Why 2025 Became India’s Breakout Year For Global Data Centres

With global tech giants increasingly zeroing in on India as a key destination for data centre investments, 2025 turned out to be a net positive year for India’s tech ecosystem, despite a turbulent year on the geopolitical front.

The year can be seen both as a catalyst and an outcome of efforts by the central government and industry stakeholders to simplify policies, improve infrastructure, and make it easier to set up and scale data centre operations in India.

As a result, tech giants like OpenAI, Microsoft, Google, Meta, to name a few, announced plans to set up data centres in India. 

Notably, the global traction for India-based data centres is also a result of a sustained policy push. After granting ‘infrastructure’ status to data centres in 2022, the government is now considering tax exemptions and other fiscal incentives to lower operating costs, ease capital deployment and accelerate large-scale capacity build-outs across key markets.

“Data sovereignty remains a core priority under the broader Digital India agenda, with policymakers signalling a strong preference for critical data to be stored and processed within the country. This creates a direct incentive for hyperscalers to localise infrastructure,” data centre technology company Submer’s India head Dev Tyagi said.

While these efforts were going on for a while now, 2025 boomed and changed the global perspective on India’s data centre outlook. The trend was also driven by the global boost in AI adoption and switch to adaptation of Cloud usage. 

As per the Economic Survey 2024-25, India’s data centre market is projected to touch $11.6 Bn by 2032, at a CAGR of 10.98% from $4.5 Bn in 2023. 

Let’s dive deeper into how 2025 welcomed data centre boom in India 

Why India’s Data Centre Landscape Now Matters More

The AI conversation has evolved in waves — first curiosity, then concern around deepfakes, followed by fears of job losses. The year 2025, however, represents a pivot towards maturity, as companies move beyond broad debates to deploy targeted AI capabilities with applications that cut across sectors. From deeper penetration of GenAI to agentic AI taking up day-to-day tasks, AI and its forms captured a broader audience, resulting in enhancing AI workload. 

With AI usage going up, demand for data centres in the country has also surged. Training large language models (LLMs) requires extremely high power consumption, clusters of GPUs, advanced cooling and uninterrupted power. As per Inc42’s data, India’s homegrown AI market is projected to surpass the $17 Bn market by 2030. 

As per a report released by Cushman & Wakefield, by 2028, India will have 3X installed capacity with a total IT load of 3.29 MW.

“What once consisted largely of small, enterprise-owned server rooms has now become a national strategic priority, driven by the world’s largest mobile user base, rapid 5G rollout, and the explosive growth of AI workloads. Industry estimates suggest India’s installed data centre capacity will grow from just over 1GW today to more than 2GW by 2026, and exceed 8GW by 2030,” said Knight Fintech founder and CEO Kushal Rastogi.

Not just vertically, but India’s AI has expanded horizontally among several sectors including fintech, ecommerce, healthcare, education, enterprise SaaS and much more — with real-time use cases like chatbot query, recommendation, image generation, or fraud detection. 

AI models perform well when data is processed close to where it is generated.

At a time when the central government is projecting India as an emerging AI superpower, data centres have become indispensable physical infrastructure underpinning the country’s AI ambitions.

How India Became The Epicentre?

India has more than 900 Mn internet users who come from speaking different languages, income levels, geographies and digital maturity. This makes a perfect ground for AI models to be trained and deployed — to learn major complexities of data.

While this trend picked up pace in the post COVID era, 2025 clearly stood out as an inflection point series of developments that firmly established it as a landmark year for data centre push in the country. 

Earlier this month, MoS MeitY Jitin Prasada said that the cloud data centre capacity in India has reached approximately 1,280 MW, largely serving critical sectors such as banking, power and other critical public infrastructure. Based on industry estimates, Prasada expects this capacity to grow by 4-5X by 2030.

This growth phase is being powered by multiple hyperscale investment commitments and executions this year. The year kicked off with AI giant OpenAI announcing its plans to set up a 1 GW capacity data centre in India to store data of Indian users as well as from smaller neighbouring countries. As of now, the company is reported to be in advanced discussions with Tata Consultancy Services (TCS) to build AI compute infrastructure in the country, as well as co-develop agentic AI solutions for enterprises.

TCS also recently roped in global investment firm TPG as a major investor in HyperVault – TCS’ subsidiary which is building AI-focussed data centres across India, to invest around INR 18K Cr in this project.

Similarly, global and hyperscale players such as NTT Data, Reliance, AWS and Google announced hefty investments to set up data centres in India. In parallel, state governments including Chhattisgarh, Maharashtra and Telangana continued to attract large investment commitments for the establishment of AI-focused data centre infrastructure during the year.

Submer’s Tyagi further explains that India’s new user base and AI use cases align well with hyperscaler business models. Apart from this, India priorities data sovereignty, highlighting the strong preference of the policymakers to keep critical data within the borders. 

“This creates a direct incentive for hyperscalers to localise infrastructure. Competitive power costs in key regions, a deep engineering and developer talent pool, and a growing shift among enterprises toward cloud‑first and cloud‑native strategies further improve the business case and reduce perceived infrastructure risk,” he added. 

Notably, with the digital personal data protection rules (DPDP) out now, global cloud providers, hyperscalers, and enterprises have confidence to invest in long-term infrastructure like data centres.

The Policy And Geopolitical Push 

To sustain this momentum, both the Centre and state governments are stepping up efforts to attract global players to India’s data centre ecosystem.

In September, the draft National Data Centre Policy 2025 indicated that the government is considering tax exemptions for data centre developers for up to 20 years. Coupled with India’s mature IT and digital infrastructure, and relatively lower real estate costs compared to global markets, these measures aim to make India a more cost-effective destination for large-scale data centre investments.

“The median price of constructing a data center in India is estimated at $6.8 Mn per MW of capacity, significantly lower than in most other APAC nations. We strongly believe India is poised to become one of the world’s largest data center hubs,” added the Cushman & Wakefield report. 

Talking about the geographical advantages, India sits at the intersection of Asia, the Middle East, and Europe, making it a natural hub for data traffic moving between the East and the West. Additionally, Mumbai, Chennai, and Kochi are connected to multiple global cable systems – which further links India directly to the US, Europe, Southeast Asia, and the Middle East. 

“Global spending on digital and compute infrastructure is now estimated in the hundreds of billions of dollars annually, and boards are under pressure not to concentrate that capacity in a single geography or regulatory regime,” added Tyagi.

With the current investment bets and policy push, India is expected to be the fastest growing data centre market, with hyperscale segments growing above 25–30% of the overall market. 

“That trajectory implies continued demand for higher‑density racks, more power‑intensive compute, and advanced cooling approaches—such as immersion and direct‑to‑chip liquid cooling—to keep both energy use and operating costs within acceptable limits,” said Tyagi.

[Edited by Akshit Pushkarna]