The Factory-Built Future: How Waterless Cooling and Modular Prefabrication Are Solving AI's Infrastructure Crisis

SINGAPORE – October 20, 2026

The artificial intelligence boom has precipitated a quiet crisis in the physical world. As generative AI models grow exponentially more complex, the digital infrastructure required to train and run them is colliding with the hard limits of real estate, power availability, and municipal water supplies. For hyperscalers and enterprise operators, the old way of building data centers—custom-designed, stick-built facilities that take years to commission and consume millions of gallons of water—is no longer viable.

This tension was the defining undercurrent at Data Centre World Asia (DCWA) 2026, held late last month in Singapore. Amid the bustling exhibition floor, EPG Global Technology emerged with a compelling blueprint for the future of compute infrastructure. By showcasing its AI Core and Edge POD prefabricated modular solutions alongside the global launch of a revolutionary waterless cooling platform, the company illustrated a fundamental shift in how the industry must adapt. The message was clear: the future of AI infrastructure is factory-built, standardized, and hyper-efficient.

Tackling the AI Water Crunch in Tropical Hubs

Perhaps the most significant technological breakthrough presented at DCWA was the AeroLev 1850, a 1.85 MW-class air-cooled, magnetic-bearing chiller platform. Developed by MagPro—a joint venture established in May 2026 between EPG and Chinese magnetic-levitation pioneer CIGU Technology Corp.—the system addresses one of the most pressing existential threats to data center expansion in Southeast Asia: water scarcity.

Tropical data center hubs like Singapore and Malaysia face a unique geographical disadvantage. High ambient temperatures and year-round humidity render traditional air cooling wildly inefficient, forcing operators to rely heavily on cooling towers. These evaporative systems, while effective at managing heat, consume vast quantities of municipal water. In regions where water security is a matter of national priority, regulatory scrutiny over Water Usage Effectiveness (WUE) is intensifying.

The AeroLev 1850 bypasses this bottleneck entirely. Utilizing a cooling-tower-free architecture, it combines CIGU's oil-free magnetic-levitation technology with intelligent controls to deliver high-capacity cooling without the massive water footprint. CIGU, a company listed on the STAR Market with over 500 patent applications, brings deep expertise in maglev compressors, which boast an isentropic efficiency of up to 87%.

Crucially, the AeroLev 1850 is designed to deliver chilled-water outlet temperatures of up to 45°C. To the uninitiated, 45°C (113°F) sounds dangerously hot for a server room. However, this perfectly aligns with the latest ASHRAE TC 9.9 Thermal Guidelines for Class A4 equipment and the rapid adoption of direct-to-chip liquid cooling architectures. Modern AI clusters, packed with high-density GPUs, require liquid cooling to prevent thermal throttling. By supplying warmer fluid, the AeroLev system utilizes free-cooling functions more frequently, drastically reducing compressor runtime and overall energy consumption.

"Water constraints are rapidly becoming the primary limiting factor in data center site selection across the Asia-Pacific," noted one industry analyst observing the MagPro launch. "A platform that can reliably support high-density liquid cooling in a tropical climate, without relying on evaporative cooling towers, fundamentally changes the calculus for hyperscale deployments in these regions."

Factory-Built Compute: Solving the Speed-to-Market Crisis

While advanced cooling solves the operational sustainability equation, EPG's modular infrastructure offerings address the industry's other massive pain point: speed to market.

At the Critical Infrastructure Theatre, EPG Chief Operating Officer Johnny Zhang highlighted the growing demands AI is placing on data center infrastructure, emphasizing the urgent need for faster deployment. Stick-built data centers often suffer from unpredictable construction delays, local labor shortages, and complex supply chain logistics. When AI compute capacity is a competitive moat, a two-year construction timeline is an eternity.

EPG's response is to move construction off-site. The company's AI Core solution takes a prefabricated, modular approach, integrating standardized power, cooling, and critical infrastructure modules in a controlled factory environment. EPG Chief Technology Officer Enso Li demonstrated how this factory prefabrication accelerates deployment, allowing operators to commission capacity in a fraction of the time required for traditional builds.

For distributed environments, the company introduced Edge POD, an all-in-one solution designed for 1–10 MW applications. Edge POD is particularly relevant for low-latency computing requirements, where power availability and deployment speed are critical. By standardizing the design and testing the modules before they ever reach the site, operators mitigate on-site construction risks and ensure predictable performance from day one.

This modular approach also offers unparalleled flexibility. As compute requirements grow, operators can seamlessly scale their infrastructure by adding standardized blocks, rather than retrofitting existing, inflexible facilities. It is a shift from treating data centers as bespoke real estate projects to treating them as deployable technology products.

The APAC AI Corridor and Global Delivery

The announcements at DCWA 2026 also highlighted the shifting geopolitical and supply chain dynamics of the global data center market. EPG, with its R&D centers in Singapore, Malaysia, and Shanghai, and six advanced manufacturing hubs across Malaysia and China, is uniquely positioned to bridge Chinese manufacturing capacity with international hyperscale demand.

This cross-border integration is creating a robust "APAC AI Corridor." Southeast Asia is rapidly emerging not just as a destination for data centers, but as a strategic integration hub where advanced components—like CIGU's maglev compressors—are assembled into deployable infrastructure modules for the global market.

During the event, EPG solidified this positioning by signing cooperation agreements with several international industry partners spanning AI data center planning, construction, and delivery. With a workforce of over 1,600 employees and nearly 300 researchers and engineers, the infrastructure provider is scaling its delivery network to support enterprise and hyperscale deployments across APAC, EMEA, and beyond.

"AI computing is reshaping the infrastructure requirements of data centers," said Alick Wan, Founder and Chairman of EPG. "Our focus extends beyond individual technological breakthroughs. By combining modular and standardized solutions with global delivery capabilities, we aim to help customers deploy and scale computing infrastructure efficiently across markets."

As the digital economy becomes increasingly reliant on artificial intelligence, the physical systems that underpin it must evolve. The innovations showcased in Singapore demonstrate that the path forward relies on engineering efficiency into the very fabric of the building process. By standardizing the complex and eliminating the unsustainable, the industry is finally building an infrastructure model capable of supporting the next generation of digital growth.