Compute-Led Urbanisation: The Next Phase of India’s Economic Development

Over the past few months, it has become increasingly clear that artificial intelligence is entering an infrastructure-heavy phase. What was less clear, until recently, was how quickly this shift would begin to express itself in physical investment patterns—particularly outside the United States.

That clarity is now emerging.

On 13 April, Bengaluru-based real-estate company RMZ announced plans to invest approximately $35 billion in India over the next five years in data-centre, office, housing, retail, and logistics infrastructure. 

Even allowing for execution uncertainty and phased deployment, the scale of this commitment is difficult to ignore. It implies sustained, multi-billion-dollar annual capital deployment into long-duration physical assets.

At first glance, this appears to be a large-scale real-estate expansion push. On closer examination, it signals something more consequential: the spatial expression of the industrialisation of AI.

Infrastructure, Not Real Estate

The nature of this massive investment warrants reclassification. It is not simply about building commercial or residential space. It is about constructing an integrated asset base that combines data centres, office towers, housing, logistics parks, and urban services—backed by institutional capital, including pension funds and private equity funds.

This is characteristic not of a traditional real-estate developer, but of a mega infrastructure developer. Such developers do not build to sell. They build to own, operate, and generate long-duration returns. In this case, the underlying demand driver is not merely urbanisation or consumption, but the rising need for compute and enterprise infrastructure.

What appears as built space is, in effect, a physical layer designed to house and sustain the deployment of AI at scale.

A Sector Entering Its Institutional Phase

This investment is not occurring in isolation. India’s data-centre sector is itself entering a new phase.

According to a recent report by Vestian (released on 13 April) India's installed data-centre capacity could expand from roughly 700 MW currently to 5 GW by 2030, and the market size is expected to hit 22 billion dollars by 2030 from 10 billion dollars in 2025. This would position India as a primary digital infrastructure hub within the Asia-Pacific region. The report further says that 80% share of this expansion is being financed by foreign institutional capital.

The implication is that India’s data infrastructure story is no longer a function of incremental demand growth alone. It is becoming a site of large-scale, globally financed capital formation.

In that context, large integrated developers are not anomalies. They are early participants in a broader structural shift.

Infrastructure in a Fragmenting World

At the same time, the geographical logic of data infrastructure is beginning to change.

Recent disruptions in West Asia have underscored that data centres are not insulated from geo-political conflicts. As digital systems become more central to economic and strategic activity, their infrastructure must be designed for resilience as much as for efficiency.

This is giving rise to a new pattern: distributed and co-located architectures, where capacity is spread across multiple geographies to mitigate concentration risk.

In this emerging framework, data infrastructure begins to follow a logic previously associated with energy security—redundancy, diversification, and strategic location.

India’s relevance, therefore, is not purely economic. It is increasingly geo-political.

Compute-Led Urbanisation

Taken together, these shifts are beginning to reorganise the structure of urban development.

Data centres, by their nature, act as anchor infrastructure. Around them, enterprise activity clusters. Offices follow. Housing follows labour. Retail and services follow population.

The result is not simply city expansion, but a different organising principle for urban growth—one anchored in compute capacity.

This can be understood as a form of compute-led urbanisation, where digital infrastructure does not merely support cities but begins to shape their configuration.

Beyond the Office: The Factory as the Next Frontier

At present, much of the demand for such infrastructure is tied to office-based use—global capability centres, IT services, and digital platforms.

However, this is unlikely to remain the primary locus of demand indefinitely.

As AI matures, its deeper economic effects are expected to emerge within physical systems—manufacturing, mining, refining, logistics, and energy systems. These sectors are far larger in economic footprint and far more consequential for productivity.

This suggests that future demand for compute may increasingly arise not only in metropolitan office clusters but also in and around industrial ecosystems, including smaller cities linked to industrial corridors.

AI may first saturate the office, but its deeper economic impact will emerge when it reaches the factory floor.

This transition may also begin to reorganise how India’s technology services sector operates geographically. In megapolitan regions, established IT companies are likely to continue serving large global and Indian enterprises, supported by nearby hyperscale data-centre clusters. 

In parallel, a different layer may emerge in industrial clusters/parks and smaller cities, where new-age AI-first IT startups work more closely with manufacturing startups, SMEs, and logistics companies — drawing on compute infrastructure located in/near these clusters/parks.

These two layers will not be mutually exclusive. However, the underlying shift is important. India’s traditional advantage in IT services was built on labour arbitrage. More recently, it has evolved towards AI integration and tailoring capabilities. In an infrastructure-bound AI environment, a further dimension may emerge—one shaped by proximity to compute itself. IT firms operating from closer to data-centre infrastructure may benefit from lower latency, tighter system integration, and improved responsiveness, particularly in industrial applications.

Building for a Moving Compute Frontier

A further complexity arises from the nature of the underlying technology itself.

AI infrastructure is inseparable from the evolution of compute hardware. Firms such as NVIDIA, AMD, and Intel, along with a growing number of specialised chip design firms, are continuously advancing performance, efficiency, and architectural design—often through modular, chiplet-based systems.

In such a context, data centres cannot be designed as static assets optimised for a single hardware paradigm. They must accommodate heterogeneous compute environments, capable of integrating multiple architectures and evolving over time.

This introduces a structural tension. Infrastructure is long-duration by nature. Compute innovation operates on short cycles.

Resolving this tension requires designing facilities with flexibility at their core. Long-duration infrastructure must be built for short innovation cycles.

The Resource Constraint Layer

Expansion at this scale inevitably raises questions about land, water, and energy. Data centres are resource-intensive assets. Their viability depends not only on connectivity and demand, but also on sustained access to power, cooling systems, and physical space.

Treating these constraints as externalities is no longer viable. Instead, they must be integrated into the planning logic itself. 

This points towards a more ecosystem-oriented approach: long-term partnerships for green energy and sustainable water supply. In this regard, greenfield ecosystems—rather than incremental additions to already-stressed energy and water systems—may prove to be more strategically durable.

India’s Strategic Position: Policy, Capital, and Direction

India enters this phase with a set of structural advantages that remain intact.

It possesses a rapidly expanding renewable energy base, a large and growing domestic market for digital services, and a demonstrated capacity to integrate complex technological systems into enterprise operations.

What has become more visible recently is the alignment of policy with these structural conditions. The Union Budget has introduced a long-horizon tax exemption framework aimed at attracting global data centre and cloud infrastructure into India. By reducing tax uncertainty and clarifying treatment of cross-border digital activity, this policy signals that AI infrastructure is being treated as strategic infrastructure. This marks a shift from passive enablement to active positioning.

At the same time, large private capital commitments—both domestic and foreign—are beginning to converge on this opportunity.

The combination of policy intent, capital availability, and rising demand creates a window that is structurally significant.

Yet, the outcome is not predetermined. The presence of infrastructure does not automatically translate into strategic advantage. There remains a risk that India becomes a hosting layer for globally owned systems, with limited domestic value increment.

Therefore, the distinction between hosting and shaping remains critical.

Conclusion: The Structural Choice

In the nineteenth century, industrial power followed coal.

In the twentieth, it followed oil.

In the twenty-first, it is increasingly following compute.

What is now visible is that this transition is no longer confined to abstract projections or hyperscaler balance sheets. It is beginning to organise capital allocation, infrastructure design, and urban form in real time.

The industrialisation of AI has moved from concept to construction.

The question is no longer whether India will participate in this transformation. It will.

The more consequential question is whether it will shape the ecosystem in which this transformation happens—or remain just a host within it.