Can India turn its data centre boom into a manufacturing win before imports dominate?

PwC India has projected that India’s data centre ecosystem could generate a potential $280 billion order book by 2035 as artificial intelligence, cloud adoption and digital transformation reshape infrastructure demand. India’s installed data centre capacity currently stands at about 1.6 GW and is expected to rise 8.5 times to nearly 13.8 GW by 2035. Direct investment in core data centre facilities is estimated at $71.6 billion, but the wider opportunity across IT equipment, power systems, cooling, construction and maintenance is significantly larger. The strategic relevance is clear: India’s data centre race is no longer just about real estate with servers inside, but about whether the country can build a domestic industrial supply chain around the physical backbone of artificial intelligence.

Why does India’s $280 billion data centre opportunity matter beyond artificial intelligence demand?

India’s data centre opportunity matters because artificial intelligence is turning digital infrastructure into a strategic asset class. Earlier data centre demand was driven mainly by cloud migration, enterprise digitisation, online platforms and regulatory pressure around data localisation. The next phase is different because AI workloads need far higher computing density, more advanced cooling, more reliable power and faster equipment refresh cycles. That changes the economics from a simple capacity expansion story into a broader industrial infrastructure story.

The $280 billion opportunity also shows that the largest value may not sit only with data centre owners. IT equipment, including chips, servers and networking systems, is expected to account for the biggest share of capital expenditure, estimated at $180 billion to $210 billion. That means the real economic prize is distributed across semiconductor supply chains, server assembly, networking hardware, electrical systems, cooling technology, design engineering, operations and after sales support.

For India, this creates both an opportunity and a warning. The opportunity is to become a trusted digital infrastructure hub at a time when global companies are diversifying cloud, AI and data operations. The warning is that without domestic manufacturing capacity, a large part of the value could flow to imported equipment suppliers. India may host the data centres, but if it imports most of the brains, nerves and cooling systems, the domestic value capture will be lower than the headline number suggests.

How could AI workloads change India’s cooling, power and data infrastructure economics?

Artificial intelligence is changing data centre design because high density computing generates far more heat than traditional enterprise workloads. Conventional air cooled systems and thermal containment models are suitable for lower density loads, but they become less effective when racks are packed with AI accelerators and high performance processors. PwC India’s assessment points to direct liquid to chip cooling and immersion cooling gaining importance as workloads move into much higher power densities.

This has major implications for capital allocation. Cooling is no longer a back office utility decision. It becomes a core design variable that can affect energy efficiency, uptime, operating cost and equipment life. Data centre operators that underinvest in advanced cooling may face higher power consumption, lower usable rack density and weaker competitiveness against hyperscale facilities built for AI from day one.

The power impact may be even more consequential. Data centres need reliable electricity, and AI data centres need even more of it. This will increase pressure on grid connectivity, renewable procurement, backup power systems and state level planning. The states that combine land availability, power reliability, connectivity, policy support and faster approvals could become the next magnets for digital infrastructure investment. In other words, the data centre boom will not be won only inside server rooms. It will also be won in substations, planning offices and power purchase agreements.

Why does import dependence create a strategic risk for India’s data centre expansion?

The import dependence issue is central because mechanical, electrical and plumbing components are long lead items. These components can be exposed to global supply chain delays, shipping disruptions, currency volatility and geopolitical tensions. When data centre construction depends heavily on imported systems, project timelines become vulnerable to events that have little to do with Indian demand.

This is especially important because data centre customers value predictability. Hyperscalers, cloud companies, banks, telecom operators, public sector institutions and AI developers cannot build mission critical infrastructure on uncertain delivery schedules. If transformers, cooling units, specialised racks, power distribution systems or critical components face delays, project commissioning can slip and revenue recognition can suffer.

There is also a strategic autonomy angle. Data infrastructure increasingly supports financial services, healthcare, government services, ecommerce, defence adjacent systems, telecom networks and enterprise AI operations. Heavy reliance on external original equipment manufacturers for equipment and after sales support can become a risk if spare parts, technical support or upgrades are delayed. Digital sovereignty is not only about where data is stored. It is also about who controls the equipment that keeps the data alive.

How could a data centre equipment PLI scheme change manufacturing and procurement?

A Production Linked Incentive scheme for data centre equipment could help shift India from being mainly a data centre consumption market to a data centre manufacturing and supply market. If designed well, such a scheme could attract manufacturing of servers, networking hardware, power systems, cooling equipment, racks, battery systems and other critical infrastructure components. This would allow Indian companies to capture a larger share of the ecosystem value rather than only providing land, construction and operations support.

Manufacturing clusters under Make in India could improve economies of scale. Data centre equipment demand is expected to be large and predictable over the next decade, which is exactly the kind of demand visibility manufacturers need before committing capital. If government agencies, hyperscalers and large enterprises aggregate demand, domestic manufacturers and global joint venture partners may have stronger confidence to set up facilities in India.

The challenge is policy design. Incentives must avoid creating assembly only operations with limited value addition. India’s goal should be to move up the stack from basic assembly toward components, systems integration, quality testing, thermal design, power electronics and maintenance capability. A weak incentive model would produce factories that depend on imported kits. A strong model would produce a domestic ecosystem that can compete on reliability, cost and service speed. The difference matters because nobody wants a data centre policy that merely puts a Make in India sticker on imported machinery.

What does the data centre buildout mean for power, real estate and regional planning?

The data centre boom could become one of India’s most power intensive infrastructure cycles. A rise from about 1.6 GW to nearly 13.8 GW of installed capacity would require a deep rethink of energy planning, grid readiness and renewable supply. Data centres need continuous power, which makes them different from many industrial users that can tolerate some flexibility. For states, the ability to guarantee reliable electricity could become as important as land incentives.

Real estate will also become more specialised. Data centres require suitable land parcels, fibre connectivity, low flood risk, reliable power access, cooling resources, security and proximity to demand centres. This could benefit established hubs such as Mumbai, Chennai, Hyderabad, Bengaluru, Pune and the National Capital Region, but it could also open opportunities for emerging locations if states offer the right mix of infrastructure and approvals.

Regional planning will need to consider water use and environmental stress. Advanced cooling can reduce some constraints, but large data centre clusters still create pressure on power, land and local infrastructure. India will need to balance digital growth with sustainability. Otherwise, the country could end up solving the AI infrastructure problem while creating a new local resource problem. That is not a digital transformation strategy. That is moving the headache from the server rack to the municipal office.

How should executives read India’s shift from data centre capacity to full ecosystem value?

Executives should read this as a shift from capacity counting to ecosystem building. The old question was how many megawatts or gigawatts of data centre capacity India could add. The new question is how much domestic value India can capture from every gigawatt added. That includes equipment manufacturing, engineering, cooling systems, power infrastructure, operations, maintenance, cybersecurity, cloud services and AI deployment layers.

For technology executives, the message is that infrastructure choices will influence AI strategy. Companies planning large AI adoption cannot treat data centre availability as someone else’s problem. Capacity, latency, cost, energy efficiency and resilience will directly affect AI economics. Firms that secure reliable compute access and data infrastructure early may have an advantage over those that assume capacity will always be available on demand.

For industrial executives, the opportunity is broader than digital services. Electrical equipment, precision manufacturing, thermal management, construction engineering, renewable energy, battery systems, cable systems and power electronics can all benefit from the buildout. India’s data centre boom could therefore become an industrial capex story, not just a technology story. The smartest companies will not ask whether they are in the data centre business. They will ask whether some part of the data centre value chain is about to enter their business.

What are the execution risks that could slow India’s data centre opportunity?

The first execution risk is power availability. Data centres need reliable and scalable electricity, and AI facilities are especially demanding. If power approvals, grid upgrades or renewable procurement lag behind capacity announcements, projects could face commissioning delays or higher operating costs. Power is not a supporting actor in this story. It is very much in the lead cast.

The second risk is policy fragmentation. Data centre approvals can involve multiple layers of government, including land, environment, power, building permissions, telecom connectivity and state incentives. If different states follow inconsistent or slow processes, India may lose time even while demand remains strong. Faster clearances can attract investment, but weak oversight can create sustainability and local infrastructure problems. The ideal framework needs speed with discipline.

The third risk is talent and maintenance capability. Data centres require specialised engineers, facility managers, cybersecurity professionals, thermal management experts and operations teams. As AI workloads become more complex, the talent requirement will deepen. If India builds capacity faster than it builds skills, uptime quality and operational resilience could become concerns. Buildings can be constructed quickly. Institutional competence takes longer, and it rarely arrives by overnight courier.

Why could India’s data centre opportunity become a policy test for Make in India?

The data centre opportunity is a natural test for Make in India because demand is large, infrastructure is strategic and the import risk is visible. Unlike some manufacturing sectors where demand must be created, data centre demand is already being pulled by AI, cloud, digital payments, ecommerce, streaming, enterprise software and government digitisation. That gives policymakers a strong base to build from.

A data centre equipment manufacturing push could also complement India’s broader electronics and semiconductor ambitions. Even if India does not immediately dominate advanced chip fabrication, it can build capability in servers, racks, cooling systems, power distribution, testing, assembly, system integration and maintenance. These are valuable parts of the infrastructure chain and can create skilled jobs if the ecosystem is designed with depth.

The policy challenge is to avoid treating data centres only as an investment headline. A $280 billion opportunity is attractive, but it will not automatically translate into domestic capability. India needs a coordinated approach across tax policy, manufacturing incentives, energy infrastructure, logistics, standards, research partnerships and workforce development. The opportunity is big enough to matter. It is also big enough to punish sloppy execution.

Key takeaways on what India’s $280 billion data centre opportunity means for business and policy

• India’s projected $280 billion data centre ecosystem opportunity by 2035 reflects a shift from basic digital infrastructure growth to AI driven industrial infrastructure expansion.
• Installed data centre capacity is expected to grow from about 1.6 GW to nearly 13.8 GW by 2035, creating demand across power, cooling, equipment, construction and operations.
• Direct data centre facility investment is estimated at $71.6 billion, but the broader value chain is much larger because equipment and infrastructure systems dominate capex.
• IT equipment, including chips, servers and networking systems, could account for $180 billion to $210 billion of the opportunity, making domestic value capture a major policy issue.
• Advanced AI workloads are pushing the industry beyond conventional air cooling toward liquid and immersion cooling systems that can support much higher rack densities.
• Import dependence in mechanical, electrical and plumbing components creates project risk because long lead equipment can be affected by global supply disruptions and geopolitical tensions.
• A dedicated Production Linked Incentive scheme for data centre equipment could help India build domestic manufacturing capability, provided it encourages real value addition rather than shallow assembly.
• Power availability, grid reliability and renewable procurement will become decisive factors in determining which Indian states attract the next wave of data centre investment.
• The opportunity extends beyond technology companies to electrical equipment makers, cooling specialists, engineering firms, real estate developers, power utilities and manufacturing partners.
• The broader signal is that India’s AI future will depend not only on software talent, but also on whether the country can build the physical infrastructure stack beneath it.