India needs a 20-year semiconductor strategy
At the inaugural Sydney Dialogue, Prime Minister (PM) Narendra Modi emphasised that semiconductors will be a crucial layer in India’s technology stack. Hinting at an upcoming policy specifically for semiconductor manufacturing units (fabs), he said, “We are preparing a package of incentives to become a key manufacturer of semiconductors”.
India has long been a major global centre for chip design. At the same time, it has failed to attract firms involved in the two subsequent stages of chip production — chip fabrication and chip assembly, testing, marking, and packaging (ATMP). Of the two, chip fabrication is the more complex, capital-intensive, and skill-intensive stage. At least four previous attempts to start such fabs in India have fallen flat. PM Modi’s latest statement suggests that the government is expecting different results on the fifth try, perhaps because it has decided to commit a much higher quantum of public funding.
To succeed, it is essential to reflect on the difference between being able to manufacture one line of chips and achieving semiconductor self-sufficiency or even becoming a key manufacturer. The current discourse masks this difference. The dominant narrative suggests that India is in a do or die situation, one in which building a fab now implies the elimination of critical strategic vulnerabilities. In contrast, another failure means India is resigned to a position of weakness in the information age. This understanding is misplaced. Getting one fab going will not make India a key manufacturer. We will still be dependent on manufacturing equipment, designs, and chips manufactured outside India.
India needs to drop the dream of swadeshi semiconductors. Instead, it should aim to become a key player in a trusted, plurilateral semiconductor ecosystem that keeps key adversaries out. In our view, at least five specific parts of the puzzle need to fall in place.
First, India needs one 20-year semiconductor roadmap, not 20 one-year plans. The Centre should begin by auditing the origins of the chips that form the core of key defence equipment and critical infrastructure. After this vulnerability assessment, the government needs to ensure that such equipment should have chips produced end-to-end within the trusted semiconductor ecosystem over time.
Second, the 20-year roadmap needs a 20-year financial support plan. The roadmap needs to sequence semiconductor initiatives depending on the government’s financial wherewithal. For instance, focusing on getting a leading chip assembly player to India can be considered immediately, at a lower monetary and opportunity cost. Co-investing in a chip production unit at a trailing-edge, speciality chip fab unit can become the government’s next big step. Concurrently, the government can fund new semiconductor materials research, new design architectures for critical equipment, intellectual property protection, and technical standards. Over a two-decade period, this can well give the confidence to global investors to co-invest in a leading-edge fab here.
Third, plurilateral strategic cooperation on semiconductors is a necessity for India, not a choice. The Quad Semiconductor Supply Chain Initiative is a good starting point. India needs to push for a Quad Supply Chain Resilience Fund to immunise the supply chain from geopolitical and geographic risks. While the United States (US) restarts manufacturing at leading-edge nodes (5 nanometres and below), Quad should fund specialised trailing-edge fabs (45 nanometres and above) in India, Japan or Australia. Over time, this initiative can coalesce other major semiconductor powers such as Taiwan, South Korea, European Union, and Israel.
Fourth, favourable trade policies are critical for building a plurilateral semiconductor ecosystem. Over the past few years, the Centre has been increasing import duties. Such policies have significant implications for the semiconductor industry’s prospects. For example, even Taiwan, which produces over 50% of the world’s contract-manufactured chips, needs to import specialised equipment. Unsurprisingly, tariff reduction was reportedly a major issue of discussion between India and Taiwan over a semiconductor collaboration earlier this year. Moreover, considering the small number of electronics manufacturers in India, fabs in India will primarily be exporting their products. In essence, a fab in India will still be deeply connected with the world — buying equipment from some countries and selling chips to others.
Finally, the 20-year roadmap needs a robust infrastructure plan. Significant quantities of reliable water and electricity supplies are non-negotiable requirements for fabs. It is not possible to meet these specialised requirements across the country. Hence, the Centre must work with a few interested state governments to build the necessary infrastructure over time.
India’s comparative advantage lies in semiconductor design. India is a global hub for chip design and design services. Doubling down on this comparative advantage should be an essential element of India’s 20-year roadmap. Reducing barriers for technology exchange, joint product development, visitation and research participation, can go a long way.
It took decades for Taiwan, South Korea, and Japan to become semiconductor powers. Their success was based not just on government investment but a confluence of the five factors mentioned above. India, too, needs to play the long game.
INDIA’S ADVANTAGE LIES IN SEMICONDUCTOR DESIGN. INDIA IS A HUB FOR CHIP DESIGN AND DESIGN SERVICES. DOUBLING DOWN ON THIS SHOULD BE AN ESSENTIAL ELEMENT OF INDIA’S 20-YEAR ROADMAP