India’s new semiconductor incentive scheme will offer homegrown startups a cushion to absorb the cost of producing prototype chips—a process that has often taken over a year and cost as much as ₹2,000 crore. To do this, the Centre will offer incentives as it waits for the country’s first commercial chip fab to come into operation in the country. In an interview with Mint on Thursday, Amitesh Sinha, additional secretary in the ministry of electronics and information technology (Meity) and chief executive officer (CEO) of the India Semiconductor Mission (ISM), said the move may help India create its own chip patent over the next five fiscal years. Edited excerpts:
Chip startups have rarely commercialized trials. Will Semicon 2.0 address this?
Yes. When small companies go to a commercial fab to try to tape out (physically build multiple copies of) their chips, they have to wait in a long queue for their turn. Further, there’s no discount for early-stage chip design startups, though large fabless chipmakers such as Nvidia and Qualcomm get up to 30% discounts in chip fabrication costs. About 30% of a chip’s cost is fabrication, for which big companies get a 30% discount, reducing their manufacturing cost by 9% overall. Under Semicon 2.0, we’ll offer Indian chip design startups this 9% as a deployment-linked incentive, so that early-stage ventures can absorb the cost impact when they are cash-strapped.
The US created billion-dollar startups like SambaNova and Groq. Is cost the only factor?
Time and chip fab agreements are also on our minds, and we’re exploring the possibility of signing a memorandum of understanding with strategic chip fabs to prioritise Indian startups. It’s not that India hasn’t done it before—domestic startups have made 12nm and 28nm trial chips from TSMC (Taiwan Semiconductor Manufacturing Co., the world’s largest chip fabrication company).
There are also chip fabrication aggregators in between, and we’re trying to see whether our startups can benefit from them, such as Belgium’s Imec (Interuniversity Microelectronics Centre). Domestically, we’ll have Tata’s Dholera fab operational by 2028. Once it is, startups can make chips much faster.
Semicon 2.0 also speaks of localizing machinery. Is this really possible?
The idea is to start somewhere and not aim right away to recreate a photolithography machine that only one company in the world makes. Right now, India’s semiconductor ecosystem and supply chain are just beginning to develop. As it matures, we will see more and more machinery companies setting up shop here in India. As they do, there will be local integrators, or the Indian entities of global machinery manufacturers, which will make it cost-effective to start manufacturing small subsystems and other infrastructure required in chip fabs and testing plants. That is why we are offering the incentives today: to help the ecosystem mature over the course of the mission.
India already has C-Dac (Centre for Development of Advanced Computing), a research body with chip design expertise. Why not leverage it?
C-Dac already plays a key role in the India Semiconductor Mission, serving as the nodal agency for validating applications from India’s chip design startups. Beyond this, C-Dac's development of the AI (artificial intelligence) chip remains independent of Semicon 2.0. With the latter, our core goal is to incentivize the right kind of private sector spending in research and development, which is crucial if India wants to create a fabless chip design company. Research grants for C-Dac are sourced separately.
Do you expect large firms to invest in chip R&D, then?
Absolutely. Meity will hold the equity and royalty investments through something akin to a special purpose vehicle. The royalty-based calculation is dedicated to large companies and conglomerates in India, which have significantly greater access to cash pools. The promise is that the Centre will match whatever capital investment they make in semiconductor design and development. There’s no equity involved here, since these companies won’t be in shortage of raising capital from private investors. The principle that we’re working with is that the Centre will take back 1.5x of what it invests as royalty, once the chip design in question starts generating revenue.