AI's Biggest Winners May Not Be AI Companies

When we think of the AI boom, names like Nvidia, Microsoft and Google come to mind. But AI needs more than chips and software! It needs data centres, reliable power, cooling systems and high-speed connectivity.

This creates an opportunity for businesses building the infrastructure behind AI, even if they don't develop AI themselves.  

India's data centre industry is expanding rapidly, supported by cloud computing, digitalisation and rising AI adoption.

According to CBRE's 2026 data centre report, India's data centre sector had attracted approximately $173 billion in cumulative investment commitments since 2021, including $38 billion in the first half of 2026.

*Source: CBRE, October 2026. The year-end figure is a forecast, not completed capacity.  

More importantly, over 15% of cumulative investment commitments were linked to AI. These are investment commitments, not necessarily funds already spent or revenue earned by suppliers.  

This brings us to an interesting investment question: Who benefits when billions are spent building data centers? 

Imagine running thousands of high-performance computers around the clock. The electricity requirement is enormous.

AI data centres need a steady, uninterrupted power supply to run thousands of high-performance servers. AI-focused racks can require 30 to 50 kW of power, compared with 8 to 10 kW for traditional racks. 

This requires higher-capacity transformers and switchgear to handle the load, UPS systems to prevent interruptions, and backup generators during power outages. Companies such as Hitachi Energy India, ABB India and Cummins India supply parts of this electrical infrastructure. 

Cables carry power across data centres and enable high-speed data transfer between servers. Although they account for around 9% of data centre capex, demand grows as capacity expands. This creates opportunities for cable makers such as Polycab, KEI Industries and Apar Industries, as well as fibre-optic suppliers such as Sterlite Technologies and HFCL.   

AI servers generate significant heat, making cooling systems essential for reliable operations. Cooling can account for 15 to 20% of data centre spending, with denser racks driving demand for liquid cooling and advanced thermal management.

Blue Star, Voltas, Thermax, Johnson Controls and Vertiv operate in this space. If you’re looking to invest in this value chain, the key is identifying companies with the technology and execution track record to win specialised data centre contracts and convert rising demand into revenue. 

Land and construction account for about 15 to 20% of data centre spending, creating an opportunity for companies such as Larsen & Toubro, Techno Electric & Engineering and Anant Raj. Building a data centre in Mumbai can cost 20 to 30% less than in major global markets, making India more attractive for new projects.  

With around 3.9 GW of capacity expected by 2030 but only 509 MW under construction in H1 2026, the investment opportunity lies in companies that can convert planned capacity into actual construction contracts and revenue. 

Of course, the AI infrastructure story is exciting, but an exciting business opportunity doesn't automatically make a stock a good investment. 

Some AI-linked infrastructure stocks have already experienced sharp rallies. For example, an October 2026 Financial Express analysis highlighted Sterlite Technologies' significant share-price appreciation alongside concerns about its valuation.  

 Before investing in any data center beneficiary stock, it is prudent to ask 4 questions:

• Are sales, margins and cash flows improving?
• Can it deliver its order book without hurting profits?
• Does today’s share price already assume years of growth?
• Is it winning actual data centre orders, or only talking about the opportunity? 

The AI revolution is not just about software, chips or chatbots. It is also about the physical infrastructure needed to keep those technologies running.

As India's data centre industry expands, businesses supplying power equipment, optical fibre, cooling systems and engineering services could benefit from growing investment. But investors should distinguish between companies with proven AI-related earnings and those benefiting mainly from market excitement.

The next big AI opportunity may not have AI in its name. It may be a company manufacturing the cables, transformers or cooling systems that make AI possible. And that's what, we think makes the infrastructure side of the AI boom worth watching!

Disclaimer: This article is for educational purposes only and does not constitute investment advice or a recommendation to buy or sell any security. Investors should assess valuations, financial performance and risks before making investment decisions.