Transforming Data Centres: The AI Revolution in Asia Pacific
As artificial intelligence (AI) adoption accelerates across the Asia Pacific region, the data centre landscape is undergoing a significant transformation. Traditional facilities, designed for earlier generations of computing, are struggling to meet the demands of modern AI systems, which are characterized by high energy consumption and cooling requirements. By 2030, GPU-driven workloads could push power densities in data centres to unprecedented levels, signaling the urgent need for purpose-built “AI factory” data centres.
The Booming AI Data Centre Market
The market for AI data centres is projected to skyrocket from $236 billion in 2025 to nearly $934 billion by 2030. This explosive growth is primarily fueled by rapid AI adoption in sectors such as finance, healthcare, and manufacturing, all of which rely on high-performance computing environments powered by dense GPU clusters. These clusters demand significantly more energy and cooling capacity compared to traditional server setups.
Government investments in digitalization, the expansion of 5G networks, and the rollout of cloud-native and generative AI applications are further amplifying this demand in the Asia Pacific region. The pace at which compute needs are rising is unprecedented, and to keep up, operators must rethink their infrastructure strategies.
Strategic Infrastructure Evolution
Paul Churchill, Vice President of Vertiv Asia, emphasizes that meeting the growing demand for AI capabilities requires more than just larger facilities. “Infrastructure leaders must move beyond piecemeal upgrades. A future-ready strategy involves adopting AI-optimized infrastructure that combines high-capacity power systems, advanced thermal management, and integrated, scalable designs,” he stated.
Tackling Cooling and Power Challenges
As rack power densities rise from 40 kW to upwards of 250 kW by 2030, cooling and power delivery present significant challenges. Traditional air cooling methods are becoming inadequate for these high-density environments. To combat this, Vertiv is developing hybrid cooling systems that combine direct-to-chip liquid cooling with air-based solutions. These systems can adapt to fluctuating workloads, thereby enhancing energy efficiency and reliability.
Moreover, power delivery is evolving to meet the demands of dynamic AI workloads. Vertiv is innovating its rack power distribution units to handle higher voltages and improve load balancing, enabling operators to manage energy consumption more effectively. This approach is crucial in regions like Southeast Asia, where power grid stability is a concern.
Redesigning Data Centres for AI Workloads
The emergence of liquid-cooled GPU pods and 1 MW racks indicates a paradigm shift in data centre architecture. New facilities are being designed from the ground up to support AI workloads rather than retrofitting existing structures. “The future of data-centre architecture is hybrid, and these infrastructures require facilities to be built around liquid flow,” Churchill noted.
Next-generation data centres will integrate cooling, power, and monitoring systems from the chip level to the grid. This integrated design is essential for keeping pace with the performance expectations and sustainability goals of rapidly expanding hyperscale campuses across Asia Pacific.
Transitioning to AI Factory Data Centres
By 2030, Asia Pacific is expected to surpass the United States in data centre capacity, reaching nearly 24 GW of commissioned power. To manage this growth, enterprises are shifting from incremental upgrades to fully integrated AI factory data centres. Churchill suggests a phased approach for this transition:
- Integrated Planning: Combine power, cooling, and IT management systems into a cohesive strategy, simplifying deployment and laying a robust foundation for future scalability.
- Modular and Prefabricated Systems: Implement modular solutions that allow for phased capacity increases with minimal disruption to operations. “Companies can deploy factory-tested modules alongside existing infrastructure, gradually migrating workloads to AI-ready capacity without disruptive overhauls,” he explained.
- Embed Sustainability: Prioritize sustainability at every stage, utilizing lithium-ion energy storage and higher-voltage distribution systems to enhance efficiency and resilience.
DC Power: A Game Changer for AI Data Centres
Vertiv has recently introduced the PowerDirect Rack, a DC power shelf tailored for AI and high-performance computing. Switching to DC power can significantly reduce energy losses by minimizing conversion steps between the grid and servers. This aligns well with renewable energy and battery storage systems, which are gaining traction in the Asia Pacific region. In energy-constrained markets like Vietnam and the Philippines, flexible power solutions are crucial for maintaining operational continuity.
Emphasizing Sustainability in Data Centres
With AI driving up energy consumption, data centre operators face stricter regulations and increased grid constraints, particularly in Southeast Asia. Vertiv collaborates with operators to incorporate alternative energy sources, such as lithium-ion batteries and microgrids, to mitigate dependency on traditional power grids and enhance resilience. There is also a growing focus on solar-backed UPS systems and advanced energy storage technologies to balance loads and manage costs effectively.
Modular Solutions for Rapid Expansion
Emerging economies in the Asia Pacific face challenges such as limited space, unstable power supplies, and a shortage of skilled labor. In these contexts, modular and prefabricated data centre systems provide a viable solution. Prefabricated modules can reduce deployment times by up to 50%, while enhancing energy efficiency and scalability. This flexibility is particularly beneficial for AI workloads, which can experience rapid growth.
Preparing for a Demanding Future
The surge in AI adoption is fundamentally reshaping how data centres are designed and operated in the Asia Pacific. As workloads intensify and sustainability challenges mount, companies can no longer rely on outdated infrastructure. The transition towards AI factory data centres—powered by advanced cooling solutions, DC power, and modular systems—reflects a strategic shift in preparing for the next era of computing.
(Photo by İsmail Enes Ayhan)
Engagement Questions
- What are AI factory data centres, and why are they important?
AI factory data centres are purpose-built facilities designed to meet the high energy and cooling demands of modern AI workloads. They are crucial for handling the rapid growth of AI applications across various industries.
- How is cooling technology evolving to meet the needs of AI data centres?
Cooling technology is evolving through hybrid systems that combine liquid cooling with air-based methods, allowing for efficient thermal management in high-density environments.
- What role does DC power play in the future of AI data centres?
DC power can reduce energy losses significantly and aligns well with renewable energy and battery storage systems, making it a vital component in energy-constrained markets.
- Why is sustainability a central focus for data centre operators?
Sustainability is critical due to increasing energy consumption, regulatory pressures, and the need to balance operational efficiency with environmental responsibilities.
- What challenges do emerging economies in Asia Pacific face regarding data centre infrastructure?
Challenges include limited land availability, unstable power supplies, and shortages of skilled labor, which can hinder the deployment of traditional data centre solutions.
