes with an enormous challenge: managing the heat generated by today’s workloads. Data centres already account for around 2% of global electricity consumption, and this share is climbing sharply as AI and high-performance computing drive demand. Traditional cooling systems, long the backbone of the industry, are struggling to keep up. In tropical regions where temperatures and humidity are high, and where energy efficiency and land use are tightly regulated, the need for scalable, high-density cooling has become a strategic imperative.
The limitations of traditional cooling
For decades, air cooling has been the standard method for managing heat. While it is a proven method, it is energy-hungry, requires bulky infrastructure, and is not very efficient for supporting racks that consume more than 50 kW, which is below the demands of today’s AI-driven systems. According to AIRSYS North America, AI servers can draw 50 to 100 kW per rack, significantly increasing the cooling load and creating pressure for alternative solutions.
Evaporative cooling offered some relief but is now under scrutiny, particularly in Asia, where regulators are restricting water usage in Singapore and Johor. Meanwhile, chilled water systems, which are effective for legacy equipment, struggle to provide the efficiency and responsiveness required by the new generation of GPUs and CPUs.
As chip density continues to increase, traditional methods are reaching their practical limits. Operators cannot afford outages or surging power bills. The search for more efficient cooling solutions has therefore accelerated across Asia.
Liquid cooling at a tipping point
This has brought the data centre industry to a tipping point. Liquid cooling, whether immersion-based or direct-to-chip, is no longer limited to early-stage trials and is being evaluated more widely as operators assess options for supporting higher-density workloads. Industry analysts predict that the data centre liquid cooling market may reach US$15 billion globally within five years, according to Data Centre Solutions, with Asia identified as a growing market.
According to Castrol ON’s The Dipping Point Report, research among global data centre experts indicates that existing immersion cooling deployments have reported average reductions of 15,000 MWh of energy and 3.5 million litres of water per facility per year compared to traditional air cooling. These findings reflect specific reported outcomes and may vary across different environments.
The same study found that 90% of data centre experts anticipate adopting immersion cooling as their primary system by 2030, with some considering earlier transitions. Among hyperscalers, 65% report using or experimenting with immersion cooling, and more than three-quarters view it as one potential approach to supporting rising compute power demands.
Navigating the transition: A fragmented problem
Despite this momentum, the transition has not been smooth. Findings from The Dipping Point survey indicate that the industry has been hesitant, with concerns centred on the fragmented nature of the cooling ecosystem. Industry analyses from IDTechEx describe the data centre cooling value chain as long and complicated, involving multiple stakeholders at the chip, server, rack, and facility levels. Operators often manage a patchwork of suppliers, including different vendors for fluids, cold plates, tanks, sensors, installation, and maintenance.
This piecemeal approach introduces risks at various stages, such as increased costs, deployment delays, and, in some cases, early system failures. External market reports identify high upfront capital requirements for retrofitting, a lack of standardised designs, and the need for specialised maintenance expertise as additional challenges. The Dipping Point also notes that these operational complexities contribute to slower adoption across some markets.
As a result, some operators are examining more integrated approaches that streamline coordination across the lifecycle of cooling systems. These can include involvement from the design and testing stages through to commissioning and ongoing support. Such approaches are intended to simplify implementation and support operational consistency as data centre development continues across Asia.
The future of data centre cooling in Asia
Operators across the region are examining more integrated approaches to cooling as part of their planning for higher-density workloads. Market forecasts from Data Bridge Market Research identify Asia-Pacific as one of the fastest-growing markets for data centre cooling, driven by digital transformation, 5G rollouts, and government support for digital infrastructure.
In edge computing environments, The Dipping Point survey notes that some operators view immersion cooling as suitable in situations where systems operate without power-intensive compressors, which can be beneficial in hot and humid climates.
Findings from The Dipping Point indicate that 74% of data centre experts view immersion cooling as one potential approach to meeting rising compute requirements. The survey also reports that 77% expect the broader transition toward liquid cooling technologies to generate financial and operational benefits, while 76% believe the industry will need to evaluate immersion cooling within the next three years to maintain performance improvements. These perspectives reflect ongoing reassessment of cooling strategies as workloads continue to grow across the region.
A defining moment
Asia’s digital economy is expanding rapidly. The region represents the largest share of generative AI users at work, with 83% of employees in APAC reporting that they use generative AI in the workplace, according to HCAMAG. As demand increases, the pressure on data centres to modernise their cooling infrastructure is expected to rise.
The tipping point is approaching, and discussions are shifting from whether cooling strategies will need to evolve to how quickly the industry can adapt.
To support growing workloads, operators are reassessing fragmented cooling approaches and evaluating scalable liquid cooling options alongside other modernisation efforts. These assessments form part of broader plans to address current thermal requirements and prepare for continued growth in Asia’s digital economy.
