China accelerates data center expansion amid growing Western pushback
Read more
TechNode
technode.com

China accelerates data center expansion amid growing Western pushback

As the global artificial intelligence industry strives to secure computing power, the construction of data centers is facing a new set of challenges. Starting in the second half of 2026, resistance to data center projects has emerged in various parts of Europe and the United States. Rising electricity costs, local community protests, and disputes over resource allocation have led to the halt of some projects.

In contrast, China continues to promote the expansion of its computing infrastructure, following a strategy based on national computing nodes and coordinated growth between eastern and western regions.

An analysis of China's computing capacity demonstrates the scale of this expansion. As of June 2026, China operated 15.56 million standard data center racks, and its intelligent computing power reached 2185 EFLOPS, representing a 177% increase compared to the previous year, making China the second-largest market by this metric.

The rapid growth in computing demand is closely linked to the boom in the large language model (LLM) industry in China. Over the past two years, daily usage of domestic LLM tokens has increased by more than 1000 times. As AI applications spread to areas such as AI agents, enterprise services, and industrial applications, the demand for computing infrastructure is entering a new phase of growth.

In the first seven months of 2026, at least 158 data center and intelligent computing center projects were registered nationwide, winning tenders worth over 100 million yuan ($15 million USD) each. Their total value exceeded 113 billion yuan ($17 billion USD). Furthermore, many other projects are in the planning, application, and construction stages, and investments in the sector continue to grow.

From a regional distribution perspective, eight national computing center nodes remain the main centers for China's computing infrastructure development. Cities and regions, including Junwei in Ningxia, Hohhot and Ulanqab in Inner Mongolia, Qingyan in Gansu, and Shaoguan in Guangdong, continue to implement data center projects. Notably, computing infrastructure is also expanding into smaller counties. Smaller computing facilities have started operations in places like Tuoli in Xinjiang and Guannan in Yunnan. This signifies the gradual formation of a nationwide computing network that connects major nodes with regional points.

However, the large number of projects does not mean that all planned computing capacity has been commissioned. In the first quarter of 2026, investments in intelligent computing infrastructure reached 728.8 billion yuan ($109.2 million USD). Nevertheless, only 13.6% of projects have started operations, while projects in the planning and preparation stages accounted for 71.6%. This indicates that most of the investment remains frozen in projects awaiting construction, equipment deployment, and final handover.

Telecommunications operators remain the dominant force in building data centers in China. They secured more than half of the major tenders worth at least 100 million yuan ($15 million USD) in the first seven months of the year. At the same time, third-party Internet Data Center (IDC) providers and state energy companies are also increasing their participation.

Given the current project portfolio and construction pace, 2027 may be a crucial period for commissioning new computing infrastructure in China.

One of the most serious problems facing the global data center industry is the need to balance growing computing demand with limited resources of energy, land, and local population. In parts of Europe and the US, opposition to data center projects largely stems from an imbalance between costs and benefits. Data centers consume enormous amounts of electricity but create relatively few long-term jobs. When local energy infrastructure cannot expand fast enough to meet additional demand, the resulting pressure can lead to higher electricity prices for residents, while economic benefits may not be distributed among surrounding communities.

Ireland serves as a clear example. In 2025, data centers accounted for 23% of national electricity consumption. In Finland, it is estimated that if construction continues at the current pace, wholesale electricity prices could rise by approximately 31% by 2030, potentially adding nearly 400 euros to the annual household electricity bill.

China's approach differs in that it began coordinating the development of computing infrastructure at a national level relatively early. The Chinese 'Eastern Data, Western Computing' initiative, launched in 2022, established eight national computing center nodes and 10 national data center clusters. The plan aims to redistribute some computing demand from Eastern China to western regions where land, electricity, and renewable energy resources are more available.

This strategy is not just about relocating data centers to the west. It also seeks to coordinate computing demand, power supply, and regional economic development. Ulanqab in Inner Mongolia is an example. More than 80 data centers have been built in the city, and 109 computing projects have been signed. Intelligent computing accounts for over 90% of the city's total computing capacity. In addition to building facilities and deploying servers, these projects have also created local jobs and supported the growth of nearby retail and other businesses. Simultaneously, the abundant wind and solar resources in western China provide opportunities for using green electricity for data centers locally, helping to reduce energy costs and lessen the burden on traditional power sources.

Global demand for AI computing power continues to grow, but the criteria for competition in the data center industry are changing. In the past, companies mainly focused on land and electricity costs. Today, energy supply stability, local infrastructure capacity, community acceptance, and the long-term economic benefits provided to host regions have become important factors in determining whether projects can move forward.

Thanks to early planning of national computing nodes, closer coordination between computing and electrical infrastructure, and stronger ties between eastern and western regions, China has formed a relatively clear development strategy. However, as more projects move into construction and commissioning, the industry still faces challenges such as preventing overinvestment, improving computing utilization, and achieving more environmentally friendly and low-cost operations.

Ultimately, the next stage of competition in computing will involve more than just chips, servers, and the sheer scale of data centers. It will increasingly depend on the ability to integrate energy resources and coordinate computing power. The real competitive advantage may not lie in who has more data centers, but in who can provide computing power more reliably, efficiently, and with fewer resource costs.

Similar stories

According to Crisil, data center capacity in India will increase fivefold by 2030
Read more
yourstory.com

According to Crisil, data center capacity in India will increase fivefold by 2030

According to a report published by Crisil Ratings on Tuesday, India's data center industry is preparing for an unprecedented investment boom. The installed capacity is expected to increase fivefold, reaching approximately 7.5 gigawatts (GW) by March 2030. This will attract investments exceeding 3 lakh crore rupees over the next four fiscal years.

This expansion is driven by generative artificial intelligence (GenAI), enterprise digitalization, and growing data consumption, which is set to transform the country's digital infrastructure landscape.

However, the industry's ability to translate demand into sustainable profit will critically depend on timely grid connectivity, permit acquisition, and disciplined project execution. Crisil Ratings notes that nearly half of the upcoming capacity is already secured through long-term contracts, which limits speculative development and reduces sales risk.

Preliminary agreements with hyperscalers, driven by the adoption of cloud technologies and computationally intensive GenAI workloads, provide revenue certainty from the outset, helping mitigate project risks despite increasing execution demands due to the scale and speed of planned construction.

Strong demand from global hyperscalers, fueled by cloud adoption and increasingly demanding GenAI tasks, also supports investment decisions and reduces sales uncertainty. Furthermore, India's competitive cost structure enhances its attractiveness as a data center destination.

Development costs here are approximately 20 percent lower than in several competing markets in the Asia-Pacific region, reflecting relatively competitive costs for land, construction, and labor. Combined with lower electricity tariffs, this allows for rental rates 30–35 percent lower than in comparable markets. The availability of large campuses powered by renewable energy also strengthens India's position for hyperscalers aiming to meet sustainability commitments.

Anand Kulkarni, Director at Crisil Ratings, emphasized that India's structural advantages further enhance its appeal as a location for data centers. He noted that development costs are about one-fifth lower than in some competing Asia-Pacific markets due to more competitive costs for land, construction, and labor.

He added that, combined with lower electricity tariffs, this results in rental rates 30–35 percent lower than in comparable markets. The ability to build large campuses powered by renewable energy aligns well with hyperscalers' sustainability goals, thereby strengthening India's competitive standing.

However, the scale of the planned expansion places significant pressure on energy infrastructure. Crisil reported that nearly three-quarters of the projected 6 GW of upcoming capacity is expected in Visakhapatnam, Mumbai, and Hyderabad. By the 2030 fiscal year, data center capacity in these cities could account for approximately 80%, 60%, and 20% of their current peak electricity demand, respectively. This concentration makes timely capacity upgrades and grid connection critical to prevent deployment delays and cost overruns.

Companies need a combination of resilience and intelligence to develop in a fragmented world
Read more
www.khaleejtimes.com

Companies need a combination of resilience and intelligence to develop in a fragmented world

Enterprises in the Middle East, Central Asia, and Africa (MECAA) are entering a new phase of digital growth. Cloud technologies are being actively implemented, Artificial Intelligence (AI) is becoming a priority at the board level, and digital services are rapidly expanding across all sectors. Simultaneously, the operating environment is becoming more complex due to geopolitical tensions, changing trade relations, and evolving data regulation norms.

This creates new expectations: organizations can no longer focus solely on innovation. They must also ensure stability, security, and control over operations even amid external uncertainty. In this context, resilience and AI merge into a single strategy: one ensures continuity, the other drives growth, and neither can exist in isolation.

The influence of geopolitics on technology has become more apparent in recent years. Requirements for data localization are tightening in several markets, control over cross-border data transfer is increasing, and dependence on global platforms and supply chains is being re-evaluated. For businesses, this means that infrastructure solutions are no longer purely technical; they have taken on strategic significance. Organizations must clearly know where their data is, how it is protected, and who controls it. They must also ensure the ability to continue operating regardless of disruptions caused by regulatory changes, geopolitical events, or supply chain constraints.

This shift is particularly relevant in regions where digital growth occurs alongside policy development. Businesses are developing rapidly, but they are also expected to comply with regulations, ensure security, and maintain accountability. As a result, resilience takes on a broader meaning—it is no longer just about recovering from failures, but about maintaining control, continuity, and confidence amidst growing unpredictability.

Disaster Recovery as a Service (DRaaS)

In this situation, the role of Disaster Recovery as a Service (DRaaS) is increasing. It is no longer viewed merely as a backup solution activated when problems arise. Instead, it is integrated into a broader approach to business continuity, data protection, and regulatory compliance. In a world of distributed multi-cloud environments, enterprises need the ability to quickly restore systems, while ensuring that recovery processes align with modern workloads, including real-time data-dependent applications, cross-border operations, and increasingly, AI.

Downtime now has broader implications: it affects not only operational activities but also compliance obligations and customer trust. In sectors such as financial services, telecommunications, and public services, these consequences are even more pronounced. This is why resilience is moving to the core of corporate strategy, enabling organizations to act confidently even when conditions around them change.

Experience and regional scale are becoming equally important components of enterprise resilience. In the MECAA region, organizations are increasingly seeking partners who understand local regulatory requirements, changing data policies, and sector-specific operating conditions. Amid growing global fragmentation, resilience is becoming a matter of sovereignty as much as it is a technological one. Businesses are assessing not only system recovery speed but also the location of critical data, who manages it, and how to continue operating during geopolitical, regulatory, or supply chain disruptions.

Concurrently, businesses across the region are accelerating the adoption of AI. Governments and businesses are investing in AI to improve efficiency, enhance services, and support economic diversification. However, scaling AI presents its own challenges. AI workloads require significant resources and depend on data. They need constant uptime, reliable data channels, and robust governance. Any infrastructure failure can directly impact outcomes, whether it is model performance, decision accuracy, or operational continuity.

There is also growing attention on how AI is deployed. Data sovereignty, privacy, and regulatory compliance are becoming critical aspects. Enterprises must ensure that AI systems are not only efficient but also trustworthy. In response to this need, cloud AI platforms are emerging, consolidating computing power, data, and AI tools in a managed environment designed to transition organizations from fragmented experiments to predictable, industrial-grade results.

It is becoming clear that resilience and AI cannot be managed in isolation. The more enterprises rely on AI, the more dependent they become on the stability and security of their underlying systems. If an AI platform fails, the consequences go beyond mere downtime—they affect analytics, decision-making, and business operations. If the data used for training or inference is compromised or lost, the reliability of the entire system can suffer.

Furthermore, traditional approaches to resilience may not fully cover the scale and complexity of AI environments. Large datasets, distributed workloads, and continuous processing require more advanced recovery and protection mechanisms. Here, alignment between DRaaS and cloud AI platforms is crucial. Together, they create a foundation where innovation can scale without increasing risk. DRaaS guarantees fast and reliable recovery of critical applications and data, while the AI cloud provides an environment for developing and deploying intelligence at scale. Integrating these capabilities allows enterprises to adopt innovations with confidence, knowing that their operations remain protected.

This integrated approach is changing how enterprises view resilience and AI. It is not about preparing for failure on one side and stimulating innovation on the other. The goal is to create an environment where both aspects can coexist. Organizations can continue to operate during outages while simultaneously scaling AI initiatives with confidence. This is especially important in regions where regulatory expectations and geopolitical dynamics continue to shift.

The integrated approach also simplifies operations. Instead of managing separate systems for recovery and AI, organizations can align their strategies, reduce complexity, and focus on achieving business outcomes. More importantly, it strengthens executive-level trust. Technical decisions are made with a clearer understanding of risks, control, and long-term value.

As enterprises in the MECAA region navigate this landscape, several points stand out. They must view resilience as a strategic opportunity that supports not only traditional applications but also modern, data-intensive workloads. It is necessary to ensure that AI platforms are designed with governance, security, and continuity in mind from the outset. Furthermore, infrastructure solutions should be aligned with broader business and regulatory priorities, including data sovereignty, compliance, and operational control. Finally, there is a need to move towards simpler, more integrated architectures. Combining resilience and AI into a single structure reduces risks, increases flexibility, and ensures greater long-term adaptability.

The global technology landscape is defined not only by stability; it is shaped by change, uncertainty, and rapid innovation. In such an environment, enterprises must balance ambition with assurance. Resilience and AI are not competing priorities, but complementary capabilities. One builds the foundation, and the other drives the future. Organizations that recognize this connection will be better prepared to overcome failures while continuing to innovate. They will move forward with greater clarity, confidence, and control, even in a fragmented world.

In the MECAA region, enterprise technology solutions are now shaped by a combination of the drive for growth and operational risks. Regulatory expectations are becoming clearer, data policies are evolving, and geopolitical factors continue to influence infrastructure design and management. In this environment, organizations are looking not only for cutting-edge technological opportunities but also for partners capable of ensuring stability, accountability, and long-term direction. Tata Communications collaborates closely with enterprises across the region, combining local engagement with global capabilities in cloud technologies, networks, and infrastructure. With over 16 years of regional presence and more than 300 clients in the MECAA region, Tata Communications combines expertise in networking, cloud technologies, security, and five broader operational structures integrated to help enterprises navigate increasingly complex technological and regulatory conditions.

Supporting enterprises in the UAE, Saudi Arabia, Qatar, Oman, and Kuwait has given Tata Communications direct insight into how resilience requirements differ across various industries. From retail and telecommunications to airlines, real estate, healthcare, and financial services, a common trend is becoming increasingly evident: the success of digital transformation initiatives now depends as much on operational resilience, governance, and continuity as it does on the implementation of technology itself. As organizations expand cloud, AI, and resource-intensive workloads, uninterrupted operation, cyber resilience, and governance are becoming board-level priorities alongside innovation and growth.

As a global cloud provider, Tata Communications integrates DRaaS, hybrid and multi-cloud platforms, managed infrastructure, security, and network capabilities into a unified operational environment. Instead of viewing resilience, connectivity, and cloud transformation as separate initiatives, organizations can align them as part of a holistic strategy supporting both operational continuity and long-term innovation. This allows organizations to advance transformation programs while maintaining clear control over data, operations, and compliance requirements.

Resilience is built into this approach at the platform level. Cloud environments unify infrastructure, managed services, and disaster recovery into a single operational model. Enterprises can deploy workloads on dedicated or hybrid architectures that provide stable performance, stronger isolation, and greater transparency across different environments. The same foundation is designed to support AI expansion. Through integrated cloud capabilities, organizations can access development tools, data platforms, and operational services via a unified structure. This accelerates the transition from experiments to production results while maintaining governance, compliance, and operational control. By combining resilience and AI capabilities, enterprises can manage infrastructure, data, and intelligence as part of one coordinated system, rather than as separate layers. This enables innovation at scale without introducing unnecessary complexity or risk. Ultimately, the conversation is no longer just about infrastructure; it is about creating the conditions for enterprises to innovate with confidence, operate under control, and remain resilient in a world where disruption has become a constant reality.

Popular