Global Data Centre Boom: Could Electricity Demand Become the Next Technology Crisis

The global technology industry is entering an era in which computing power is becoming as important as physical infrastructure. Artificial intelligence, cloud computing, digital payments, online entertainment, autonomous systems and connected devices are generating enormous volumes of data that must be processed, stored and transferred continuously. Behind these digital services are data centres: large facilities containing servers, networking equipment, storage systems and cooling infrastructure.

For years, the expansion of data centres was primarily viewed as a technology and investment story. Companies competed to build larger facilities, develop faster processors and increase computing capacity. However, the rapid adoption of generative artificial intelligence and high-performance computing has introduced another major consideration: electricity availability.

Modern data centres require continuous power to operate thousands of servers and maintain stable temperatures. AI-focused facilities can consume particularly large amounts of electricity because advanced processors operate at high utilisation levels and generate substantial heat. As technology companies accelerate investment in AI infrastructure, electricity demand is becoming an increasingly important factor in determining where new facilities can be constructed and how quickly they can become operational.

The International Energy Agency (IEA) estimates that global data centre electricity consumption reached approximately 485 terawatt-hours (TWh) in 2025. Its updated outlook projects that consumption could rise to around 950 TWh by 2030, nearly doubling within five years. This growth raises an important question: could electricity availability become the next major constraint on technological progress?

Why the Global Data Centre Boom Is Accelerating

The expansion of data centres is closely connected to the increasing dependence of businesses and individuals on digital services. Cloud platforms allow organisations to store information and run applications without maintaining their own computing infrastructure. Streaming services require large-scale storage and content delivery networks, while financial institutions depend on data centres for payment processing, transaction management and cybersecurity.

Artificial intelligence has significantly accelerated this demand. Training advanced AI models requires large computing clusters containing specialised processors, high-speed networking equipment and extensive memory capacity. Once these models are deployed, they continue consuming electricity whenever users submit queries, generate content, analyse information or interact with AI-powered applications.

The growth of AI agents and automated business systems may create an additional source of demand. Unlike a simple search query, an AI agent may perform several connected tasks, access external information, execute software operations and repeatedly process information before producing a result. These activities can require more computing resources than conventional digital interactions.

Investment trends reflect this growing demand. According to the IEA’s 2026 analysis, capital expenditure by five major technology companies exceeded $400 billion in 2025 and was expected to increase substantially in 2026. The same analysis found that electricity consumption by data centres increased by approximately 17% in 2025, significantly faster than the growth of global electricity demand overall.

This investment is encouraging the construction of hyperscale data centres and specialised AI facilities. However, the expansion of computing infrastructure is now increasingly dependent on the availability of power, grid connections, cooling systems and supporting electrical equipment.

Why AI Data Centres Consume So Much Electricity

A data centre does not consume electricity only when users access websites or applications. Its infrastructure must remain operational continuously, including during periods of relatively low activity. Servers, storage systems, networking equipment, backup systems and cooling infrastructure all contribute to its total energy requirements.

Computing equipment is one of the largest sources of electricity consumption. AI workloads can be particularly demanding because graphics processing units and other specialised accelerators perform large numbers of calculations simultaneously. Training an advanced model may require thousands of processors operating together for extended periods.

Cooling is another major requirement. Servers generate heat while processing information, and excessive temperatures can reduce performance, damage components or interrupt operations. Data centres therefore use cooling systems to maintain suitable operating conditions. Depending on facility design and climate, this may involve air cooling, liquid cooling, chillers, pumps and heat-exchange systems.

Power conversion and distribution also create additional energy requirements. Electricity received from the grid must be converted and distributed at appropriate voltages before reaching servers. Uninterruptible power supplies and backup systems are used to protect critical workloads against interruptions.

The efficiency of these operations is commonly measured through Power Usage Effectiveness, or PUE. It compares the total electricity consumed by a data centre with the electricity used by its IT equipment. A lower PUE indicates that a smaller proportion of energy is being consumed by supporting infrastructure such as cooling and power distribution.

However, improving PUE alone cannot eliminate the electricity challenge. If computing demand grows faster than efficiency improves, total energy consumption can continue rising even when individual servers and cooling systems become more efficient.

Could Electricity Demand Become a Global Technology Bottleneck?

The central concern is that the construction of new data centres may move faster than the expansion of electricity generation and transmission infrastructure. Technology companies can invest substantial capital in servers and buildings, but those facilities cannot operate without reliable power connections.

The IEA’s analysis suggests that global data centre electricity consumption could approach 950 TWh by 2030, equivalent to approximately 3% of projected global electricity demand. Although this remains a relatively small share of worldwide electricity consumption, the impact is much greater in locations where large numbers of data centres are concentrated.

This geographical concentration is important. Several data centre projects may compete for electricity within the same regional grid. Even if a country has sufficient electricity generation capacity overall, local transmission networks may not have the capacity to deliver the required power to a particular industrial cluster.

The result can be delayed grid connections, higher infrastructure costs and restrictions on the timing of new projects. The IEA has warned that electricity grid constraints could put a significant proportion of planned data centre projects at risk of delays unless these issues are addressed.

Such delays could affect the technology industry directly. Companies may have to postpone the launch of new AI services, move computing workloads to other regions or invest in dedicated energy infrastructure. Electricity availability could therefore become a competitive factor in determining which countries attract the next generation of digital investment.

The Pressure on Electricity Grids and Transmission Infrastructure

Electricity grids were traditionally designed around established patterns of residential, commercial and industrial consumption. Data centres introduce a different type of demand because they operate continuously and can require large amounts of power within relatively small geographical areas.

A large AI facility may require electricity comparable to that used by a substantial industrial installation. When several such facilities are developed within the same region, utilities must assess whether existing substations, transformers and transmission lines can support the additional load.

Transformers and other grid equipment are particularly important. Even when sufficient electricity can be generated, limited transformer manufacturing capacity or delayed equipment delivery can slow the connection of new facilities. Grid upgrades also require land acquisition, regulatory approvals, engineering work and coordination between multiple agencies.

Another concern is the changing pattern of electricity consumption. AI workloads can produce rapid variations in power demand as computing tasks are distributed across processors. Electricity systems must maintain a stable balance between supply and demand, making forecasting, energy storage and flexible consumption increasingly important.

The challenge is therefore not simply to build more power plants. Countries must also modernise transmission networks, strengthen distribution infrastructure, improve grid monitoring and introduce mechanisms that allow large consumers to manage their electricity requirements more efficiently.

Renewable Energy: Can Clean Power Support the Data Centre Boom?

Technology companies are increasingly seeking renewable electricity to reduce the environmental impact of their data centre operations. Solar and wind energy can support this objective, particularly when combined with long-term power purchase agreements and energy storage.

Renewable energy has several advantages. Solar and wind projects can be developed in many regions, operating costs after construction are relatively predictable, and large technology companies can support new generation capacity through long-term electricity contracts.

However, renewable generation is variable. Solar power depends on daylight and weather conditions, while wind generation changes according to local wind patterns. Data centres, on the other hand, require reliable electricity throughout the day and night.

This creates a need for energy storage, grid balancing and complementary generation sources. Battery storage can help manage short-term fluctuations, while pumped-storage hydropower can support longer-duration electricity balancing. Nuclear energy, hydropower, geothermal power and other dispatchable sources may also contribute to reliable low-carbon electricity supply.

The IEA’s analysis suggests that renewables will meet a substantial share of additional data centre electricity demand through 2030, supported by storage and wider electricity networks. Natural gas and nuclear power are also expected to contribute to the generation mix.

This means that the sustainability of data centres will depend not only on how much renewable electricity companies purchase, but also on how electricity is physically generated and delivered when their facilities require it.

The Water and Cooling Challenge

Electricity is not the only resource affected by data centre expansion. Cooling systems can also place pressure on local water supplies, particularly in regions that experience water scarcity or high seasonal temperatures.

Some facilities use water-based cooling systems because water can transfer heat efficiently. Depending on the technology, location and operating conditions, water may be consumed through evaporation or used within cooling systems that require regular replenishment.

The water footprint of a data centre varies considerably. Facilities using air-based cooling, direct-to-chip liquid cooling, immersion cooling or closed-loop systems can have different water requirements. Local climate and the temperature at which cooling equipment operates also influence consumption.

The challenge becomes more serious when large facilities are constructed in water-stressed regions. A data centre may bring investment, employment and digital infrastructure, but its resource requirements must be evaluated alongside the needs of nearby households, agriculture and industries.

Companies are increasingly exploring more efficient cooling approaches. Direct-to-chip liquid cooling transfers heat away from processors through liquid systems, while immersion cooling places electronic equipment in specially designed cooling fluids. These technologies can support higher computing density and reduce certain cooling requirements, although their overall environmental performance depends on system design.

Sustainable data centre development must therefore consider electricity consumption, water availability, heat management and local environmental conditions together rather than treating them as separate issues.

India: A Growing Data Centre Market With Rising Power Requirements

India is becoming an increasingly important destination for data centre investment because of its expanding digital economy, cloud adoption, AI development and growing demand for secure data infrastructure. Digital payments, online government services, e-commerce, financial technology and enterprise software all require reliable computing and storage facilities.

The country’s installed data centre capacity has expanded considerably. According to the Government of India’s August 2026 update, installed capacity increased from approximately 375 megawatts (MW) in 2020 to around 1,575 MW. The government also cited a Central Electricity Authority estimate that data centre power demand could reach approximately 17 gigawatts (GW) by 2031–32.

Another government estimate presented in March 2026 placed expected electricity demand from data centres at 13.56 GW by 2031–32. These estimates reflect different assessment points and should not be treated as identical forecasts. Their common message, however, is that data centre electricity requirements are expected to increase substantially.

India’s major data centre locations include Mumbai, Navi Mumbai, Chennai, Hyderabad, Bengaluru, Delhi NCR and Jamnagar. New investment destinations are also emerging as states compete to attract digital infrastructure projects.

The expansion creates economic opportunities, but it also requires careful electricity planning. Data centres must be connected to reliable power systems, and their demand must be incorporated into regional generation, transmission and distribution plans.

Why India Could Face a Localised Electricity Challenge

India’s overall electricity generation capacity has expanded, but national capacity figures do not automatically guarantee that every region can supply large new data centre facilities without infrastructure upgrades.

The location of generation resources and data centre demand may not always match. Renewable electricity may be generated in one state while a major data centre is located in another. Transmission infrastructure must therefore be capable of moving electricity across regions without creating congestion.

Urban concentration creates another challenge. Mumbai and Navi Mumbai have emerged as major data centre markets because of their connectivity, commercial activity and access to digital infrastructure. However, the concentration of large facilities can increase pressure on local substations, transmission networks and land availability.

India must also balance data centre growth with electricity requirements from households, agriculture, transport and manufacturing. Cooling demand during hot weather can increase peak electricity consumption, while industrial expansion and electric vehicle adoption create additional demand.

The objective should not be to restrict data centre investment. Instead, new projects should be located and approved with reference to realistic assessments of grid capacity, water availability, renewable energy access and future infrastructure requirements.

India’s Policy Response and Sustainable Data Centre Planning

The Indian government has recognised that data centre expansion requires coordinated planning across electricity, water, land and connectivity infrastructure.

The National Informatics Centre already operates national data centres and smaller facilities across different states to support government digital services. Meanwhile, the private sector is expanding capacity to serve cloud computing, enterprise applications and AI workloads.

In October 2026, NITI Aayog’s Frontier Tech Hub released a Data Centre Evaluation Framework to help state governments assess proposed projects more systematically. The framework is intended to shift attention beyond attracting investment towards infrastructure readiness, resource protection and long-term economic value.

This is an important development because states compete to attract data centre investment through land, electricity and fiscal incentives. However, a project that receives approval without adequate consideration of local infrastructure may create additional costs for utilities and communities.

India is also promoting renewable energy access through mechanisms such as Green Energy Open Access Rules and the Green Energy Corridor programme. The government’s broader energy strategy includes expanding renewable generation, strengthening transmission capacity and supporting storage technologies.

The Sustainable Harnessing and Advancement of Nuclear Energy for Transforming India (SHANTI) Act has also strengthened the policy framework for nuclear energy development, including the possibility of future small modular and micro-reactor deployment. These technologies may eventually contribute to reliable electricity supply for energy-intensive industries, although their commercial deployment will require time and further development.

How Technology Companies Can Reduce Electricity Consumption

Technology companies cannot treat electricity as an unlimited resource. Improving computing efficiency will be essential if AI and cloud services are to expand without creating unnecessary pressure on electricity systems.

One important approach is hardware optimisation. More efficient processors can perform more calculations using less electricity. Specialised AI accelerators can also improve performance per unit of energy for selected workloads.

Software optimisation is equally important. AI models can be made more efficient through improved architectures, quantisation, model compression and inference optimisation. Companies can also schedule certain computing tasks during periods when electricity is cheaper or renewable generation is more abundant.

Data centre operators can improve server utilisation by consolidating workloads and reducing the number of underused machines. Advanced cooling systems, efficient power distribution and real-time energy monitoring can further reduce supporting energy requirements.

Artificial intelligence can also help manage the energy systems themselves. Predictive models can forecast electricity demand, optimise cooling, identify equipment faults and coordinate battery storage. However, using AI to improve data centre efficiency does not eliminate the need to evaluate the additional electricity required to operate AI systems.

Could Nuclear Power and Energy Storage Become Part of the Solution?

The increasing electricity requirements of data centres have renewed interest in energy sources that can provide reliable power throughout the day. Nuclear energy is one option because it can generate electricity continuously with relatively low operational carbon emissions.

Some technology companies are exploring agreements with nuclear power operators to support their long-term electricity requirements. Small modular reactors are also attracting attention because their proposed designs aim to provide smaller, potentially more flexible nuclear generation units.

However, nuclear projects involve complex regulatory, safety, financing and construction requirements. Small modular reactors should not be treated as an immediate solution to current electricity shortages. Their contribution will depend on technological maturity, licensing, project economics and successful deployment.

Energy storage offers a more immediate supporting role. Battery systems can help manage short-term fluctuations, provide backup power and reduce pressure during periods of peak demand. Pumped-storage hydropower can provide longer-duration balancing where suitable geographical conditions exist.

The most reliable strategy is likely to combine multiple sources of electricity rather than depending on one technology. Renewable energy, storage, nuclear power, hydropower and other dispatchable generation can work together to support data centre operations.

The Economic Consequences of an Electricity Constraint

Electricity availability could influence the economics of the global data centre industry. Power costs already represent an important part of operating expenditure, and increasing demand may make locations with limited generation or transmission capacity less attractive.

If electricity becomes expensive or difficult to secure, companies may need to reconsider the location of new facilities. They could move projects to regions with stronger grids, invest in dedicated generation or negotiate long-term electricity agreements.

Higher infrastructure costs may also affect smaller cloud providers and technology startups. Large technology companies can often negotiate substantial power contracts and finance dedicated infrastructure, while smaller operators may have fewer options.

The consequences could extend beyond data centre operators. Electricity demand from large facilities may influence grid investment decisions, industrial power prices and public infrastructure spending. Governments will therefore need to establish transparent arrangements to ensure that the costs of new connections and grid upgrades are allocated fairly.

A successful digital economy requires both computing capacity and affordable electricity. If the cost of supporting digital infrastructure becomes excessive, it could affect the competitiveness of businesses and the affordability of digital services.

Conclusion

The global data centre boom is creating a new relationship between technological progress and electricity infrastructure. Artificial intelligence, cloud computing and digital services are expanding rapidly, but their growth depends on physical systems that require continuous and reliable power.

Electricity demand is not yet a universal barrier to data centre development, and the sector remains only a portion of total global consumption. Nevertheless, its rapid growth and geographical concentration can create serious local challenges. Grid connections, transformers, power generation, cooling systems and water availability are becoming essential considerations in technology investment decisions.

For India, the expansion of data centres presents an important opportunity to strengthen digital infrastructure, attract international investment and support AI development. However, this opportunity must be accompanied by coordinated electricity planning, renewable energy integration, efficient cooling and responsible use of local resources.

The next technology crisis may not arise from a shortage of data, computing algorithms or advanced processors. It could emerge when the electricity infrastructure needed to operate them cannot expand quickly enough. Countries and companies that treat energy availability as a core part of digital strategy will be better positioned to sustain the next phase of technological growth.

Online Internship with Certificate