The government is working on a plan to aggregate orders from Indian fabless semiconductor chip companies to improve their bargaining power with global foundries, said S. Krishnan, secretary at the Ministry of Electronics and Information Technology (MeitY).
This pooling of orders—a longstanding industry demand—is part of a series of measures the government is considering to help Indian semiconductor companies scale, attract more private capital and retain intellectual property in the country, even as they remain dependent on global foundries and suppliers.
The domestic semiconductor industry’s strength lies in designing chips that power most modern electronics, but for manufacturing they depend primarily on companies in Taiwan, South Korea and China that count some of the world’s largest tech companies including Nvidia Corp. and Apple Inc. among their clients.
That leaves Indian semiconductor companies with little clout to influence global chip manufacturers to prioritise their orders. “Our bargaining power with the foundries is not as great, because a big company will get precedence,” Krishnan said in an interview with Mint detailing several aspects of the government’s semiconductor policy.
Krishnan also said the government is rethinking conditions around foreign acquisitions of Indian chip companies, with the focus shifting from ownership to keeping the intellectual property (IP) in India. The government is not in favour of protectionist measures, he added.
Edited excerpts:
Where do you see the biggest gaps in India’s semiconductor ecosystem? Where should India be focusing or doubling down?
We all know that it’s a global value chain. It doesn’t sit in any one country. Geopolitics and the pandemic have changed how we think about semiconductor supply chains. The pandemic taught us that you can’t be reliant on a single source. You need to diversify. That is where the opportunity for India truly arises.
India needs to do it strategically because we can’t assume that we can rely on everybody else in the world for everything on an ongoing basis. This is what I would describe as strategic autonomy. I would not describe it as having to do everything ourselves. We have to have a significant enough share in certain segments of this ecosystem to make sure that we won’t be taken lightly.
China has built capacity in legacy nodes and the world has become reliant on one source of supply. You really need another big source of supply. That is where we believe India fits in. The investment required for legacy nodes is significantly lower, and the idea is to start with something tried and tested, where people can teach you, before going straight to the higher end.
The second part is design. Twenty percent of the world’s semiconductor design workforce is in India, but 95% works for global fabless companies (firms without their own fabrication or manufacturing plants). That’s good too, but it means there is very little Indian IP. If we want to become a product nation, we need Indian companies that are able to grow with their IP.
The third part is the ecosystem, including materials, gases, chemicals and testing. You can’t just build a fab (semiconductor chip fabrication plant) or OSAT (outsourced semiconductor assembly and test unit) and source everything from around the world. Some of these will need to be located closer to the facilities. This is why India Semiconductor Mission 2.0 also focuses on building the ecosystem.
But India’s fabless startups are, for the foreseeable future, dependent on a global manufacturing and supply-chain ecosystem, where they are seen as too small and do not have much bargaining power. Do you see a government or policy role there to help such startups?
Our bargaining power with [global] foundries is not as great because a big company will get precedence. The way we are trying to [approach] this is, in some ways, trying to sort of aggregate the demand… intended for overseas firms and see whether we can actually then negotiate on better terms.
Second, we have to look at diversification… We don't expect that we will have capacity in every area, ever. We then look for diversified sources of supply. If it’s not there, then maybe that’s an area we will enter in the future.
Has work already started on this consolidation of demand?
We are working on this. Many of the DLI companies (firms selected for the government’s design-linked incentive scheme) are only now coming to that stage. And this is a request that they have raised with us. This consolidation of demand is an approach we would like to take.
As fabless chip companies reach production, they are now talking about commercialization. Do you see a role for the government here?
We are not a closed economy. We are not going to say you should use only chips made here. Also, anything that we make should be part of the global value chain, right? It’s not meant for just the Indian market… If you’re not able to produce for the export market, then you’re not being fair to your own country.
Which is why whatever we support will have to be globally competitive. They may need some support to reach scale before they become globally competitive. That we can support, but in the long term there has to be a roadmap by which time they will be able to manage without specific government support, whether in terms of buying support, procurement support or subsidy support. They should be able to sell in the global market competitively. I think that is a non-negotiable.
Is there then a role for government standards or procurement requirements to create that initial market?
There is a lesson we learned from our experience with CCTV cameras. CCTV cameras used even in high-security locations were using chips which come from countries that are not very friendly with India, which means they could be potentially backdoored, essentially the data could be going somewhere else.
So, from a security perspective, we brought in a certain set of guidelines around a trusted value chain from which you have to get these cameras. Incidentally, it happened to be an opportunity for some of our DLI companies. When you’re looking for a chip which comes from a trusted value chain, these DLI players also got to play. We’ve learned something from that.
We are exploring whether a similar approach can be used in other areas with a security angle, such as smart meters, IoT (internet-of-things) devices, industrial IoT devices, and e-passports. There’s an opportunity there to build similar standards and then you have the possibility that an Indian chip company could benefit.
If you look at the long term, from a semiconductor ecosystem in India point of view, what would you call success?
I think we should have an adequate number of fabs (chip fabrication plants) of different types. We are right now supporting up to 28 nanometer, but we should eventually build capability for more cutting-edge nodes. We should have a robust research system which keeps pushing in that direction.
We should have a packaging industry which sufficiently meets our requirements in certain categories and is also able to export. We should have the entire ecosystem in terms of manufacture of equipment, manufacture of materials, chemicals and energy. It’s not just for India, but for the world.
We should have a robust system which will take care of an adequate part of India’s needs directly, while also playing an important role in the global value chain, so that we can never be ignored in that process.