India’s semiconductor industry: India has money behind its semiconductor ambitions. What it lacks is the accumulated industrial knowledge that turns investment into a functioning technology ecosystem. The government has put together an elaborate policy architecture to build a domestic semiconductor industry, covering fabrication and assembly, chip design, equipment, research and talent. The second phase of the India Semiconductor Mission, approved this year with an outlay of ₹1.275 lakh crore, is intended to take the programme beyond the first phase, which focused largely on getting projects off the ground.
The difficult question is whether India can build the capabilities needed to make those investments commercially and technologically durable.
READ | Success of semiconductor mission depends on ground-level delivery
India has considerable strengths in software, engineering and chip design. But it has limited experience of high-volume semiconductor manufacturing. Taiwan has spent decades building its foundry industry. South Korea has developed formidable capabilities in memory chips and advanced manufacturing. Japan remains important in semiconductor materials and equipment. The US dominates several parts of the technology chain, including advanced chip design. China has invested heavily across semiconductor manufacturing, equipment and AI hardware.
India is trying to build several of these capabilities at once. That makes the shortage of accumulated knowledge more important than the size of the subsidy on offer.
A fab cannot be built with capital alone
Semiconductors are an industry in which capital does not automatically create expertise. A fab is not simply an expensive factory. It is a tightly integrated system involving process engineering, materials science, equipment maintenance, chip design, packaging, testing and quality control.
The manufacturing process is unforgiving. Small errors in a production process can affect yields and make an entire batch uneconomic. Knowing semiconductor theory is therefore very different from knowing how to operate and troubleshoot a semiconductor facility.
This is one reason India’s semiconductor challenge is different from the one it faced in building its IT services industry. A software company could hire programmers and add capacity without reproducing an industrial production environment for every new employee. Semiconductor manufacturing requires specialised equipment, controlled environments and production processes that workers must learn to handle.
India therefore needs a workforce pyramid, not simply more highly qualified engineers. Chip architects, process engineers and researchers are important, but so are technicians, equipment specialists, quality-control professionals and operators who keep fabs and assembly and testing facilities running.
The government says the first phase of its Chips to Startup programme has already met its target of training 85,000 engineers in semiconductor design, four years into a programme that had initially envisaged 10 years. Semicon 2.0 aims to train another 100,000 engineers over six years.
Those figures indicate the scale of the government’s effort. They do not, however, measure manufacturing experience. An engineer trained in chip design is not necessarily a process engineer who has worked on improving yields inside a fab. A classroom course can provide technical foundations, but it cannot reproduce the knowledge acquired by solving problems on an operating production line.
That distinction is central to India’s semiconductor ambitions.
READ | India Semiconductor Mission 2.0 faces the import test
India’s semiconductor industry: Lessons from Taiwan and Singapore
The proposal to bring foreign research institutions such as Taiwan’s Industrial Technology Research Institute, or ITRI, and Singapore’s Agency for Science, Technology and Research, A*STAR, into India’s talent-development effort is therefore significant.
ITRI has an unusually relevant history. Established in 1973, it played an important role in developing Taiwan’s technology capabilities and helped lay the foundations for its semiconductor industry. One of its early achievements was the transfer of integrated-circuit technology that contributed to the creation of United Microelectronics Corporation. Taiwan’s semiconductor success was not produced by subsidies alone. Research institutions such as ITRI helped create the technical capabilities and industrial links on which the industry could grow.
India cannot reproduce that history. It can learn from the institutional model behind it.
If ITRI can bring some of the knowledge accumulated through decades of research and industrial collaboration into Indian training centres, its value will extend well beyond the number of people who complete a course. The more important outcome would be the creation of Indian engineers and institutions capable of solving manufacturing problems without having to depend indefinitely on outside expertise.
A*STAR offers a somewhat different example. Singapore has developed a significant semiconductor industry despite having a small domestic market. Its research institutions have built close links with global manufacturers and technology companies, while developing capabilities in areas such as advanced packaging, silicon photonics and semiconductor-related research.
The experience of Singapore is relevant to India for another reason. A country does not need a huge domestic market to become important in semiconductors. It needs specialised capabilities that fit into the global production chain.
Germany’s Fraunhofer-Gesellschaft could offer another useful model. Its network of applied-research institutes has long connected scientific research with industrial applications. Such institutions are particularly relevant to India’s challenge because the gap is not simply between universities and companies. It is between theoretical knowledge and the ability to solve problems in an industrial setting.
READ | India Semiconductor Mission enters its hardest phase
Foreign partnerships must have an Indian end point
There is a legitimate question about what these foreign institutions gain from working in India.
The answer is access to a semiconductor market that the government is trying to build at considerable cost. India is supporting fabs, packaging facilities, design centres, equipment capabilities and research. Institutions such as ITRI and A*STAR would gain relationships with Indian companies, universities and technology organisations as the ecosystem expands.
India should use that interest to its advantage.
Foreign institutions should not become permanent providers of expertise to subsidised Indian facilities. Their role should be to help create Indian capabilities that remain after the partnerships mature. That means training programmes need to be connected to actual manufacturing facilities, equipment suppliers, universities and domestic research institutions.
The test should therefore be whether an arrangement leaves India with stronger institutions and a deeper pool of people who can train others, solve production problems and develop new processes.
Self-reliance in semiconductors cannot mean producing everything domestically. Semiconductor manufacturing itself is global, with different countries specialising in design, equipment, materials, fabrication, packaging and other parts of the value chain. For India, self-reliance should mean having enough domestic capability to make strategic choices and participate in that chain from a position of technological competence.
India needs to choose where it can build depth
India is entering the semiconductor industry late. Taiwan, South Korea, Japan, the US and European countries have spent decades developing specialised capabilities. China has invested enormous resources in trying to close the gap.
India cannot compress all that accumulated experience into a few years through incentives.
It can, however, decide where to concentrate its efforts.
Mature-node chips, automotive semiconductors, power electronics, advanced packaging and testing could offer more realistic opportunities for building industrial depth than an immediate attempt to compete at the leading edge of logic-chip manufacturing. Success in these areas would also create experience, supplier networks and engineering capabilities that could support more ambitious goals later.
The temptation in a national technology programme is to measure progress by the number of projects approved, the money committed or the engineers trained. Those are useful indicators, but they are not the same as technological capability.
The harder measure will be whether Indian engineers can run a production line, improve its yield, solve equipment problems, develop processes and eventually teach others to do the same.
The easier part of the semiconductor mission was getting projects started. The harder part is building the knowledge base around them.
Bringing ITRI, A*STAR and potentially Fraunhofer into that effort could help India shorten the learning curve. The value of these partnerships will ultimately depend on whether the knowledge they bring becomes embedded in Indian institutions and companies.