Even if AI demand fades, India need not worry - about data centres

A lot of questions flank the excitement around AI-led data centre growth – will it last, will it be free of dents on the planet, and how much will the anatomy of a data centre change ahead? Manoj Paul, MD, Equinix India, spells out how new rack density requirements, hybrid cloud trends, AI training, AI inference, interconnect, etc., are changing not just ‘what’ data centres do but ‘how’ they do it all. And why will this be sustainable?

What has changed about datacentres after AI workloads?

AI opportunities are proliferating while simultaneously becoming more distributed and dynamic. Enterprises need a distributed AI infrastructure located everywhere their data lives to connect that data with compute and decision-making tools securely, and with speed, flexibility and ease. Equinix is the World Digital Infrastructure service provider with a focus on enterprises. As the next wave of AI innovation unfolds, integrating real-time intelligence and AI automation with multicloud workloads will support faster inference and edge-to-cloud connectivity.

AI workloads have fundamentally changed how Data Centres are designed and consumed in India. The industry is now entering a high-growth phase driven by two simultaneous shifts: the massive expansion of hyperscalers and the rapid acceleration of AI adoption. Today, nearly 60-70 per cent of new colocation capacity consumption comes from hyperscalers such as AWS, Google and Microsoft, who are not only expanding cloud infrastructure but are also preparing to deploy large-scale AI infrastructure within the country.

How is AI changing demand for interconnect, cloud on-ramps, and ecosystem density in India?

Until recently, most AI workloads of enterprises, especially proofs of concept, were run on the cloud. As Indian consumers and enterprises become increasingly AI-eager, AI inferencing will need to happen much closer to users. Over the next two to three years, deployment of AI inferencing is expected to outpace AI training by several multiples, creating one of the largest new demand waves for data centres. This shift is driving demand for dense interconnection, cloud on-ramps and ecosystem proximity rather than standalone capacity.

Domestic enterprises, particularly in BFSI and broking, are moving away from in-house data centres toward third-party colocation facilities.

What makes a data centre AI-ready?

AI pushes rack densities from ~5–10kW to 50–100kW+, making liquid cooling, greater power capacity, and purpose‑built ‘AI‑ready’ Data Centre campuses essential — whether for regional training clusters or dense inference. What makes a Data Centre AI-ready is the ability to support advanced cooling, predictable scalability and direct access to clouds, networks and partners in a sustainable manner.

Is there anything you can share from your Chennai setup here?

Equinix’s Chennai IBX was built with these requirements in mind. With liquid cooling capability, the facility supports high-density workloads, multi-cloud and ecosystem interconnection-first architecture, along with proximity to subsea cable landings, enabling low-latency global AI training and inferencing use cases for enterprises operating from India.

What are the most common enterprise patterns you see now—hybrid, multi-cloud, distributed inference, or cross-region designs?

Globally, we’re seeing strong and sustained growth in digital spend, with hybrid multicloud expanding at around nine per cent CAGR, reaching approximately USD 94 billion by 2029. What’s even more remarkable is the pace of growth in AI‑driven investments specifically, which are accelerating at a much faster rate—more than 20 per cent CAGR—and projected to reach around USD 94 billion by 2029.

In India, enterprises are rapidly adopting hybrid and multi-cloud architectures as they modernise their digital infrastructure. Domestic enterprises, particularly in BFSI and broking, are moving away from in-house data centres toward third-party colocation facilities to gain scalability, efficient interconnection with their required ecosystem, operational efficiency and access to specialised talent. This shift is being further accelerated by distributed AI, hybrid multi-cloud architectures and a growing focus on sustainability.

Are tier-2-3 towns becoming a significant trend- why?

Mumbai has long been a digital hub in India, but cities like Chennai are emerging as important extensions of the core ecosystem rather than standalone alternatives. As AI inferencing and digital services move closer to users, regional infrastructure is being integrated into national interconnection frameworks anchored in major metros. This approach allows enterprises to balance latency, resilience and scale while supporting India’s expanding digital footprint.

In an interconnect-heavy facility, what levers create the biggest sustainability gains—efficiency, renewables, or operational optimisation? Can you tell us more about your captive solar projects here?

At Equinix, sustainability is embedded into our business strategy under our global ‘Future First’ commitment, which focuses on delivering digital infrastructure that drives responsible innovation. In interconnection-heavy facilities, the biggest sustainability gains are achieved through initiatives around energy efficiency, renewable energy adoption and operational optimisation.

Globally, Equinix has achieved 96 per cent renewable energy coverage and continues to invest in projects that contribute new renewable generation to the grids in which we operate. In India, Equinix’s first renewable energy Power Purchase Agreement in APAC, signed with CleanMax, went live in November 2025. The 33 MW captive power project in Maharashtra, comprising 26.4 MW of solar and 6.6 MW of wind capacity, will supply clean energy to Equinix’s Mumbai data centres while advancing India’s renewable energy ambitions.

We have also invested in energy-efficiency projects globally, achieving an average PUE of 1.39, representing a year-over-year six per cent improvement in 2024. Our facilities in India are designed to international ASHRAE A1A efficiency standards, allowing optimised cooling and reduced energy use without performance trade-offs. Hosting critical workloads such as cloud, AI, and IoT closer to users further reduces latency and energy waste associated with long-haul data transfers.

What is unique about India’s Data Centre market versus other regions you operate in, and what will change fastest in the next two years?

India’s Data Centre market is distinctive because of the scale of its digital consumption, combined with the early stage of ecosystem development. India generates a significant share of global data, yet its installed data centre capacity remains comparatively low, creating strong long-term growth potential. This growth is now being amplified by hyperscalers and AI-led demand.

India aims to become a USD 1 T digital economy by 2028. It is already making significant progress, supported by the country’s thriving startup ecosystem, the third largest in the world, and initiatives like Startup India.

Widespread 4G and 5G networks and affordable telecom services have accelerated growth, attracting 100 million new 5G users in 2024. Telecom companies are attracting new users with some of the lowest pricing in the world. Affordable mobile internet means that Indians from all walks of life are able to consume massive amounts of content, including streaming video. In many cases, they’re also creating content themselves, driving further demand for compute and storage resources in the country.

And this translates into data centre demand?

For every megawatt (MW) of installed colocation capacity, users generate approximately 13.2 PB of data monthly. This ratio indicates India’s tremendous potential for Data Centre expansion compared to more mature markets. In other Asia-Pacific countries, data generated per 1 MW of installed colocation capacity is commonly much lower: 0.3 PB for Australia and just 0.01 PB for Singapore.

India’s low power and construction costs, combined with expanding subsea cable connectivity and interconnection density, also position the country to potentially host regional and global AI deployments for non-GDPR-bound datasets. Unlike more mature markets, India is still building ecosystem depth, making neutral interconnection platforms particularly critical.

Over the next two to three years, deployment of AI inferencing is expected to outpace AI training by several multiples, creating one of the largest new demand waves for data centres.

Over the next two years, the fastest changes will be seen in AI-driven capacity demand, continued hybrid multicloud adoption and interconnection growth. Chennai, Mumbai and other major hubs will see rapid expansion as enterprises prioritise resilience, regulatory readiness and ecosystem access over pure scale.

What are the top questions CIOs should ask to select an interconnection-centric provider with credible sustainability and resilience?

When selecting an interconnection-centric colocation service provider, CIOs should focus on ecosystem access, sustainability credibility and operational resilience rather than just available capacity and price. Questions should include how many networks, cloud providers and partners are available within the facility, how multicloud interconnection is delivered, whether they can pay per use of interconnection service with flexibility to modify capacity online and whether the company can scale their presence dynamically as AI and digital workloads grow.

Sustainability is increasingly a board-level concern. CIOs should examine renewable energy sourcing, efficiency metrics and transparency in environmental reporting.

Are outages and downtime serious concerns?

Resilience is equally critical. Outages and downtime remain serious concerns, particularly for AI, BFSI and real-time digital workloads. Facilities must be designed with redundant power, cooling and network paths, supported by rigorous operational processes and proven uptime performance.

Also, providers that enable cross-region architectures and offer diverse connectivity options significantly reduce single points of failure while supporting business continuity and long-term digital growth.

What happens to data centre capacities if the AI demand and hype turn out to be a bubble?

India’s evolution as a strategic digital corridor and a leading digital economy requires industry-leading digital infrastructure. Even if AI demand moderates in the short term, the fundamentals of India’s data centre market remain strong. Demand is being driven by BFSI, digital-native enterprises, hyperscalers, OTT platforms and enterprise AI inferencing, not by a single technology cycle.

Cloud adoption, regulatory requirements and enterprise modernisation continue to support sustained growth.

Data Centres in India are being built as foundational digital infrastructure rather than speculative AI-only assets, reducing exposure to short-term hype cycles.

Capacity expansion is increasingly being executed through phased and modular deployment models, allowing supply to closely track real demand. Interconnection-centric and multi-tenant facilities like Equinix retain long-term relevance because enterprises continue to require secure, low-latency connectivity and hybrid cloud access, regardless of individual workload trends.

AI may accelerate demand, but Data Centres in India are being built as foundational digital infrastructure rather than speculative AI-only assets, reducing exposure to short-term hype cycles.

Why are cloud-adjacent infra and last-mile connectivity, as being focused on by you, important? Would edge compute, modularity, etc., redefine data centre configurations in a serious way ahead?

AI-powered technologies require AI inference at the edge. So, it’s not surprising that edge AI is trending. The global edge AI market is expected to grow at a compound annual growth rate of 21.7 per cent over the next five years. Meanwhile, AI is being integrated into mobile networks, and mobile networks are being used for AI data transport.

Cloud-adjacent infrastructure is critical because it enables direct, low-latency access to major cloud platforms, improving application performance while reducing data movement costs. As AI, analytics and real-time digital services scale, last-mile connectivity becomes essential to delivering consistent user experiences.

One simple mobile AI query can involve data and infrastructure in many locations. Edge inference requires a dynamic hybrid multicloud network backbone to transport data quickly and reliably across the inference architecture. Private interconnection can play an important role here, offering deterministic performance and high security for sensitive inference data.

pratimah@cybermedia.co.in