India’s data centre industry is entering a new phase of expansion. Infrastructure execution now matters as much as investment appetite. Union Commerce and Industry Minister Piyush Goyal said India could attract around $200 billion in data centre investment. He made the statement on September 2 during a CEO roundtable in New Delhi. The meeting focused on improving the ease of doing business for the sector. Officials discussed power availability, land readiness, approvals and fiscal conditions. They also highlighted rising demand from cloud services and artificial intelligence workloads. Hyperscalers accounted for more than 80% of data centre absorption in the first half of 2026. The investment figure therefore matters less as a headline than as an indicator of investment potential.

The Investment Ambition Moves Beyond Capacity Additions

The government’s latest projection connects data centres with a wider economic strategy. That strategy also includes artificial intelligence, semiconductors and manufacturing. Goyal has linked these sectors with India’s longer-term economic ambitions. He also called for at least 10 GW of data centre capacity by 2030. This target gives the industry a physical reference point for future growth. Investment and capacity, however, remain different measurements. A large capital opportunity does not automatically create an equal amount of operational capacity. Projects still need land, power, financing, equipment and customer commitments. A project can therefore have substantial potential investment without immediately adding equivalent live IT load.

The distinction matters because data centre investment passes through several development stages. Developers first need suitable sites and reliable utility access. They must then secure approvals and complete detailed engineering. Financing and customer commitments can influence the pace of construction. Equipment procurement can also affect project schedules. Transformers, generators, switchgear, cooling systems and networking equipment require careful planning. Construction completion does not always mean that a facility can immediately support its full planned load. Commissioning, testing and customer deployment must follow before capacity becomes operational. This makes infrastructure readiness an important measure of India’s data centre growth.

Why AI Is Changing the Infrastructure Equation

Artificial intelligence is changing the technical profile of data centre demand. Accelerated computing can place large processing loads inside dense server environments. Training and inference workloads rely heavily on GPUs and other accelerators. These systems can create higher rack densities than many traditional workloads. That change affects power distribution, cooling, backup systems and network design. India’s AI infrastructure plans already include public access to high-performance computing. The IndiaAI Mission has established access to more than 38,000 GPUs through 14 cloud partners. Inference workloads can create infrastructure needs across locations, while training can use specialised high-density clusters.

AI also changes how operators approach facility design. Higher compute density can increase power requirements within individual racks. It can also raise the amount of heat that cooling systems must remove. Network infrastructure must handle large volumes of data between processors, storage systems and users. Backup systems must support critical workloads during power disturbances. These requirements can make AI facilities more infrastructure-intensive than conventional enterprise data centres. Operators therefore need to consider compute, power and cooling as connected design elements. The growth of AI can also create demand for both large training clusters and distributed inference capacity. That combination could broaden the infrastructure requirements across India.

Power Becomes a Strategic Infrastructure Requirement

Power availability sits at the centre of data centre development. Facilities require continuous electricity at significant load levels. The September roundtable examined cluster-based transmission and substation planning. Officials also discussed sanctioned load from the first day of operations. Dual power feeders formed another part of the discussion. These measures address a practical development risk. A facility can reach construction completion before local power systems are ready. Data centre operators need certainty about capacity, connection timelines and redundancy. The power discussion also includes renewable procurement and access to cleaner electricity.

Electricity requirements become more complex as AI workloads expand. A conventional data centre may already operate at substantial continuous load. AI clusters can add much higher density within selected server areas. This can increase demand on transformers, switchgear and distribution systems. Developers therefore need accurate power planning before facility construction begins. Grid operators also need visibility into future data centre demand. Transmission and substation investments can take time to plan and execute. Early coordination can reduce the risk of infrastructure arriving after the data centre itself. Power readiness will therefore remain one of the strongest factors shaping the speed of capacity deployment.

Grid Connectivity Will Shape Where Facilities Can Scale

Future data centre clusters will depend partly on grid capacity. Electricity networks must support concentrated and continuous demand. The government has proposed cluster-based transmission and substation planning. This approach could align power infrastructure with expected data centre development. Power-ready parcels could also reduce delays during site selection. Dual feeders can add another layer of resilience for critical facilities. Day-one sanctioned load is equally important for new campuses. Developers cannot rely on future power capacity when a facility starts operations. The government’s reform agenda therefore places power, land and approvals alongside investment attraction.

Grid connectivity can also influence the economics of site selection. A site with strong utility infrastructure can reduce development uncertainty. Another location may offer cheaper land but require major grid upgrades. Developers must compare these factors before committing capital. The same assessment applies to network connectivity and expansion potential. Data centre campuses often need multiple fibre routes for resilience. They may also require strong connections to cloud platforms and internet exchanges. As a result, infrastructure readiness can become a stronger location factor over time. States with coordinated power and digital infrastructure may gain an advantage in attracting new projects.

Land Readiness Becomes Part of the Digital Infrastructure Strategy

Land is more than a real estate issue for data centres. Large facilities require significant technical infrastructure. Sites must accommodate buildings, substations, backup systems and cooling equipment. Network connections and future expansion also need space. The government has called for data centre-ready land banks. It has also highlighted pre-identified power-ready parcels. Such preparation can reduce uncertainty during early project planning. Suitable sites also need roads, zoning and telecommunications connectivity. Land preparation therefore involves much more than making industrial plots available.

Site selection can also determine the long-term scalability of a data centre campus. Operators often need additional land for future capacity. Electrical yards and cooling infrastructure can require substantial supporting space. Emergency access and fire-safety arrangements must also fit within the site plan. Network routes need protection from common physical failure points. Water requirements can become relevant for facilities using water-based cooling systems. Environmental and local planning requirements can further affect development schedules. A prepared site can therefore remove several early-stage obstacles at once. This makes land readiness an important part of India’s wider digital infrastructure strategy.

Faster Approvals Could Reduce Project Uncertainty

Regulatory timelines can affect large infrastructure projects. Construction schedules determine when invested capital can generate operational capacity. The government has therefore discussed streamlined single-window approvals. It has also examined building regulations for data centre projects. The National Building Code 2026 now recognises data centres under Group E. It includes a dedicated annex for data centre requirements. These include fire-risk assessment and data-centre-specific performance indicators. Clear requirements can help developers understand technical expectations earlier. The government expects streamlined approvals to improve timelines and investment certainty.

Data centres involve several technical and administrative approvals. Building permissions represent only one part of that process. Developers may also need electricity connections and other utility clearances. Environmental requirements can apply depending on the project and location. Local authorities can have additional requirements for construction and operations. Multiple approval layers can increase uncertainty when processes do not align. A single-window approach can provide a clearer administrative route. Its effectiveness will depend on coordination between the authorities involved. Better coordination can help projects move from planning to construction with greater predictability.

Hyperscalers Are Setting the Pace of Capacity Demand

Hyperscalers are driving a large share of current data centre demand. Government figures show that they represented more than 80% of absorption in H1 2026. Large cloud platforms require substantial computing and storage capacity. Their workloads include cloud services, enterprise applications and consumer platforms. Artificial intelligence is adding another major demand source. Hyperscaler requirements can influence the timing of new campuses. They can also shape facility design and expansion plans. Customer capacity commitments can give developers better visibility into expected demand. Actual deployment still depends on project development and technical readiness.

Hyperscaler demand also affects the infrastructure surrounding a facility. Large campuses require strong fibre connectivity and reliable network routes. Cloud customers may need direct connections to their infrastructure. High-capacity power systems must support both current and future requirements. Facilities can also need flexible expansion zones for additional server halls. These factors encourage developers to plan campuses for long-term growth. Customer requirements can influence rack density, redundancy and cooling architecture. The result is a development model where customer demand and infrastructure planning remain closely connected. India’s ability to support this demand will influence the pace of new capacity additions.

The Cloud Market Creates a Wider Infrastructure Chain

Data centre growth extends beyond server halls. Digital services depend on networks, power systems and physical facilities. India’s large internet base provides a broad demand foundation. Goyal has previously cited nearly one billion internet users. Affordable connectivity and 5G are also supporting digital applications. These trends can increase demand for computing and storage. Cloud adoption can support hyperscale, colocation and edge facilities. AI workloads can require infrastructure close to users or data. Large training workloads can instead operate on specialised high-density computing clusters.

This wider infrastructure chain creates several opportunities for operators. Hyperscale campuses can support very large cloud deployments. Colocation facilities can provide shared infrastructure for enterprises. Edge facilities can bring computing closer to users and applications. Network providers connect these locations to customers and cloud platforms. Power companies provide the electricity needed to operate the infrastructure. Equipment manufacturers supply the electrical and mechanical systems inside facilities. Engineering firms support design, construction and commissioning. Together, these segments form a broader ecosystem around India’s digital infrastructure expansion.

Sustainability Will Become an Infrastructure Design Constraint

Energy consumption will remain a major issue as capacity expands. Data centres operate continuously and require both IT and facility power. Goyal has connected technology growth with India’s clean energy ambitions. Renewable power can come through several procurement models. These include direct contracts and open-access arrangements. The latest roundtable also discussed renewable procurement across state boundaries. Operators must balance renewable sourcing with the reliability needs of critical facilities. Energy efficiency also depends on servers, cooling and power distribution. Sustainability can therefore become part of engineering and procurement decisions.

The sustainability challenge extends beyond electricity sourcing. Operators must also examine cooling efficiency and equipment utilisation. Higher server utilisation can improve the efficiency of available computing capacity. Better power distribution can reduce avoidable electrical losses. Cooling systems can also be designed around actual rack requirements. Water consumption may become an important factor for some cooling architectures. Renewable procurement does not remove the need for reliable backup power. Operators therefore need a combined approach to energy, cooling and resilience. Sustainability will increasingly influence both facility design and operating strategy.

Cooling Requirements Will Evolve With Compute Density

Higher-density computing changes the thermal requirements of data centres. Nearly all computing power eventually becomes heat. Facilities must remove that heat continuously. Traditional air cooling can support many workloads. Dense accelerator deployments can create more demanding thermal conditions. Liquid cooling can provide direct heat removal for suitable high-density deployments. It can also support rack-level thermal management where air cooling faces practical limits. Cooling choices depend on equipment, rack density and facility design. Operators must therefore consider cooling during the early engineering stage.

AI facilities can place significant thermal loads inside individual racks. This can challenge conventional room-level cooling approaches. Direct-to-chip liquid cooling is one option for suitable accelerator systems. Rear-door heat exchangers can provide another approach for certain deployments. The right solution depends on server design and operating requirements. Cooling infrastructure also consumes power through pumps, chillers and other equipment. Water requirements vary according to the selected system and operating model. Operators must balance thermal performance, energy use, maintenance and resilience. Cooling has therefore become a core consideration in high-density AI infrastructure.

AI Infrastructure Is Expanding Beyond Private Hyperscale Campuses

India’s AI strategy includes public and shared computing resources. These resources complement commercial data centre expansion. The IndiaAI Mission aims to widen access to high-performance computing. Eligible users include startups, researchers and government organisations. Shared infrastructure can reduce the need for every organisation to build its own facility. It can support model development, research and experimentation. Commercial data centres will still serve larger production requirements. Different workloads need different capacity levels and service models. India’s AI infrastructure landscape therefore includes several models for accessing computing capacity.

This structure can broaden the base of India’s AI ecosystem. Startups can access computing resources without building dedicated infrastructure. Researchers can use shared capacity for experiments and model development. Government programmes can support applications with specialised computing requirements. Commercial providers can continue serving large enterprise and production workloads. These models do not need to compete for the same infrastructure role. Instead, they can address different stages of AI development. The combination can also create additional demand for data centre operators and cloud providers. AI infrastructure growth will therefore extend beyond a small number of hyperscale campuses.

Investment Will Need a Deeper Domestic Supply Chain

Data centre construction creates demand across many equipment categories. These include electrical systems, cooling equipment and networking hardware. Generators, transformers and switchgear are also important. Construction and engineering services form another major requirement. India’s manufacturing expansion intersects with these infrastructure needs. Domestic suppliers can participate as data centre capacity grows. They can support electrical, cooling, networking and construction-related requirements. Domestic manufacturing can also provide another sourcing option. That option works where local production meets technical and certification standards.

The infrastructure requirement also extends into facility operations. Data centres need skilled technical personnel after construction. Electrical and mechanical specialists support critical systems. Network professionals maintain connectivity and computing infrastructure. Security teams protect facilities and sensitive equipment. Facility-management personnel also support daily operations. AI infrastructure creates additional demand for specialised skills. IndiaAI programmes already include skilling initiatives and planned Data and AI Labs.

A deeper domestic supply chain can support the sector in several ways. Local sourcing can provide another option for equipment procurement. Domestic engineering firms can also participate in facility design and construction. Service providers can support maintenance after commissioning. Manufacturers may develop products for the specific needs of high-density computing. However, technical certification and performance requirements will remain important. Global suppliers will continue to serve categories that require specialised technology. India’s opportunity therefore lies in expanding capable domestic participation alongside established international supply chains. This could strengthen the broader economic impact of data centre investment.

State-Level Execution Will Determine the Pace of Development

India’s data centre market operates across a federal structure. States influence several important parts of project development. These include land, local approvals and electricity distribution. Industrial policies can also differ between states. Developers therefore assess several factors before choosing a location. Power availability is one key consideration. Network connectivity and land readiness are equally important. Incentives alone cannot solve infrastructure or approval delays. Project execution still depends on land, power, approvals, financing and construction readiness.

The September roundtable included representatives from eight states. They were Andhra Pradesh, Gujarat, Haryana and Karnataka. Maharashtra, Tamil Nadu, Telangana and Uttar Pradesh also participated. Their participation reflects the importance of state-level execution. Each state can offer a different infrastructure proposition. That proposition may include power, land, connectivity and industrial support. Central and state coordination can reduce duplication across approval processes. The participation of multiple states places greater attention on practical infrastructure readiness.

State-level execution can also influence how quickly investment moves into construction. Developers need predictable access to land and utilities. They also need clarity on local regulations and approval procedures. Power distribution arrangements can differ between locations. Network availability can also influence the attractiveness of a site. States that prepare infrastructure in advance may reduce early project delays. This can become increasingly important as developers compare several potential locations. Competitive policy frameworks can support investment attraction, while project execution depends on practical readiness. The result could be stronger competition between states based on infrastructure capability as well as incentives.

Infrastructure Readiness Will Matter More Than Announcements

The $200 billion figure represents significant investment potential. Its operational impact will depend on individual project execution. A typical project passes through several development stages. These include site selection, power assessment and regulatory approvals. Financing and detailed design follow those early steps. Equipment procurement and construction then determine physical progress. Commissioning must occur before capacity becomes operational. Infrastructure planning can reduce some of these development risks. Execution quality will influence how effectively investment potential becomes working infrastructure.

The government’s latest roundtable focused on actionable and time-bound reforms. That approach places attention on implementation rather than investment announcements alone. Cloud and AI customers also require capacity within commercially relevant timelines. Delays can therefore affect both developers and customers. Operators also need room to expand facilities as demand changes. Investors will assess India’s digital market and its infrastructure processes. Power, land, regulation and connectivity will remain key factors. Workforce availability will also influence operational readiness.

The Next Phase Will Connect Digital Infrastructure With Physical Infrastructure

India’s data centre push represents a broader infrastructure shift. Digital services now depend heavily on physical infrastructure. Electricity networks support computing capacity. Industrial land supports large-scale facilities. Telecommunications networks connect those facilities with users and cloud platforms. Construction and engineering capabilities determine how quickly campuses can become operational. AI adds further pressure through high-density computing requirements. Cloud growth creates additional demand for scalable infrastructure. The $200 billion opportunity will ultimately need to translate into projects, investment and commissioned capacity.

Power planning will remain central to this development cycle. Ready-to-use land can reduce early project uncertainty. Dual feeders can support power resilience. Renewable procurement can support energy objectives. Updated building regulations can create clearer technical requirements. IndiaAI can expand access to specialised computing resources. Commercial hyperscalers will continue to influence large-scale capacity demand. These developments connect investment policy with physical infrastructure planning. The next stage of growth will depend on how effectively those systems develop together.