Industry Odisha Bureau, Sep 28: India’s data-centre buildout is lifting power, equipment and semiconductor suppliers. Goldman Sachs’s 42 “AI enablers” rose about 60% on average this year. The Nifty 50 fell roughly 12%.
India lacks the pure-play AI giants that dominate US markets. Yet artificial intelligence is reaching its equity market through a different route. That route runs through electricity, transmission equipment, generators and data centres. Semiconductor-linked businesses form another part of the chain.
Goldman Sachs Research has grouped 42 such Indian companies as “AI enablers”. They span power infrastructure, data-centre hardware, semiconductors and other supply-chain segments. Names include Adani Green Energy, Tata Power, Cummins India, CG Power and Kaynes Technology.
A rally beneath a weak benchmark
The cohort has risen about 60% on average since 1 January. Over the same period, the Nifty 50 has fallen roughly 12%. The 42 companies have a combined market capitalisation of about $670 billion.
The average masks wide variation between stocks. ABB India gained 36.2% and CG Power 38.9% by 25 September. Power Grid Corp rose just 1%.
Goldman said a rapidly compounding group had emerged beneath the weak index. The group makes up about 2% of India’s 1,800 listed companies. Yet it accounts for roughly 13% of the $5-trillion market capitalisation.
Data centres widen the footprint
The surge reflects rising demand for AI’s physical backbone. India has attracted more than $250 billion in investments over the past 12 months. Amazon, Google, Microsoft, Tata, Adani and Reliance are among the companies involved.
That investment creates demand well beyond technology firms. Needs range from electricity and transmission to generators, cooling, servers and semiconductors.
Cushman & Wakefield says India aims to reach 5.65 GW of operating capacity by 2030. That would more than triple current operating capacity.
The economics of a gigawatt
One gigawatt of data-centre capacity needs roughly ₹70,000 crore of investment. About 60% goes on hardware such as GPUs, storage and memory. Much of that spending currently flows to overseas suppliers.
The remaining 40% opens an addressable opportunity for domestic firms. It spans power, electrical equipment and other data-centre infrastructure. Capturing that share is not guaranteed.
Amit Chandra of HDFC Securities said India’s buildout is still early. He said construction has only started and is far from its peak.
A different kind of AI market
Goldman estimates AI-exposed companies make up just 16% of MSCI India’s market capitalisation. The comparable figure is 70-80% in Korea and Taiwan. It ranges between 30% and 50% in China and Japan.
Goldman described India as the default “anti-AI” trade at index level. The label reflects low index exposure, not an absence of AI investment.
In the US, JP Morgan links 42 generative-AI companies to 65-75% of S&P 500 returns. India’s exposure, by contrast, is emerging largely through infrastructure.
Orders support the move
The gains are relatively broad-based. Power, data-centre and semiconductor stocks rose between 40% and 80% in 2026. All nine Goldman sub-segments beat the broader market, and six gained over 20%.
Amit Anwani of PL Capital linked the rise to growing orders. Some companies now cite data centres as 10-15% of revenue, Anwani said. That revenue has grown over 20-25% in two to three years, he added.
Infrastructure before pure-play AI
India’s listed AI exposure remains modest by global standards. But the physical infrastructure behind AI is expanding. Power, equipment and data centres now offer India’s clearest route into the cycle.
