India has spent years announcing semiconductor ambition in rooms full of scale models. Dholera is where the PowerPoint must finally meet chemistry, yield and factory discipline. Tata Electronics’ planned fabrication plant is important precisely because chips are unforgiving: patriotic enthusiasm cannot improve a defective wafer.
The project has credible technology partners and a large public commitment. Reporting on ASML’s agreement with Tata Electronics describes a 300mm facility using process technology licensed from Taiwan’s PSMC, with mature nodes rather than a theatrical attempt to manufacture the world’s most advanced chips immediately. That is the correct starting point.
Mature nodes are not backward
Technology coverage trains consumers to think only the smallest transistor matters. Cars, power systems, industrial equipment, appliances and communications infrastructure use enormous quantities of mature-node chips. Reliability, cost and long product lives often matter more than packing maximum computing power into a tiny area.
India’s automotive and electronics industries need precisely these components. Domestic fabrication can reduce exposure to geopolitical disruption, but resilience is not autarky. Chemicals, gases, tools, software and specialised parts will continue crossing borders. The goal should be a competitive node in a global network, not an isolated national island.
The government’s broader semiconductor programme recognises the scale of the challenge. Its electronics ministry reporting places semiconductor development inside a larger manufacturing ecosystem. That connection is essential. A fab without customers, packaging, testing and design capability becomes an expensive monument.
I give the government credit for accepting that markets alone would not create the first plant. Every major semiconductor geography grew through public support, defence demand, infrastructure or coordinated finance. Complaining that India is subsidising its entry while established powers subsidise expansion is economically naïve.
The real product is knowledge
The first wafers will attract cameras. I care more about what happens after defects appear. A fab learns by improving yield: diagnosing contamination, tuning equipment, refining recipes and coordinating suppliers. That tacit knowledge lives in teams and routines. It cannot be purchased once and installed like a server.
India should therefore measure engineer retention, training hours, supplier qualification and process improvement. If foreign specialists solve every difficult problem and leave, the plant may produce chips without building national capability. Partnerships are necessary; dependency without learning is not.
Universities must update curricula with industry rather than simply renaming electronics courses “semiconductor programmes.” Students need clean-room practice, materials science, statistical process control and equipment maintenance. Technician training is equally important. A fab employs highly qualified engineers, but it also depends on disciplined operational roles rarely celebrated at technology conferences.
Water and power deserve unglamorous attention. Semiconductor plants demand stable electricity and sophisticated water systems. Dholera’s infrastructure must be audited transparently, including recycling and environmental performance. Communities should not discover the true resource cost after production begins.
Chinese dominance will not vanish
India’s phone market remains dominated by Chinese brands because those firms combine supply-chain speed, price discipline and retail execution. A domestic fab will not suddenly create an Indian smartphone champion. Nor will it make every component local. The semiconductor value chain contains thousands of specialised dependencies.
What it can do is anchor suppliers and create credible demand for Indian design. Packaging and testing firms may locate nearby. Equipment-service teams can deepen. Local companies may design power-management, display or automotive chips knowing a domestic manufacturing relationship is possible. Capability accumulates around repeated transactions.
Policy should avoid forcing Indian electronics firms to buy locally regardless of price or quality. Captive demand can protect inefficiency. Better tools include long-term procurement visibility, co-development grants, shared research facilities and incentives tied to globally competitive output.
The biggest risk is announcement inflation. India has seen proposed fabs change partners, locations and timelines. Government communication should separate approved, financed, under-construction and producing projects. Treating each stage as a finished achievement weakens trust.
Publish what success means
The public is financing a substantial portion of India’s semiconductor push. It deserves milestones that reveal progress without exposing commercial secrets: construction status, equipment installation, workforce training, trial production, yield bands and qualified product categories. A first-wafer ceremony should be the start of reporting, not the end.
The counterargument is that such disclosure may embarrass a strategic project during inevitable delays. I prefer managed embarrassment to hidden failure. Chip manufacturing is difficult; reasonable citizens can understand setbacks if targets were honest. What they will not tolerate forever is a sequence of triumphant announcements without commercial output.
Dholera should also be judged over a decade. The first facility may not deliver ideal returns, because some public value lies in skills and ecosystem formation. But “strategic” cannot become a magic word protecting every cost. Later investment should depend on demonstrated learning and market demand.
I am optimistic because the approach is finally specific: a real company, a defined site, equipment relationships and mature processes with clear markets. I remain cautious because India excels at inaugurating ambitions and underinvesting in the connective tissue.
The wafer is not sovereignty. The supplier who can repair a tool at midnight, the engineer who improves yield and the Indian designer whose chip sells abroad are closer to it. If Dholera creates those people and firms, Make in India will have moved from assembly to capability. That is the standard worth funding.
