India's $20 Billion Chip Bet Enters Next Phase as AI Sparks Data Centre Boom

India's semiconductor ambitions are moving from blueprint to reality. Four years after launching the Semicon India programme, the government says the country now has 12 semiconductor projects worth around $20 billion, with three facilities already manufacturing chips. At the same time, a parallel infrastructure boom is taking shape, with AI and cloud computing expected to drive a massive expansion in India's data centre capacity over the rest of the decade.

Union Electronics and IT Minister Ashwini Vaishnaw said the government's semiconductor strategy has entered a new phase after learning from earlier failed attempts to build a domestic chip industry.

"2022: We started Semicon India. Learnt from past attempts. Clear policy and execution," Vaishnaw said in a post on X.

The minister said the programme has now attracted 12 semiconductor units involving nearly $20 billion in investments, with three already producing chips. He also announced that the government has approved Semicon 2.0, signalling a renewed push to expand India's semiconductor manufacturing ecosystem.

"2026: 12 semiconductor units with $20 billion investment. Three units already making chips. Semicon 2.0 approved. And we are just getting started," Vaishnaw wrote.

The comments came in response to a post recalling how chipmaker AMD had once planned to invest $3 billion in a semiconductor fabrication facility in India in 2005 before the proposal collapsed amid policy and execution hurdles.

India's semiconductor push is aimed at reducing dependence on imported chips that power everything from smartphones and automobiles to telecom equipment, defence systems and artificial intelligence applications. With Semicon 2.0 now approved, the focus is expected to shift from announcing projects to expanding manufacturing capacity, strengthening the supplier ecosystem and positioning India as a competitive player in the global semiconductor value chain.

AI Boom to Fuel India's Next Data Centre Wave

The semiconductor push comes as India's digital infrastructure is preparing for another phase of rapid expansion.

According to an Axis Capital report based on expert assessments, India's installed data centre capacity could rise from the current 1.3-1.4 GW to around 3-3.6 GW by 2030, driven largely by hyperscalers such as Amazon Web Services, Microsoft and Google, alongside a new generation of AI-focused cloud providers.

The report estimates that hyperscalers will account for nearly 90 per cent of data centre demand by 2030, up from about 60 per cent today, as artificial intelligence workloads and large language model hosting accelerate investment in computing infrastructure.

While industry announcements point to an ambitious target of 6-8 GW of capacity by the end of the decade, Axis Capital believes execution challenges—including power availability, equipment shortages, construction timelines and supply-chain bottlenecks—could keep actual operational capacity closer to 2.8-3.6 GW.

Building AI-ready facilities is also becoming more expensive. Traditional data centres cost around Rs 44-46.6 crore per MW, but AI-enabled facilities require denser computing infrastructure, pushing costs up by 10-15 per cent.

Even so, the long-term outlook remains robust. If announced projects are executed successfully, India's data centre capacity could touch 7-8 GW by 2035, creating significant opportunities across power, cooling systems, electrical equipment, construction and digital infrastructure.

Taken together, India's semiconductor manufacturing drive and the rapid build-out of AI-ready data centres represent two sides of the same story: one focused on producing the brains of the digital economy, the other on powering them. As artificial intelligence reshapes global technology, India is betting that chips and computing infrastructure will become twin pillars of its next growth chapter.

(With inputs from ANI)