Build the ecosystem alongside the fab, not after it

India's semiconductor story is usually told in fabs: how many have been approved, how much they cost, when the first wafers will come out. That is the right place to start. It is the wrong place to stop. The countries that built lasting chip industries did not succeed because they built fabs. They succeeded because they built everything around the fabs at the same time.

What the winners have in common

Hsinchu grew up with hundreds of local suppliers of materials, chemicals, parts and equipment service within easy reach of TSMC. When something goes wrong on a line, the supplier's engineer can be on site the same day. Dresden's "Silicon Saxony" grew over decades into a network of hundreds of supporting companies around a few anchor fabs, and that network is now one of the reasons new fab investment keeps landing there. Korea's experience after Japan's 2019 export controls on hydrogen fluoride, photoresist and fluorinated polyimide showed the other side: a strong fab base exposed by a thin local materials base, and an expensive scramble to fix it.

None of these clusters retrofitted their supplier base after the fabs were running. The ecosystem and the fabs grew together, and the advantage compounded into faster yield ramps, lower costs and more resilience.

Where India is today

India has committed the capital. The first phase of the India Semiconductor Mission approved 12 manufacturing projects with more than ₹1.64 lakh crore of investment, and in July 2026 the Union Cabinet approved a second phase with a ₹1.27 lakh crore outlay. Assembly plants are already producing. The first commercial fab is under construction at Dholera.

The layer in between is much thinner. By our estimate, well under a tenth of the high-purity chemicals, specialty gases, quartz parts, reclaimed wafers, equipment service and specialist technicians that these plants will need is available in India today. Most of it will arrive by air and sea, at least at first.

A fab can be built in two years. A supplier ecosystem takes ten. That is why the ecosystem has to start now.

Why the sequence matters

Building the ecosystem after the fabs are running is more expensive in three ways.

• Ramp speed. A new fab's economics depend heavily on how fast its yield rises. Every problem that needs a part, a chemical lot or a service engineer from overseas slows that ramp by days or weeks.
• Qualification lock-in. Once a fab qualifies a supplier for a material or service, switching is slow and risky. Suppliers who are not in India when the qualification decisions are made may wait years for the next chance.
• Cost structure. Imported consumables carry freight, duty, inventory and risk premiums that a local supplier does not. For mature-node fabs competing on cost, that difference is significant.

What would help most

Three moves would make the biggest difference over the next five years.

1. Treat suppliers as part of the mission. Incentives for chemicals, gases, quartz, reclaim and service companies should be as easy to understand and access as the fab incentives, and timed to the fabs' qualification windows.
2. Co-locate. Supplier parks next to the fab and OSAT clusters in Gujarat, Assam and Uttar Pradesh, with shared utilities, waste treatment and bonded logistics, would shorten response times and lower entry costs for each supplier.
3. Pair Indian industry with global technology. India has strong chemical, engineering and industrial companies. Joint ventures that combine their capacity with proven semiconductor processes from global technology holders are the fastest way to reach semiconductor-grade quality.

Our view

The fabs are the anchor, and they are coming. The question now is whether the supporting ecosystem arrives in time to make them competitive. That work is unglamorous and distributed across dozens of companies, which is exactly why it needs deliberate attention. It is the work Mantra was set up to do.