Blitz India Business
NEW DELHI: Every national semiconductor programme reaches a point where the celebratory milestone and the economic question diverge. India is approaching that point, and it is a better moment to ask the question than after the ribbon is cut. More than $21 billion of projects have been approved under the India Semiconductor Mission. The Tata–PSMC fabrication plant at Dholera is past the halfway mark in construction, targeting trial production in December 2026 at 50,000 wafer starts a month on 28-nanometre technology.
The value chain of a finished chip splits, very roughly, into four stages: design; front-end fabrication, where the wafer is patterned; assembly, test, marking and packaging, where the die is cut, mounted and made usable; and the equipment and materials that all three depend on. India’s position differs sharply across them. In design, India has been globally significant for two decades — a large share of the world’s chips carry engineering work done in Bengaluru, Hyderabad or Noida, even where no Indian company owns the product. In assembly and test, India is now physically present: Micron’s facility at Sanand was inaugurated in February 2026, Kaynes Semicon’s OSAT plant followed at the end of March with a roughly ₹3,300 crore investment and capacity around six million chips a day, and Tata Electronics is building at Jagiroad in Assam. In fabrication, India has nothing yet — Dholera will be the first. In equipment and materials, India has almost nothing at all.
Four stages, four different positions: India is strong in design, newly present in assembly and test, about to enter fabrication, and largely absent from equipment and materials — the stage with the highest and most durable margins.
Packaging is where the jobs are. Equipment and materials is where the margin is. A serious programme eventually has to want both.
At a Glance
• Approved under the mission: more than $21 billion of projects
• Dholera fab: Tata–PSMC, past halfway in construction, trial production targeted December 2026
• Design capacity: 50,000 wafer starts per month at 28 nm; site of about 66 hectares; roughly 21,000 jobs projected
• Assembly & test live: Micron at Sanand (inaugurated February 28, 2026); Kaynes Semicon OSAT (March 31, 2026, about ₹3,300 crore, around 6 million chips a day); Tata Electronics at Jagiroad, Assam
• Value chain: design · front-end fabrication · assembly, test and packaging · equipment and materials
• India’s gap: equipment, materials and specialty chemicals
Being clear-eyed about 28 nanometres is part of taking the programme seriously. It is not a leading-edge node, and it is not meant to be. It is the workhorse geometry for automotive controllers, power management, industrial sensors, display drivers and the large category of chips that go into appliances, meters and machinery — a market that is large, growing, price-competitive and, critically, one where supply security matters more to the buyer than transistor density. For a country whose electronics and automotive assembly base is expanding quickly, a domestic 28 nm supply is exactly the right first fab. Chasing a leading-edge node as a first project would have been a far riskier use of the same capital, and several countries have learned that lesson expensively.
Where the constructive agenda now lies is the fourth stage. Equipment, materials and specialty chemicals — photoresists, ultra-pure gases, wafer handling, metrology — carry the highest and most durable margins in the industry, and are the least contested by other emerging entrants because they require deep chemical and precision-engineering capability rather than capital alone. India has the chemical industry and the precision-engineering base to attempt some of it, and a domestic fab about to enter trial production creates, for the first time, a domestic customer to qualify against. Three steps would compound the advantage: a qualification programme that pairs Indian materials suppliers with the Dholera and Sanand lines from the start; sustained public funding for materials research at the institutes already working on it; and, most simply, a published multi-year demand forecast so that a chemicals firm can justify a plant. India is about to make its first chip. Making the things that make chips is the harder achievement, and the one that would still be paying in 2050.