On a base of 1.5 gigawatts of installed capacity and a national plan to reach six to eight, the French infrastructure group sees its India unit becoming its single biggest business within five years.

Schneider Electric expects its India data-centre business to grow faster than the rest of the company, and faster than the core electrification and automation businesses that have driven its results everywhere else.

Deepak Sharma, the group’s managing director and zone president for Greater India, told Reuters on Monday that the unit could become Schneider’s single largest business within three to five years, on the back of the country’s planned scale-up from roughly 1.5 gigawatts of installed data-centre capacity to between six and eight.

The framing is consistent with what Sharma told Indian outlets earlier this spring. In an exclusive interview with BusinessToday in April, he described the opportunity as “exponential”, noting that data centres are not yet Schneider’s biggest line of business in India but that generation, data centres, and homes look set to lead the next growth cycle.

Globally, roughly 30% of Schneider’s revenue, which the company reports at about €40bn a year, already comes from data centres.

The India backdrop is unusual in the same way the country’s broader AI infrastructure story has become unusual. India produces and consumes around 20% of the world’s data while housing only 3% of global data-centre capacity, a gap that has set off a wave of hyperscaler and domestic commitments: Google’s $15bn AI hub in Visakhapatnam, Adani Group’s $100bn ten-year build-out, and Microsoft’s and Amazon’s multi-tens-of-billions India programmes.

Larsen & Toubro announced earlier this year that it would partner with Nvidia to build a sovereign AI factory in Chennai. Each of those builds is a Schneider customer, or could be.

Schneider has spent the last year arranging itself for that. It bought out the remaining 35% stake in its India subsidiary, SEIPL, from Temasek for €5.5bn last year, a move the company said would speed up decision-making locally.

India is now its third-largest market, with roughly 38,000 employees, 31 factories, and exports to more than 30 countries including the United States. The company has flagged that the buyout will add about €150m to its 2026 financing costs.

The wider numbers explain the urgency on both sides. Schneider’s Q1 2026 revenue rose 11.2% organically to €9.77bn, with the energy-management segment, which supplies data centre power and cooling, up nearly 13%.

The combined 2026 AI capital-expenditure programme of the major US hyperscalers is now expected to exceed $650bn, with much of it landing as long-dated orders for the kind of grid-to-rack power equipment Schneider sells.

India’s particular role in that flow is partly geography and partly arithmetic. Deployment cost per megawatt sits at least 30% below the global average, Sharma said, which has pushed builders out of the traditional Mumbai/Chennai/Delhi/Bengaluru cluster and into tier-two and tier-three cities.

Edge sites, designed for the low-latency inference workloads that generative AI keeps producing, are where the next wave of capacity is going.

The constraints are familiar from elsewhere. Sharma flagged copper and silver costs, both of which Schneider’s products use at volume, as a continuing pressure that the company has been passing through to the market.

A prolonged conflict in West Asia, he added, would compound that on plastics and transportation. Power supply, the harder constraint behind every AI data-centre announcement, was not addressed directly.

What Schneider is selling, in effect, is the assumption that India’s 1.5 gigawatts becomes six to eight inside the decade, that most of it gets built, and that the bill of materials runs through equipment vendors like itself. The next several quarters will show whether the orders match the announcements.