Impact of Data Centers in the Era of Artificial Intelligence

The advent of artificial intelligence (AI) has significantly transformed the landscape of data centers, making them integral to the technological advancements driving our modern world. As AI continues to evolve, its symbiotic relationship with data centers becomes more pronounced, highlighting the importance of these facilities as the backbone of our digital infrastructure.

Computational Power and AI Applications

AI is transforming data processing, analysis, and utilization, leading to exponential growth in computational power requirements. The AI market is projected to reach $407 billion by 2027, with an annual growth rate of 37.3% between 2023 and 2030. The rise of generative AI (Gen AI), such as ChatGPT, is driving this growth, necessitating greater investment and infrastructure for AI-powered solutions.

Data centers serve as the critical infrastructure supporting the AI ecosystem – Marc Garner, SVP Secure Power Europe at Schneider Electric

How AI Pressuring Data Centers

AI technologies require immense computational power and vast amounts of data to function effectively. Data centers, equipped with supercomputers, provide the necessary infrastructure to support these requirements. AI is exerting unprecedented pressure on data centers, driving them to operate at maximum capacity. This section delves into how AI applications are pushing the limits of data center operations, focusing on energy use, computational power, and infrastructure demands.

In terms of data, many people may not realize that the near-instantaneous nature of digital interactions relies heavily on the extraordinary computing capabilities of data centers. In 2022, it was estimated that a staggering 1.145 trillion megabytes (MB) of data were generated daily. Finances Online reports that every minute, WhatsApp users sent 41,666,667 messages, while Facebook users exchanged 150,000 messages and 147,000 images. Additionally, Instagram users posted around 347,222 stories each minute, and Netflix users streamed 404,444 hours of video content every minute.

Rising Demand and Energy & Water Consumption

With the digital transformation accelerating across sectors, power-intensive AI applications are skyrocketing global data service demand. This linkage has accelerated the development and deployment of AI applications across various sectors, from healthcare to finance, and beyond. According to a report by Goldman Sachs, AI is expected to drive a 160% increase in power demand by 2030, underscoring the critical role of data centers in sustaining AI growth. The International Energy Agency estimates that data centers currently account for around 1% of global electricity demand. Schroders Global estimated, by 2030, data center power consumption is expected to reach 35 gigawatts annually, up from 17 gigawatts in 2022.

For cooling alone, data centers require vast amounts of electricity and water for cooling, were not originally designed to handle this continuous influx of data management demand. In 2020, it is estimated that data centers consumed between 196 terawatt hours (TWh) and 400 TWh of electricity—approximately 2% of global electricity consumption—to manage cooling requirements. This energy usage accounted for about 2% of global greenhouse gas emissions.

While data centers play a crucial role in enabling digital advancements that help combat climate change, they also necessitate a transformation driven by innovation. According to a media roundtable on data center sustainability hosted by Vertiv Malaysia, data centers consumed between 200 to 250 TWh of electricity in 2020. Schroders Global estimated, by 2030, data center power consumption is expected to reach 35 gigawatts annually, up from 17 gigawatts in 2022.

Infrastructure Challenges: Power, Cooling, and Racks

AI applications, particularly deep learning algorithms, require substantial computational power to process and analyze vast amounts of data. This drives the need for powerful and efficient hardware, including GPUs and TPUs. High-density racks hosting numerous servers in a compact space elevate power requirements. For instance, training large AI models like GPT-3 consumed 1,287 megawatt hours of electricity and generated 552 tons of CO2. To address these demands, data centers are enhancing power distribution systems and adopting energy-efficient hardware and software. Traditional air-cooling systems are becoming insufficient for high-density setups. The transition to liquid cooling, specifically Direct-to-Chip liquid cooling, is crucial for managing the concentrated heat generated by AI clusters. Liquid cooling enhances processor reliability and performance, improves power utilization, and reduces water usage.

Network Bandwidth Requirements

AI systems rely on large datasets for training and continuous learning, increasing the need for high-speed data transfer within data centers and across networks. Improvements in network infrastructure, such as higher-capacity switches, routers, fiber optic cables, and internal system interconnects, are essential to ensure efficient data movement.

Hardware Upgrades – Edge Computing and Distributed AI

AI’s growth drives the need for real-time, low-latency processing. Edge computing, where AI computations occur closer to the data source, reduces latency and bandwidth requirements. This shift necessitates developing distributed AI architectures and integrating AI capabilities into edge devices and networks. With AI progression, hardware components must evolve in parallel. This necessitates frequent upgrades of chips, hard disks, fiber optics, and other components to cater to AI’s computational and storage needs. Data centers must continuously invest in newer, more powerful hardware to stay ahead of AI demands.

Malaysia’s Strategic Move Towards Data Centers

NIMP is launched by Prime Minister of Malaysia

Recognizing the potential of data centers, Southeast Asia has seen a surge in investment in this sector. According to recent reports, green data centers have significantly boosted green investment in the region, attracting as much as USD 6.3 billion. Countries like Singapore and Malaysia are leading the charge with energy efficiency regulations and incentives that encourage the development of sustainable data centers.

Malaysia has incorporated data center development into its national strategy under the New Industrial Master Plan (NIMP) 2030. Based on MDEC database (as of June 2024), currently, Malaysia hosts 18 data centers, with several more under construction, particularly in Johor. This strategic push aims to position Malaysia as a key player in the global data center market, attracting significant investments from tech giants like Google, which recently announced a $2 billion investment in Malaysia’s data center and cloud market.

Malaysia’s Emergence as a Data Center Hub

Over the past decade, Malaysia has positioned itself as a burgeoning hub for data centers in Asia, driven by a confluence of strategic advantages and proactive government policies. As highlighted by CNBC, Malaysia’s ascension in the data center industry is fueled by its strategic location, abundant natural resources, and a supportive business environment . The Malaysian government has been instrumental in this growth, implementing policies that attract foreign investments and foster technological advancements .

The development of data centers in Malaysia is expected to create numerous economic opportunities. It will attract foreign direct investment, create high-value jobs, and stimulate the local economy. Moreover, it will foster technological advancements and innovation, positioning Malaysia as a hub for digital transformation in the region.

Malaysia’s data center landscape is thriving, with significant investments from global technology giants and local enterprises alike. According to The Edge Malaysia, the country is home to numerous data centers with a combined capacity that continues to expand . Major players such as Microsoft, Google, and Amazon Web Services (AWS) have established a strong presence, contributing to the robust growth of the sector .

Recent data from the government underscores this expansion, with substantial investments pouring into the sector, reflecting a buoyant market outlook . For instance, KKD reported a surge in data center construction and expansion projects, underscoring the burgeoning demand for digital infrastructure .

Driving Growth and Future Prospects

One of the key factor in propelling the growth of data centers in Malaysia is the economic incentives and government policies. The Malaysian government has introduced a suite of incentives to attract data center investments, including tax breaks, grants, and streamlined regulatory processes. Other than NIMP mentioned before, The National Investment Aspirations (NIA) and the National Digital Economy Blueprint (MyDIGITAL) are pivotal frameworks that have bolstered the sector .

Continuous improvements in technology and infrastructure have been critical in supporting the data center boom. Investments in high-speed internet connectivity, robust power supply, and cooling technologies are essential for the efficient operation of data centers

Looking ahead, Malaysia’s data center industry is poised for continued growth and innovation. Projections indicate a robust expansion over the next decade, driven by advancements in artificial intelligence (AI), edge computing, and the Internet of Things (IoT). These technologies are expected to revolutionize data center operations, enhancing efficiency and enabling new capabilities .

Sustainable Development of Data Centers

The rapid expansion of data centers brings with it the challenge of ensuring they are built and operated sustainably. The construction and operation of data centers can have substantial environmental impacts, primarily due to their high energy consumption and the need for robust cooling systems. As discussed before, the data center industry is under increasing pressure to reduce its carbon footprint and embrace green initiatives.

What is a Green Data Center?

A green data center is an environmentally friendly facility that houses computer systems and associated components, such as telecommunications and storage systems. The primary objective of green data centers is to reduce environmental impact through energy-efficient technologies, sustainable design practices, and optimized operational procedures. However, a comprehensive approach to green data centers goes beyond energy efficiency, encompassing the entire lifecycle of the facility from construction to operation and maintenance.

The previous section discussed on the impacts of data center, environmentally. To mitigate these impacts, data centers should adhere to stringent sustainability standards, such as LEED (Leadership in Energy and Environmental Design), EDGE (Excellence in Design for Greater Efficiencies), GBI (Green Building Index), MyCREST (Malaysian Carbon Reduction and Environmental Sustainability Tool), GreenRE and others.

Implementing Green Standards – Beyond Green Building Certification

Adopting these standards helps ensure that data centers minimize their carbon footprint, utilize renewable energy sources, and implement efficient cooling and power management systems.

Some Considerations and Criteria

Energy Efficiency: At the core of green data centers is the emphasis on energy efficiency. Techniques such as using advanced cooling technologies, optimizing server utilization, and integrating renewable energy sources are standard practices. According to IBM, green data centers employ various strategies to reduce energy consumption, including the use of energy-efficient hardware and infrastructure, smart grid technologies, and dynamic power management systems .
Sustainable Design: Green data centers are designed with sustainability in mind from the outset. This includes the use of sustainable building materials, efficient architectural designs that maximize natural lighting and ventilation, and landscaping that reduces the heat island effect.
Water Usage: Water usage is a critical aspect often overlooked in discussions about green data centers. Efficient water management systems, such as rainwater harvesting and water recycling, are integral to reducing the environmental footprint. These systems ensure that water is used efficiently for cooling and other operational needs, minimizing wastage and reducing the overall water footprint of the facility .
Waste Management: Effective waste management practices, including recycling programs and the responsible disposal of electronic waste, are essential components of green data centers. Facilities are designed to minimize waste production and promote the reuse and recycling of materials wherever possible.
Renewable Energy Integration: Incorporating renewable energy sources, such as solar, wind, or geothermal energy, into the power supply of data centers is a significant step towards sustainability. Facilities like YTL Power International Berhad’s data center integrate renewable energy sources to achieve superior energy efficiency and reduce dependency on energy from the grid.

Beyond Design and Construction – Operations and Maintenance

Adopting green data centers goes beyond just design and construction; it necessitates a comprehensive approach that includes efficient and sustainable operations and maintenance. While the initial design and construction phases are crucial for establishing a sustainable foundation, the long-term impact of a green data center is largely determined by how it is operated and maintained. Ensuring that these facilities continue to perform efficiently requires continuous monitoring, optimization, and adherence to best practices in energy management, cooling, and waste reduction throughout their operational lifespan.

Operational Efficiency: Sustainable operations involve continuous monitoring and optimization of the data center’s performance. Advanced monitoring systems track energy usage, cooling efficiency, and overall performance metrics in real time. This data is used to make informed decisions about operational adjustments and improvements.
Maintenance Practices: Green maintenance practices ensure that the data center operates efficiently throughout its lifespan. This includes regular maintenance of HVAC systems, use of environmentally friendly cleaning products, and adherence to best practices for equipment upkeep. Preventive maintenance schedules and proactive equipment management help in reducing energy consumption and extending the lifespan of the data center infrastructure

Hong Leong Bank receiving the Titanium+ Tier Certification from Green Climate Initiative’s (GCI) – Source: HLB Website

Innovations in Sustainable Technology

In addition to green certifications, innovations in sustainable technologies are being implemented in data centers. These include advanced cooling techniques like liquid cooling and free cooling, which reduce energy consumption significantly. Renewable energy sources, such as solar and wind power, are also being integrated to power data centers, further reducing their environmental impact .

Case Studies and Success Stories

Several organizations have made notable strides in developing green data centers. For instance, Hong Leong Bank’s new data center received the highest green certification in from the Green Computing Initiative (GCI), demonstrating the bank’s commitment to sustainability . Similarly, YTL Power International Berhad’s data center project highlights the integration of renewable energy sources and cutting-edge cooling technologies to achieve superior energy efficiency . Delta Power Solutions’ LEED-certified green data center serves as another exemplary model, showcasing how advanced design and operational strategies can create environmentally sustainable IT infrastructures.

The Importance of Technical Competency

As the previous sections discussed, with the rapid expansion of data centers in Malaysia, the focus on green and sustainable development has become a paramount priority. As the nation positions itself as a burgeoning data center hub, the question arises: Do we have the necessary manpower and technical competency to maintain these sophisticated, eco-friendly facilities?

Do we have the necessary manpower and technical competency to maintain these sophisticated, eco-friendly facilities?

The Importance of Technical Competency

Building and physical aspects of sustainable data centers are only part of the equation. Maintaining these facilities requires a workforce equipped with specialized skills in green data center maintenance. This includes knowledge of energy-efficient technologies, renewable energy integration, and advanced cooling systems, which are distinct from the maintenance of conventional commercial buildings.

Current State of TVET in Malaysia

However, Malaysia’s current technical and vocational education training (TVET) programs do not specifically address the unique requirements of green data center maintenance. This gap highlights the need for curriculum development to include training on sustainable data center operations, ensuring that the local workforce is prepared to support the country’s ambitions in this sector.

Students from KKTM Sri Gading conducted Energy Audit as part of its TVET course (Source: Facebook – Green Building KKTM Sri Gading)

Addressing the Skills Gap

To address this skills gap, collaboration between industry stakeholders, educational institutions, and the government is essential. Programs like the Data Centre Certified Technician course provide a foundation, but more specialized training focused on sustainable practices is needed. The expansion of TVET programs to include modules on green data center maintenance can equip the workforce with the necessary skills to manage and operate these facilities efficiently.

Industry Engagement and TVET Expansion

The Malaysian government and industry players are aware of this necessity. According to a report from the International Labour Organization, digital transformation and technical and vocational education and training (TVET) are critical in addressing these emerging needs . Moreover, the National Institute of Technology (NIT) in Malaysia has been advancing industry engagement in TVET expansion, promoting the concept of teaching factories where students can gain hands-on experience in real-world scenarios . This approach is pivotal in ensuring that graduates are job-ready and possess the competencies required by the industry.

Collaborative Efforts for Curriculum Development

The collaboration between government bodies, educational institutions, and the private sector is crucial for developing and updating TVET curricula. This collaborative effort will ensure that the curriculum includes comprehensive training on sustainable data center operations and maintenance. By aligning TVET programs with the industry’s needs, Malaysia can cultivate a workforce that is not only competent but also adept at managing green data centers.

Conclusion

As we stand on the precipice of a new era driven by artificial intelligence, the role of data centers has never been more crucial. These facilities are not merely the backbone of our digital infrastructure; they are the powerhouses that fuel the rapid advancements in AI technologies. The symbiotic relationship between AI and data centers underscores the immense computational power and sophisticated infrastructure required to support the ever-growing demand for data processing and analysis.

The exponential rise in AI applications, from generative AI to deep learning algorithms, has placed unprecedented pressure on data centers, necessitating continuous innovation and investment in power, cooling, and network bandwidth capabilities. The staggering data generation and the subsequent demand for real-time processing further highlight the critical role data centers play in our digital lives. As power consumption and cooling needs soar, the challenge of maintaining sustainable and green data centers becomes more pronounced.

Malaysia’s strategic initiatives have position the country as a burgeoning hub for data centers in Southeast Asia. The nation’s commitment to sustainable development, supported by significant investments and government policies, is paving the way for a thriving data center industry that aligns with global sustainability standards. However, the journey towards sustainable and green data centers is complex and requires a multifaceted approach.

To meet the environmental challenges posed by data centers, it is imperative to adopt green building certifications, integrate renewable energy sources, and employ advanced cooling technologies. Additionally, the focus must extend beyond construction to encompass efficient operations and maintenance practices, ensuring long-term sustainability.

The success of Malaysia’s data center industry hinges not only on infrastructure and technological advancements but also on the technical competency of its workforce. Bridging the skills gap in green data center maintenance through enhanced technical and vocational education training (TVET) programs is essential. Collaboration between industry stakeholders, educational institutions, and the government will be key in developing a workforce adept at managing sustainable data centers.

In conclusion, the future of data centers in the era of artificial intelligence is both promising and challenging. As AI continues to evolve, data centers must adapt and innovate to meet the demands of this transformative technology. Malaysia’s proactive approach and commitment to sustainability, coupled with a skilled workforce, will ensure that it remains at the forefront of the global data center industry, driving digital transformation while safeguarding the environment for future generations.

Remark:

Thank you to Ts. Mohd Mawardi bin Hussain for his insights for TVET on maintenance of green building.

References:

IIRAI