A global shortage of semiconductor chips is disrupting automobile manufacturing. From the US to India, automakers have had to reduce production and shut down plants. The losses of revenues and profits add up to millions of dollars. It is not just the auto industry that is worrying about chip scarcity. Chips are in everything today—from consumer durables like televisions and washing machines to cameras and WiFi routers, from cars to mobile phones, power grids and telecom towers.
The pandemic had increased the demand for devices, and mobile phone, tablet and laptop manufacturers have snapped up all available chips. Auto firms had initially reduced chip orders during the pandemic and manufacturers had diverted supplies to other buyers.
Chip production is concentrated in the Asia Pacific region—Taiwan, South Korea, Japan, China and Singapore. Chip fabrication plants are running at full capacities but many analysts predict the shortage could go on till 2022 or longer because chip demand in every device is going up.
The chip business is segmented into integrated players such as Intel that design and manufacture their own chips; chip design companies such as Qualcomm; and pure chip fabrication foundries or FABs such as Taiwan Semiconductor Manufacturing Company (TSMC), GlobalFoundries of the US or Semiconductor Manufacturing International Corporation (SMIC) of China. Some companies like Samsung design chips but also have huge FAB capacity for other designers.
Chip nationalism—where countries with chip capacities are trying to ensure that they tie-up supplies and stocks for their requirements first—is going to create additional problems for those that do not have their own production facilities.
In the US, both Republicans and Democrats are worried enough about the chip shortage to propose a law that will allocate billions of dollars to increase development and manufacturing of chips in the nation and reduce outside dependence. Though the US has a lot of production capacity, its demand outstrips its supplies and lately, it has depended heavily on global suppliers in South Korea, China and Taiwan.
In China, the government is ramping up capabilities too. Till recently, China was considered to be lagging behind the US when it came to chip designs. And it was also behind Taiwan, South Korea and the US in terms of sophisticated FABs that churn out the most advanced chips. China has many foundries but Taiwan and South Korea’s newest FABs are far more advanced.
India finds itself in a not-too-happy position in this regard. While it has a decent chip design talent, it never built up chip FAB capacity. The ISRO has a FAB foundry and the DRDO has one too. But they are primarily for the requirements of the respective organisations and also not as sophisticated as the latest in the world.
Since 2006, different governments have tried to get companies to set up FAB units in India without success. Two separate consortia had expressed interest earlier but failed to get their act together or backed out for some reason.
Part of the problem is the investment required and the technology that needs to be mastered. The latest TSMC state-of-the-art FAB unit costs over $13 billion to build. Even smaller and less sophisticated foundries would require several billion dollars of investment. FAB plants also require a lot of land, clean water, stable power and a highly skilled workforce. Above all, FAB companies look at the long term—10 to 15 years—when deciding on geography. They need policy stability, assured demand and long-term incentives.
The MEITY has again sought expression of interest for setting up a FAB plant in India or in acquiring one abroad. Will it yield any better results than earlier?
One thing that has turned in India’s favour is domestic demand. By some estimates, India accounts for 5% of the world’s chip demand and the country’s consumption is growing at 25% per annum. This could be attractive to a global FAB manufacturer.
But that alone might not be enough. The government needs to look at chip fabrication as a strategic industry where it needs to get involved and invest in, just as China is doing.
While the government plans to reduce its business footprint by divesting off most non-strategic Public Sector Enterprises (PSEs), it needs to remember that the original reason for setting up government-owned units in any country has been the need to build capacity and expertise in areas where the private sector was either unwilling or incapable.
The DRDO and ISRO foundries only cater to their requirements. The government could look at using one of the other PSEs—say, Bharat Electronics Ltd or Hindustan Aeronautics Ltd—to set up a foundry with the help of a global major. Allowing management a free hand in a joint venture where the global major has brought in technical expertise, as was done when Maruti was originally set up, and providing long-term policy stability along with proper incentives could well lead to success where earlier attempts failed.
India needs to understand what the US, China, South Korea and Japan have long known—that being entirely dependent on global supply chains for something as critical as chips, which are at the heart of all modern machines, is not a wise policy.
Prosenjit Datta
Senior business journalist
(prosaicview@gmail.com)