Data centre stocks: 12 India picks for 2026 AI capex

Why India’s data centre buildout is moving markets

India is positioning itself as a global hub for artificial intelligence by building the physical infrastructure needed for AI computing. That shift is drawing investor attention away from pure software narratives and towards power, construction, electrical equipment, cooling, connectivity, and data centre real estate. Market moves have reflected that change, with a reported USD 4 billion added to the combined market value of 10 companies linked to the data centre supply chain in a single week. Another market snapshot referenced a roughly USD 1 billion jump in the combined market value of 10 related companies, highlighting how quickly sentiment can turn when large investments are announced. The rally has also been framed by a simple reality: India does not have many listed “pure-play” AI model or semiconductor manufacturers. As a result, listed proxies tied to data centre capacity and AI compute become the primary way investors express the theme.

Big-ticket announcements driving the narrative

Alphabet’s Google has announced a USD 15 billion plan to invest over the next five years, tied to a gigawatt-scale AI data centre campus in Andhra Pradesh. The project was first announced in October 2025 and includes an AI hub comprising three data centre campuses. AdaniConnex and Nxtra by Airtel are expected to lead construction of the data centre buildings and connecting infrastructure to help deploy and scale AI capabilities and digital services across India. Separately, OpenAI has said it will partner with the Tata Group to build a data centre starting at 100 megawatts (MW) of capacity and scaling up to 1 gigawatt (GW). Alongside Google and Microsoft commitments, and investments by Reliance Industries and Adani Group, about USD 17 billion in data centre investments have been announced across India.

The capex supercycle: from 1.5 GW to 10.5 GW capacity

A key framing in the market has been the scale of the planned buildout. India is embarking on a USD 60 billion capital expenditure cycle to expand data centre capacity to 10.5 GW by 2031, from roughly 1.5 GW today. That capital intensity pulls in multiple parts of the industrial ecosystem, not just the data centre operator. Power distribution, cooling systems, cabling, automation, and EPC work all rise in importance as facilities become larger and more power-dense due to AI workloads. This is also why the “AI-power nexus” theme has shown up repeatedly in investor discussions.

Hyperscale builders and operators: where the capacity is being planned

A useful way to separate “theme” from “execution” is to look at who is actually deploying capital and where. Several groups are building multi-campus footprints, often measured in MW and GW rather than in small enterprise racks.

Direct listed exposure: data centres and AI hardware

Bharti Airtel sits in the middle of the stack through Nxtra Data, which will lead construction of data centre buildings and connecting infrastructure for Google’s plan, as per Google’s release. Nxtra is described as one of India’s largest data centre operators, running hyperscale and edge sites hosting cloud and AI workloads for companies including Google Cloud and Amazon Web Services. The thesis presented is both direct (data centre revenue) and indirect (integration with Airtel’s fibre and enterprise services). On the hardware side, Netweb Technologies is positioned as an India-based manufacturer of servers, storage and AI compute systems, and is described as India’s only Nvidia OEM server partner in the provided material. Netweb is also referenced as a beneficiary of the PLI scheme, with demand linked to hyperscalers seeking India-made server infrastructure.

Real estate as data centre infrastructure: Anant Raj’s pivot

Anant Raj is presented as a real estate developer that has pivoted towards data centres and a sovereign cloud platform. According to its FY25 annual report (as cited), Anant Raj plans to expand its total data centre footprint to 307 MW over the next four to five years. The broader real estate group is treated more cautiously in the provided material, with names like Prestige, DLF, Brigade, and Mahindra Lifespace described as having exposure that is “optionality rather than core business.” The key distinction made is whether capital allocation follows announcements.

Pick-and-shovel beneficiaries: power, electrification, EPC, cables, cooling

The supply chain list in the provided text is broad, but it consistently highlights electrification, EPC, and thermal management as recurring bottlenecks in AI-era facilities. Siemens India is framed as a leader in grid automation and smart-building management needed for complex power distribution and cooling. ABB India is presented as a critical enabling player, with a specific datapoint that around 30% of hyperscale data centres in India already use ABB’s solutions. EPC exposure is led by Larsen & Toubro, described as integral to building data centres and associated power infrastructure, and also described as the only Indian firm currently building Nvidia AI factories. L&T also operates L&T Cloudfinity, with 14 MW of live capacity and another 18 MW in the pipeline, giving it both construction and operating exposure.

Cooling and HVAC are highlighted through Blue Star and Voltas as primary listed plays, with Blue Star described as focused on precision data centre cooling and running a dedicated data centre vertical. Connectivity and internal wiring exposure is represented through Polycab India, KEI Industries, Apar Industries and Finolex Cables. The provided material also names smaller thermal management exposures such as Johnson Controls-Hitachi Air Conditioning India and KRN Heat Exchanger.

Market impact: what moved, and why it mattered

The rally described in the provided text followed a high-profile AI gathering in New Delhi where Prime Minister Narendra Modi reiterated India’s goal to become a global AI hub, in the presence of AI leaders including OpenAI’s Sam Altman and Anthropic PBC’s Dario Amodei. In that week, E2E Networks, described as a cloud computing infrastructure provider that partners with Nvidia, rose more than 18%. Netweb Technologies India climbed by a similar measure. Other gainers cited include AurionPro Solutions and Techno Electric & Engineering Co., while larger names such as Larsen & Toubro and Adani Green Energy also advanced. The same narrative contrasts those moves with pressure on software and services providers, where investors are cautious about AI’s potential to hurt margins in productivity-linked services.

What the story signals for investors tracking India’s AI infrastructure

The provided material repeatedly frames the opportunity as infrastructure-led rather than application-led, because power supply, electrical equipment, cooling, and construction demand rises regardless of which AI application becomes dominant. It also points to category thinking: power and electrical equipment, cables and connectivity, cooling and HVAC, and data centre real estate. Several utilities and power-related names are discussed as indirect beneficiaries, including NTPC for rising power demand and Adani Green for renewable supply linked to hyperscaler preferences for green energy. The common thread is not a single “AI winner,” but a multi-year buildout where MW and GW of capacity translate into sustained orders and operating demand across the value chain.

Key datapoints snapshot

Conclusion

India’s AI ambition is increasingly being expressed through data centre capacity, power availability, and industrial execution rather than through listed AI model developers. Announcements such as Google’s USD 15 billion plan, OpenAI’s MW-to-GW scaling approach with Tata Group, and the broader USD 60 billion capex cycle to 2031 have pulled attention towards listed “picks-and-shovels” names across EPC, electrification, cooling, cables, and AI server manufacturing. The next set of milestones to track will be project timelines, MW additions, and follow-through on capital allocation by operators and real estate developers that have announced capacity plans.