Perspectives
Future-proofing data centres in Asia
Amid the ongoing global expansion of data centre construction, Asia has emerged as a key hub fuelled by the rise of Artificial Intelligence (AI), cloud-based platforms, supportive government policies, and the widespread use of the Internet of Things (IoT).
But the rapid growth of data centres raises important environmental concerns, posing a critical question to the industry: how can we meet the increasing demand in Asia while scaling up responsibly?
Asia offers new opportunities due to cost-effective land, energy and construction. Driven by the developing countries in the region, the continent also provides a dynamic and flexible landscape for data centre companies to maximise and learn from.
While the cost of building data centres varies worldwide, typically ranging from US$9 million to US$15 million per megawatt, Asia offers more affordable options, with costs around US$8 to US$10 million.
Therefore, countries in Asia are eager to support data centre construction, as it boosts economic growth, accelerates innovation, and enables local businesses to leverage high-performance computing.
However, this growth also poses environmental and community risks. Companies are being challenged to ensure sustainability by implementing solutions for efficient energy consumption, effective heat management, and an integrated approach to designing, building, and operating data centre facilities.
In legacy data centres, power is supplied from diverse local grids at low or medium voltage. Standby generators, transformers, load banks, and cooling system heat rejection are typically located outside the buildings, with underground fuel oil storage tanks. Generators were typically placed on the ground floor or roof, providing enough power for IT equipment and cooling needs, while power distribution was done at low voltage, with two separate circuits for each cabinet.
With the rise of AI installations, power demands have significantly increased. AI cabinets now require much higher power loads and densities, leading to more frequent use of duplicate high voltage feeders and block distribution, reducing the number of cabinets per power block.
Data centres must adapt their power infrastructure to accommodate clusters of AI cabinets. The high utilisation rate of AI processors within these clusters requires careful power infrastructure planning, as well as improvements in IT equipment to enhance temperature tolerance.
As a global leader in this field, Mace is supporting efforts to rethink how data centres are delivered in the region, making them more efficient, resilient, and environmentally sound through initiatives like the following:
1. Renewable energy integration: Using solar, wind, and hydroelectric power to reduce reliance on fossil fuels.
2. Green building certifications: Pursuing LEED certification and other green building standards for sustainable construction practices.
3. Energy efficiency: Implementing advanced cooling solutions like liquid immersion cooling and direct-to-chip cooling to lower energy consumption.
4. Water conservation: Using water-efficient cooling towers and closed-loop cooling systems to minimise water usage.
5. Recycling and waste reduction: Implementing comprehensive recycling programs for electronic waste and construction materials.
However, in order to achieve these sustainability initiatives and technological advancements, significant changes in business practices and project management are needed, including the following:
1. Investment in R&D: Increased funding for research and development to innovate new cooling and power solutions.
2. Collaboration with stakeholders: Engagement with governments, utility providers, and technology partners to create supportive ecosystems.
3. Training and development: Upskilling the workforce to manage and operate advanced data centre technologies.
4. Regulatory compliance: Ensuring all projects meet local and international regulations for sustainability and efficiency.
5. Scalability and flexibility: Designing data centres with modular components to allow for scalability and adaptability to future technologies.
The critical aspects of project delivery cannot be achieved without multi-faceted expertise that combines project management, cost and commercial management, procurement, design, and sustainability. Introducing the right partner to seamlessly bring together global knowledge with local expertise will ensure that delivery meets the demand and that data centres thrive both now and in the future.
But the rapid growth of data centres raises important environmental concerns, posing a critical question to the industry: how can we meet the increasing demand in Asia while scaling up responsibly?
Growing demand
As tech giants develop their AI solutions, data centre construction is growing rapidly in Asia and with the rise of AI, driven by cloud-based platforms and digitisation, the need for safe, secure and resilient data centres is increased.Asia offers new opportunities due to cost-effective land, energy and construction. Driven by the developing countries in the region, the continent also provides a dynamic and flexible landscape for data centre companies to maximise and learn from.
While the cost of building data centres varies worldwide, typically ranging from US$9 million to US$15 million per megawatt, Asia offers more affordable options, with costs around US$8 to US$10 million.
Therefore, countries in Asia are eager to support data centre construction, as it boosts economic growth, accelerates innovation, and enables local businesses to leverage high-performance computing.
However, this growth also poses environmental and community risks. Companies are being challenged to ensure sustainability by implementing solutions for efficient energy consumption, effective heat management, and an integrated approach to designing, building, and operating data centre facilities.
Rapid growth poses environmental risks
AI, especially advanced forms like deep learning, requires significant computational power and energy, posing sustainability challenges. The AI lifecycle, from training to inference, demands both substantial and efficient energy usage.In legacy data centres, power is supplied from diverse local grids at low or medium voltage. Standby generators, transformers, load banks, and cooling system heat rejection are typically located outside the buildings, with underground fuel oil storage tanks. Generators were typically placed on the ground floor or roof, providing enough power for IT equipment and cooling needs, while power distribution was done at low voltage, with two separate circuits for each cabinet.
With the rise of AI installations, power demands have significantly increased. AI cabinets now require much higher power loads and densities, leading to more frequent use of duplicate high voltage feeders and block distribution, reducing the number of cabinets per power block.
Data centres must adapt their power infrastructure to accommodate clusters of AI cabinets. The high utilisation rate of AI processors within these clusters requires careful power infrastructure planning, as well as improvements in IT equipment to enhance temperature tolerance.
Future-proofing data centres
Asia, with its immense potential, is at the forefront of this revolution. AI is fundamentally changing the landscape, requiring bold solutions for designing, building, and operating data centres.As a global leader in this field, Mace is supporting efforts to rethink how data centres are delivered in the region, making them more efficient, resilient, and environmentally sound through initiatives like the following:
1. Renewable energy integration: Using solar, wind, and hydroelectric power to reduce reliance on fossil fuels.
2. Green building certifications: Pursuing LEED certification and other green building standards for sustainable construction practices.
3. Energy efficiency: Implementing advanced cooling solutions like liquid immersion cooling and direct-to-chip cooling to lower energy consumption.
4. Water conservation: Using water-efficient cooling towers and closed-loop cooling systems to minimise water usage.
5. Recycling and waste reduction: Implementing comprehensive recycling programs for electronic waste and construction materials.
However, in order to achieve these sustainability initiatives and technological advancements, significant changes in business practices and project management are needed, including the following:
1. Investment in R&D: Increased funding for research and development to innovate new cooling and power solutions.
2. Collaboration with stakeholders: Engagement with governments, utility providers, and technology partners to create supportive ecosystems.
3. Training and development: Upskilling the workforce to manage and operate advanced data centre technologies.
4. Regulatory compliance: Ensuring all projects meet local and international regulations for sustainability and efficiency.
5. Scalability and flexibility: Designing data centres with modular components to allow for scalability and adaptability to future technologies.
The critical aspects of project delivery cannot be achieved without multi-faceted expertise that combines project management, cost and commercial management, procurement, design, and sustainability. Introducing the right partner to seamlessly bring together global knowledge with local expertise will ensure that delivery meets the demand and that data centres thrive both now and in the future.
