Will Be In Trouble If We Continue To Import Everything: HCL Co-Founder Ajai Chowdhry On Manufacturing Semiconductors In India

With a financial outlay of Rs 76,000 crore, the Ministry of Electronics and Information Technology (MeitY) in 2021 launched the India Semiconductor Mission (ISM) with an aim to make India a "global hub" of semiconductor manufacturing. The Centre expects the semiconductor industry to be worth $63 billion by 2026. 

ISM is a specialised business division of the Digital India Corporation with an aim to build a vibrant semiconductor and display ecosystem to enable India's emergence as a global hub for electronics manufacturing and design. 

This isn't the first time India approached the idea of semiconductor manufacturing units. In 2013, under the Congress-led UPA government, IBM and STMicroelectronics had proposed to build semiconductor wafer plants in India at a cost of $8 million then - a plan that didn't take off.  

Fundamental to technologies, semiconductors are also at the core of tensions between the US and China after Joe Biden's administration blocked exports of advanced computer chips to China in 2022. Against the backdrop of these developments, India has the potential to attract semiconductor manufacturers and even become a hub. 

To understand how semiconductor manufacturing can benefit India and what is India's potential in the domain, The Core's founder and editor Govindraj Ethiraj spoke to Ajai Chowdhry, a member of the advisory board of ISM. He is also the co-founder of HCL and has been closely involved in the setting up of the computer hardware industry in the country.  

"India is a very large country. And we have a huge market for electronics. Now, with this size of demand for electronics, if we continue to import everything then we are going to be in trouble at some stage or the other," Chowdhry said on the importance of manufacturing semiconductors in India.  

Here are the edited excerpts from the interview:

What is a semiconductor?

Semiconductors are pretty much the heart of any computer or computing device whether it is a phone, tablet or any desktop or whatever. And around that microprocessor or semiconductor are other semiconductors and components that make up the whole product. This is the heart and this is where the power of the product lies.

Semiconductors which are really chips and very small, could go into an iPhone14 or an F35 fighter jet, my laptop, my refrigerator or a toaster. What is the difference between the chips?

For general applications, like washing machines and cars, etc, the type of chips that are required are what are called higher nanometer chips - 28/20 nanometers and 40 nanometers and above. For the phone where the product is really small and needs a lot of power, you go down to 5 nanometers or 3 nanometers. So, that is really where the differentiation comes. Our country majorly needs the higher nodes, as it is called. In India, a lot of people think that there is a demand for the most advanced chips. The problem is that we do not make such products today. Everything is important and therefore, if you have a demand. For example, for phones, the phone chip is not integrated in India. Everything comes integrated in a motherboard and all we do in this country is put it all together and brand it and sell it.

If we want to take a further step back, we want to manufacture a semiconductor in India catering to both or only the less advanced products.

I think we should not aim at spending effort, time and money on the lower end at all because that demand for India is very low. Our real demand sits in the higher nodes, which is 28/20 nanometers, 40 nanometers etc. And the second reason is that the investments for 28 nanometers and 40 nanometers are much lower than the latest technology for phones and other products.So I think we should not bother about that to start with because our requirements are what we should look at .

So why would we want to manufacture these semiconductor chips in India?

India is a very large country and we have a huge market for electronics. Now, with this size of demand for electronics, if we continue to import everything then we are going to be in trouble at some stage or the other. You look at Covid. During Covid, cars could not be produced in India. Close to 250 industries in India were impacted by the shortage of semiconductors.So, I think we cannot allow that to happen in the future at all.

The second reason is that the supply chains have become very different now. Everybody is talking about more diversified supply chains and less risky supply chains, and more trusted sources. And this is where the opportunity for India is because till now China was dominating in a lot of these areas. Now is the opportunity for India to step in and get the benefit of  being a trusted source and also a very reliable source for all over the world.

Many years ago, the chairperson of America's Intelligence Committee said that the next war will be fought with semiconductors. So I think that we should be cognizant of this fact that we need to be in this. We are a very large country. The point is we are a big designer of chips in India but do not make anything and that is the dichotomy because we have a large force of people who can design chips but we do not do any manufacturing in it. So, we just cannot leave it for others to do. We need to have our own-you never know when there is a crisis or a war, you will have to have your local supplies very much intact.

Also Read: ChatGPT's Impact: An In-Depth Conversation on AI's Role in the Future of Work and Its Potential Dangers

But to your point on designing, these are being designed by companies in captive units like Texas Instruments, for example. Isn't it?

Not really. There are organisations like HCL, Infosys, Wipro,Infosys, TCS and many other engineering research and development (ER&D) companies. There are the GCCs (global capability centres) which are the global companies that are designing in India, but there are Indian companies also designing for the world. If you look at the ER&D portion of NASSCOM sales, it is around 38 billion dollars. That really does not give you the right picture of the revenue because a lot of that is done by GCCs and they only reflect the transfer price. So, interestingly Intel designs a lot of products in India. Texas Instruments designs a lot of products in India. But nobody knows that those products are made in India because the IP resides abroad and that is the missing piece as far as we are concerned as a country. So the time has come for us to have both - our own design to manufacturing capability as far as the semiconductors are concerned.

You talked about nodes which you are saying is more appropriate to India and India's needs. Now, India buys maybe 150 million phones every year and more. That would be among the major electronic devices that it is buying. Isn't it? So when you say India's needs, are you talking about other appliances?

You tell me one product that does not have electronics. Maybe a toothbrush - but even the most modern toothbrush has electronics, the electric ones. So, pretty much it is there everywhere, it is ubiquitous, it is a meta resource and therefore we need to be there in every part of the game on electronics.

You talked about the strategic importance of chips. Would that chip, the kind of chip that would go in an appliance or kitchen appliance, be as strategically important as a chip that goes into a mobile phone or let us say some defence device or a fighter jet?

Look it is a range - for defence devices, aerospace and nuclear we already have semiconductor complexes, they are doing chip design and chip manufacturing in low volume for them already. So, for strategic electronics we do not need anything better. What we need is for the volume business. So, volume sits in the refrigerators, ovens, cars and all those appliances and that volume is where we should focus on all our attention. Because we do not need to initially start at the most advanced stage because putting up a plant of that technology will cost 20 billion dollars. We should not be in that area. We should be in the area where we can genuinely get the best ROI as far as initial investments go. So, USD5 billion will typically be the size for a 40 nanometer node plan.  And that is really where we should aim. And that is one part of the business. In addition to silicon there is a requirement in India for compound semiconductors too, so we should focus there as well.

What is the difference between the two?

See, the compound semiconductor is Gallium Nitride. It is a mix of chemicals that actually produce the semiconductor. Silicon chips come from silicon that comes from sand. So it is a completely different technology. The investments in compound semiconductors are much lower. Typically Rs 200 million or Rs 300 million rupees are good enough for setting up a compound semiconductor. Now, where are the compound semiconductors typically used today? They are typically used in power electronics, for example. So, for making a charger for your phone or making chargers for EVs or making power supplies for different products–that is where a compound semiconductor is used.

We are looking at a situation where we have had one Expression of Interest-Vedanta, Foxconn which is also a Taiwanese company. Now, that has not gone through because the government has sent it back saying: for what you want to manufacture, you need a technical partner. So, explain to us what that means?

The challenge that the country is facing is, we are now pretty much giving almost 70% of the capex - Center 50%, from the State 20%. So, when 70% of the money for putting up a plant is being put in by the Indian government, it has to be very sure that the combination of management quality, funding, technology and partnership is all taken care of. If any of these things are missing, the government has a major challenge taking a decision. And that is why decisions are not happening. The second reason, each of these four points might be taken care of, the fifth point is also very important - which is having a good business plan. This business plan should have a clear strategy of saying where my customers are going to come from. So, all this needs to be put together by anybody who applies. Then the application will be looked at by the ISM, and within ISM  there are backend people who are looking at it and there are close to 5-6 very very well known global Indians who evaluate each project. And they then go into such detail, where a lot of people who applied in the past could not answer their questions. So, I think it is the right approach that we are taking because if we are going to spend this kind of money, we do not want a problem tomorrow in the Parliament when somebody says we spent 70% and nothing came of it. We cannot afford to do that. So, I think the government is absolutely right in taking its time over it. Let us not rush these things and these things are not to be done overnight. After all, we have lost 30 years, we can afford to lose another six months.

If the subsidy or the incentive has to be up to 70%, then it also suggests or reflects that fact that there is no real business case in doing so for the entrepreneur or the business house.

Incentives are given by many many countries. In the past we never got a semiconductor plant going because our incentives were very poor. And the second reason why we could not get it all going was that the incentives were to be paid post investment. Here, what the government is saying is that we will pay pre-investment that is why it is a much more risky proposition. And there is a big challenge in the world today. Every country is upgrading its semiconductor capability because everyone has gone through this disruption of supply chains. So, America is very clear that they want the latest and they want huge volumes but they are giving only 30% incentive. Europe is also doing the same. Japan has invited TSMC (Taiwan Semiconductor Manufacturing Company Limited) to set up there. Everybody in the world today is wanting a semiconductor industry. So, in this environment where do you go as a country? You need to position yourself very differently. And therefore your positioning is we want an older node, we will give you 70% if you manufacture in India and we want to make sure that you are successful and we will ensure that we will get the business to be successful. So, I think all of this is a total policy of the ISM.

Two companies, Intel and TSMC, are also building very large plants in Arizona right now. That is going to be on stream very soon on the model you just talked about. So why aren't these companies particularly Intel, that has been in India for such a long time and designing and doing so much of development and manufacturing here at this point, and why isn't TSMC looking at India as an option?

TSMC has too much business. They do not need to come to India and in India there are no customers for TSMC. So there is no need for them to look at India. And as far as Intel is concerned, Intel is reworking its complete strategy. Intel was originally a micro processor company and they made the chips for themselves. They are now moving into the services business which is what TSMC has occupied. So that is the reason why Intel is aggressively going after a completely new market. See because the PC business is gone in a lot of ways. What is left is the phone business. They are nobody in the phone business – everything is MediaTek, everything is Arm Technology. So, where is the position for Intel left? So they need to reinvent themselves. You look at their numbers and the lack of profitability and the questions that are raised around Intel today. There are many many questions

If you were to manufacture in India, the proposition that was there or the proposal that was there was to set up in Gujarat. So, tell us how we stand or fare specifically in Gujarat or in India when it comes to all the other factors which are bringing the material in and the finished product going out, if it has to go out?

I am of the view that the government along with the state governments should work together to create a single area for semiconductors. There should be just one area where we put our semiconductor plants to start with so that we do not need to replicate everything that is to be done in every plant, so that we do not have to worry about all the supply chains for all the materials that are required to run a semiconductor plant. 

What is very critical is that these locations that we identify should be close to an airport so that in and out of material is very quick. And logistic cost in this country has been traditionally very high and in electronics where you are talking of a dollar or a 50 cent product, if you have a 16% cost of transit or movement of goods, it is too high. I mean if we take a decision that it is going to be Gujarat let it be there for the moment and let them develop a complete ecosystem end to end for the semiconductor capability and in the present situation where India is today, we have one hole in our total strategy of semiconductor - which we do not have today,  is the R&D capability for developing the process for manufacturing. And I think that capability was identified in ISM and recently a report was created for creating the right resources for the country and the whole world. 

So to make these chips you need to get machines from what appears to me only one company in the world, a European company called ASM Lithography. Would we be also importing from them?

No. There are some companies in the US also but each of them do different things. So for Lithography, yes, that is the company. There are only 3 to 4 companies that supply equipment for manufacturing which is why I am saying that do not worry about trying to get the latest because those machines you will never get-that will get prioritised for the Europeans and the Americans and the Japanese. Do not bother about that and go for the higher node products and in the higher nodes you can even get a second hand facility. So, you could get, for example, Samsung has a 40-node facility in Korea. If Samsung wants to bring it here, welcome them to bring it here. That is really the way we should look at it.

Samsung is a chip producer. They too are not coming to India, at least for doing this.

You see today a lot of them will not come to India because there is so much demand globally. So they will go to where the customers are. One area which is definitely open for India, where I expect that you should be seeing some definite activity is the memory area. Memory customers are abundant in this country and all the phone companies can buy memory locally. So, I think that is where someone should look at the whole incentive that we are giving. Look at it very seriously and say that yes we should come and invest for memory in India. And that for me is a very important area for us to attract people and come and invest here.

In your understanding or estimation today, where could this take us, assuming the first plant was to come up in three years time?

You see, semiconductors alone are not going to give the answer that you are looking for in terms of value addition but semiconductors are an important part of that. What we need to do is that as we create the semiconductor ecosystem in India, we should also create the electronic product ecosystem in India because Indian brands have gotten lost in the last 15-20 years. We need to bring the glory of Indian brands back and create new Indian brands, for example BOAT That is a new Indian brand and they can be scaled. They can design in India, buy components from India, buy microprocessors from India - all those things can happen tomorrow. So it is very critical that we parallelly create a ‘Make In India', a product nation for electronic products as well as semiconductors. Our next phase of growth will be in making India a semiconductor product nation. But that is phase 2 of what we should be doing.