RMZ targets up to 3 GW of India data centre capacity in $35bn expansion push

RMZ Group is seeking to expand its Indian data centre platform from approximately 250 megawatts to between 2 gigawatts and 3 gigawatts over five years, placing digital infrastructure at the centre of a broader $35 billion investment programme. The Bengaluru-based private real estate and investment group is in final-stage discussions concerning three projects that could lift its data centre capacity beyond 1 gigawatt, while it is also targeting land purchases by the end of 2026 that could support another 2 gigawatts. The plan extends beyond colocation facilities into AI factories, graphics processing unit infrastructure, power systems and software, signalling an attempt to move from property development into vertically integrated computing infrastructure. RMZ Group is also considering an initial public offering to secure permanent long-term capital for a buildout that will require sustained equity, debt and customer commitments. Strategically, the announcement makes RMZ Group one of the most ambitious Indian challengers in a market where power availability, rather than investor enthusiasm, may become the decisive bottleneck.

Why does RMZ Group’s 2 to 3 GW target mark a major escalation in India’s data centre race?

The scale of RMZ Group’s target is significant when compared with the current size of India’s data centre market. India had approximately 1.53 gigawatts of operational data centre capacity by the first nine months of 2025, meaning RMZ Group is aspiring to build a platform that could eventually be comparable to, or larger than, the country’s recent total installed base. Even if the final buildout lands at the lower end of the proposed range, RMZ Group would become a major national operator rather than a real estate company with a limited digital infrastructure allocation.

The immediate expansion pathway centres on three projects that are reportedly in advanced discussions. These projects could move the company beyond 1 gigawatt, although their commercial importance will depend on whether RMZ Group secures binding hyperscaler, cloud provider or enterprise commitments. Data centre pipelines often look impressive before power, permits, financing and tenant contracts are secured. The true milestone will therefore not be the announcement of additional sites, but the conversion of those sites into powered, pre-leased and financed capacity.

RMZ Group’s land acquisition strategy is equally important. The company aims to secure land capable of supporting a further 2 gigawatts before the end of 2026. In the data centre business, however, land without assured electricity is merely expensive soil. Site value depends on grid connectivity, fibre access, water availability, permitting timelines and the ability to accommodate high-density cooling systems. RMZ Group’s background in assembling and developing large commercial properties provides a useful starting advantage, but AI infrastructure demands a substantially different operating discipline.

The company’s target also reflects a wider change in how Indian developers view data centres. These facilities are no longer being treated simply as specialised real estate assets leased to technology companies. They are increasingly becoming platforms through which operators can capture revenues from colocation, cloud infrastructure, GPU access, managed services, connectivity and energy systems. RMZ Group appears to be positioning itself for that broader value chain.

How does the $35 billion headline translate into actual data centre capital deployment?

The $35 billion investment figure requires careful interpretation because it covers a wider RMZ Group expansion programme rather than data centres alone. The group previously indicated that its broader five-year capital plan would include colocation data centres, AI factories, commercial office developments and a return to residential projects. Within that programme, the company had outlined approximately $12 billion to $15 billion of spending to develop around 1.5 gigawatts of colocation capacity.

The newer target of 2 gigawatts to 3 gigawatts indicates that RMZ Group’s digital infrastructure ambitions may be expanding faster than originally framed. It could also indicate that the company expects to use partnerships, project-level debt, external investors or customer-backed structures rather than funding the entire programme directly from its own balance sheet. Large data centre developers frequently separate land, powered shell development, technology equipment and tenant fit-outs across different pools of capital.

That distinction matters because the capital required for an AI-ready facility can be dramatically higher than the cost of constructing a conventional data centre shell. Global data centre construction costs are expected to average approximately $11.3 million per megawatt in 2026, while tenant technology fit-outs involving GPUs, networking systems and specialised cooling can cost considerably more. A multi-gigawatt programme can therefore absorb tens of billions of dollars before accounting for electricity generation, transmission upgrades, financing costs and future hardware refresh cycles.

RMZ Group must consequently demonstrate capital discipline alongside ambition. Committing too much equity before securing tenants could weaken project returns, while relying heavily on debt could create refinancing pressure if construction schedules slip. A phased model, where new capacity follows pre-leasing milestones and power delivery certainty, would reduce speculative exposure. The company’s ability to bring institutional partners into individual campuses could also determine whether the 3 gigawatt ceiling is realistic or remains an aspirational pipeline figure.

Can RMZ Group convert its real estate capabilities into hyperscale infrastructure execution?

RMZ Group possesses several capabilities that can transfer from commercial property into data centre development. The company has experience in acquiring large sites, navigating planning processes, delivering complex buildings and working with multinational occupiers in Bengaluru, Mumbai, Hyderabad and other important Indian business centres. These capabilities can reduce early-stage execution risk and provide access to markets where demand for cloud computing, enterprise software and artificial intelligence infrastructure is growing.

Data centres, however, require engineering and operational capabilities that extend far beyond office development. Customers expect continuous uptime, redundant power systems, advanced physical security, resilient network connectivity and tightly controlled thermal environments. AI workloads further increase power density and cooling complexity, particularly as operators move towards liquid-cooled racks and larger GPU clusters. A property developer cannot simply add servers to a familiar building design and call the result an AI factory.

RMZ Group has reduced part of this capability gap through its equal joint venture with Colt Data Centre Services. The partnership was originally established with a planned investment of approximately $1.7 billion and a development pipeline of roughly 250 megawatts across Navi Mumbai, Chennai and another future location. Colt Data Centre Services contributes experience in building and operating hyperscale facilities across Europe, Japan and India, while RMZ Group contributes local development expertise and market access.

The partnership provides technical credibility, but RMZ Group’s expanded ambition raises questions about operating structure. It remains unclear whether the additional projects will sit within the Colt Data Centre Services venture, be developed through new partnerships or be controlled directly by RMZ Group. That choice will affect ownership economics, customer confidence, capital requirements and the speed at which the platform can scale.

Why are power access, land aggregation and customer commitments the real tests?

Electricity is likely to be the most consequential constraint facing RMZ Group. Data centres require not only large volumes of power but also reliable delivery schedules, grid redundancy and long-term pricing visibility. AI facilities can create particularly intense demand because advanced chips consume more electricity and generate substantially more heat than conventional enterprise computing systems.

Global data centre developers are increasingly pursuing behind-the-meter generation, renewable energy contracts, battery storage and private transmission arrangements because grid connection timelines in major markets can stretch for several years. RMZ Group’s stated interest in moving deeper into power infrastructure is therefore strategically logical. Controlling or partnering around generation and transmission can improve project bankability, accelerate commissioning and create a second revenue stream.

The strategy also exposes RMZ Group to a new set of risks. Power generation and transmission require regulatory approvals, specialised operating capabilities and considerable capital. Renewable energy output may need to be balanced with storage, grid power or other firm generation sources to meet round-the-clock computing demand. If RMZ Group moves too far into the energy value chain without the appropriate partners, the expansion could become operationally unwieldy.

Customer commitments represent the second test. RMZ Group has indicated that hyperscaler engagement has been encouraging, but it has not publicly identified customers attached to the proposed projects. Hyperscalers can significantly de-risk a campus through long-term leasing agreements, but they also possess substantial negotiating power. Large customers may demand pricing concessions, sustainability commitments, expansion options and strict delivery penalties.

The third test is whether the company can match land acquisition with confirmed power availability. Acquiring multiple large sites may create a valuable development pipeline, but an oversized land bank can tie up capital without producing cash flow. RMZ Group must resist the temptation to equate announced capacity with economically productive capacity.

How does RMZ Group compare with AirTrunk, CtrlS, Adani Group and other Indian challengers?

RMZ Group is entering an increasingly competitive market. AirTrunk has outlined plans to invest around $30 billion to develop approximately 5 gigawatts of Indian capacity by 2030. CtrlS Datacenters has secured a commitment of up to ₹70 billion from CPP Investments, including capital for an ownership stake and a joint venture focused on hyperscale campuses. International Finance Corporation has also committed capital to Sify Technologies for new facilities in Navi Mumbai and Chennai.

Adani Group is pursuing an even broader infrastructure strategy. The group has discussed plans involving renewable-powered, AI-ready data centres and has entered a partnership with Jabil to explore an integrated manufacturing platform for data centre infrastructure. Reliance Industries and Meta Platforms are also developing AI-enabled capacity in Gujarat, while global cloud providers continue expanding their direct and leased infrastructure exposure.

RMZ Group’s competitive advantage could come from combining institutional-grade real estate execution with a dedicated hyperscale operating partner. Its presence in major Indian commercial hubs may also create opportunities to serve global capability centres, financial institutions and enterprise customers that need cloud connectivity and data localisation alongside AI computing capacity.

The competitive danger is that the market becomes crowded before enough demand reaches contractual maturity. Announced Indian data centre pipelines now substantially exceed operational capacity, and not every proposed gigawatt will be delivered on schedule. Companies with secured power, anchor customers and low-cost capital will progress. Projects dependent on optimistic future demand may remain stuck between land acquisition and financial close.

RMZ Group must therefore compete on delivery certainty rather than announcement size. Hyperscalers are unlikely to select a facility merely because its developer owns attractive land. They will prioritise power availability, network ecosystems, technical standards, sustainability, cost and the operator’s record of meeting deadlines.

What does a potential RMZ Group IPO reveal about financing India’s AI buildout?

RMZ Group’s consideration of an initial public offering reflects the financing demands created by digital infrastructure. Data centres require substantial upfront expenditure, but revenue arrives gradually as phases are commissioned and leased. Permanent equity capital can support longer development cycles and reduce dependence on repeated asset sales or short-duration private funding.

An IPO could also provide a valuation mechanism for a business that increasingly spans commercial property, data centres, AI infrastructure and potentially power assets. Public investors may value digital infrastructure more favourably than conventional office development because of expected cloud and artificial intelligence demand. However, combining several asset classes within one listed entity can also create complexity and lead to a conglomerate discount.

The company has not disclosed whether a future listing would involve the entire RMZ Group, a digital infrastructure subsidiary or a separately capitalised operating platform. A dedicated data centre listing could provide investors with clearer exposure to capacity growth, utilisation and long-term customer contracts. A broader group listing could offer diversification but make it harder to assess capital allocation between offices, residential development and digital infrastructure.

Public-market scrutiny would also force RMZ Group to provide greater transparency around committed capacity, commissioned capacity, power agreements, leasing, development expenditure and project returns. This could improve capital discipline. It could equally reveal how much of the headline pipeline is genuinely contracted and how much remains dependent on future demand.

What could RMZ Group’s strategy mean for GPUs, software, power and India’s technology supply chain?

RMZ Group’s stated interest in GPUs, power infrastructure and software suggests that it does not intend to remain solely a landlord. A vertically integrated platform could earn revenues from powered space, computing capacity, managed infrastructure and enterprise services. This would place RMZ Group closer to emerging GPU cloud operators and AI infrastructure providers than traditional commercial developers.

Moving into GPU infrastructure could create higher revenue per megawatt, particularly as Indian companies seek domestic access to artificial intelligence computing. It would also expose RMZ Group to faster hardware depreciation, technology selection risk and volatile utilisation. Data centre buildings can operate for decades, while AI chips may face much shorter commercial cycles. That difference changes the risk profile of the business.

A software layer could help RMZ Group manage workloads, energy consumption, cooling and capacity allocation across multiple campuses. It could also support managed services for enterprises that lack the expertise to deploy large AI systems independently. However, developing a credible software and cloud operating business requires talent, customer support and cybersecurity capabilities that are not natural extensions of property development.

The company’s power ambitions may ultimately prove more important than its software plans. Developers that can bundle land, electricity, network access and computing infrastructure will hold stronger negotiating positions with hyperscalers. RMZ Group’s broader strategy therefore reflects a structural shift in which digital infrastructure companies are beginning to resemble combinations of utilities, property developers, telecom operators and technology service providers.

What must happen next for RMZ Group’s five-year data centre plan to become bankable capacity?

The first requirement is greater visibility on the three projects currently under discussion. Locations, expected megawatt capacity, development phases, power arrangements, customers and investment structures will determine whether the projects represent near-term construction opportunities or longer-term pipeline options.

The second requirement is evidence of hyperscaler or large enterprise pre-leasing. Signed customer commitments would support debt financing and validate demand. Without these commitments, RMZ Group could face pressure to fund substantial site and power expenditure before revenue visibility is established.

The third requirement is a clear partnership model. RMZ Group must explain how Colt Data Centre Services fits into the next phase and whether other investors, utilities, cloud providers or infrastructure funds will participate. A multi-gigawatt programme is unlikely to be financed efficiently through a single corporate balance sheet.

The fourth requirement is disciplined sequencing. RMZ Group should prioritise projects where power, fibre, land and approvals align rather than pursuing maximum geographic coverage. The Indian data centre race will not necessarily be won by the operator announcing the most capacity. It will be won by the operator that commissions profitable capacity while competitors are still waiting for substations.

RMZ Group’s proposal is strategically credible because it builds on established property capabilities and an existing data centre partnership. It is also financially and operationally demanding. The gap between 250 megawatts and 3 gigawatts is not merely a construction challenge. It is a transformation of the company’s business model, capital structure and risk profile.

What are the key takeaways from RMZ Group’s $35 billion India data centre strategy?

• RMZ Group’s proposed 2 gigawatt to 3 gigawatt platform could make it one of India’s largest data centre developers if the capacity is fully delivered.
• The $35 billion figure covers a broader investment programme, while the data centre component will probably rely on partnerships, project debt and external equity.
• Three projects under advanced discussion provide a potential route beyond 1 gigawatt, but customer commitments and power access remain undisclosed.
• RMZ Group’s commercial real estate experience helps with land and project execution but does not eliminate hyperscale engineering and operating risks.
• The Colt Data Centre Services partnership provides technical credibility and an existing foundation of approximately 250 megawatts.
• Power availability, transmission access and energy costs are likely to matter more than the size of RMZ Group’s announced land pipeline.
• Competition from AirTrunk, CtrlS Datacenters, Adani Group, Sify Technologies and global cloud providers will intensify pricing and execution pressure.
• Expansion into GPUs and software could improve revenue potential but would introduce hardware depreciation, utilisation and cybersecurity risks.
• A potential IPO could provide permanent capital and valuation transparency, although the proposed listing structure remains unclear.
• RMZ Group’s success will ultimately be measured by commissioned, leased and profitable megawatts rather than announced gigawatts.