Semiconductor manufacturing – are we staring at a glut down the road?

The union government tried to attract chip manufacturing in the country in the 1980s, in the 1990s, and once in 2005-06

By Prosenjit Datta

Will India’s ambitious plans for manufacturing silicon chips succeed this time when earlier attempts have failed? The government has received proposals worth $20.5 billion from five companies, including a Vedanta-Foxconn joint venture, for semiconductor and display fabrication units. This was after the government announced a $10 billion program for the development of a semiconductor and display ecosystem in India.

These are encouraging signs but it might be too early to start celebrating just yet. So far, none of the big chip fabrication companies have signed up for setting up a plant in the country even though most of them have announced their capital expenditure roadmap for the next few years. The companies that have given proposals have no or negligible experience in chip manufacturing. Even Foxconn has no real experience in the complex technology required for chip fabrication. Its expertise lies elsewhere.

Then and Now

There are a couple of things going on for India this time, which were absent in the past but whether these will be enough remains to be seen. The union government tried to attract chip manufacturing in the country in the 1980s, in the 1990s, and once in 2005-06. Each time, companies had shown interest in setting up and, in at least two cases, established chip companies had done detailed project plans but they all came to naught.

The chip industry is broadly divided into those who design chips, those who fabricate chips, and those who do both. Chip fabrication is an immensely complex, capital intensive, and high technology business. While there are a number of countries that have chip foundries or chip fabs, only three countries have the technology and expertise to fabricate the latest generation, cutting edge chips using the 7 nm and lower chip process.

TSMC in Taiwan and Samsung in South Korea play at this end of the market and Intel in the US is now getting into it in a serious way. Intel had largely abandoned by chip fab business some time ago, but under CEO Patrick Gelsinger, and spurred on by the US government’s incentives, it is building capacities again at all ends of the business.

Other chip fabrication companies play largely in the older generation of chip technologies that nevertheless have a big market. Chips are currently used in everything from consumer electronics to automobiles and their requirement is only going up. China, despite having a lot of chip fab capacity, is not at the cutting edge of the market though it has been trying hard for several years to catch up.

If India had managed to push chip manufacturing in a big way in the 1980s when Rajiv Gandhi tried to bring an electronic revolution in the country, it may just have not been far behind in terms of technology compared to the rest of the world. The semiconductor complex of India set up in Mohali was capable of fabricating chips only a generation or so older than the latest at that time.

To succeed, a company will need to build scale. But because of the global chip shortage, every country in the world has started building new capacities.

But it would have been a hard task even if the industry had taken shape then for India to remain in the cutting edge. Several countries in the EU and Japan once boasted of being at the leading edge of chip manufacturing but have all fallen behind. The latest generation of chips requires connecting and packing millions of transistors that cannot even be seen through naked eyes in a tiny piece of the wafer, in the correct order. It is an engineering feat that requires the highest level of expertise – something that even countries with a long history of chip manufacturing lack.

Luckily for India, it need not focus on building the latest generation of chip fabrication foundries. It might be better off focusing on older generation chips which are used in automobiles and consumer electronics. This also fits in with its manufacturing thrust.

The one advantage that India has that it didn’t have in the 1980s is that there are far more buyers of chips in India today than there were then. This is because our auto and consumer electronics manufacturing industries have evolved and also because each of these products requires more chips than ever before.

The second advantage is that there is a global shortage in chips and all countries have realised that some domestic capacity is important to protect their downstream manufacturing interests – and if India plans to become a manufacturing hub in autos or consumer electronics, it needs a secure supply of chips required for these. Finally, big domestic business houses are showing the kind of interest they did not exhibit earlier.

Will there be enough to surmount the hurdles? Let us look at the roadblocks one by one. The first is the requirements of chip manufacturing – it needs cheap and reliable power, absolutely clean water, and a “shake-free” environment. It also generates a lot of pollution. It requires a highly specialised workforce that is not available in India at the moment. It also requires its capacities to be pre-booked if it expects to make money. Enormous capital is required – even to set up older generation fabs – and the payback could take four or five years before it breaks even.

To meet these conditions, any Indian manufacturer will have to build captive power and water purification systems and also be in a place that does not see earthquakes or any sort of disturbance. These would increase its costs over competitors in countries that already have better infrastructure. It also needs to convince buyers to book capacities by taking an enormous leap of faith – unless the group is also a big buyer of chips.

The second big issue is that chip manufacturing has evolved into ecosystems with deep, long-term relationships. The ecosystems have developed over several decades and breaking into them is more difficult than most imagine.

Finally, there is the issue of capacities. To succeed, a company will need to build scale. But because of the chip shortage, every country in the world has started building new capacities. Most of these capacities will come into operation in two to three years. Given the stage India is in, its capacities – if they come up – will probably come when new plants in other countries have already started producing. And given the number of plants being built, we might end up in a situation where we are faced with a market glut by the time our capacities come on stream.

The proposals the government has received so far sound promising – but it is not time to uncork the bubbly just yet.

Datta is a Former Editor, of Business Today and BusinessWorld Magazines

feedbackvnd@cybermedia.co.in