The current brouhaha over semiconductors is reminiscent of the Y2K bug at the turn of the century, when there was a fear that the bug would affect software and bring computer systems to a crashing halt. This time around, the sudden and unexpected shortage of semiconductors that make up the core of most electronic items threatens to give a massive shock to the global economy as virtually everything from cars to toys to military drones has a chip inside it. Unlike the Y2K bug, it is not merely a technical issue but has acquired overtones and ramifications far beyond that.
By Cherian Samuel
From the geo-political perspective, it has become a part of the technological tussle between the United States and China; whilst the microchip was invented in the US in the 1950s, with Europe and the US manufacturers providing much of the world’s supply, by the 1990s, manufacturing shifted to Asia, driven by reduced cost of production and government incentives.
Thus, though US headquartered companies still have a 44 per cent market share in sales currently, only 11 per cent of chips are manufactured in the US, with the bulk of the manufacturing happening in South Korea (28 per cent), Taiwan (22 per cent), Japan (16 per cent), and China (12 per cent). Europe’s share is a measly 3 percent. Much of the production is gobbled up by Asian companies such as Samsung (8.1 per cent), Huawei (4.1 per cent), Lenovo (4.1 per cent), and Xiaomi (2 per cent), even though Apple by itself consumes as much as 12 per cent.
Huawei saw a 23 per cent drop in purchase share in 2020 following the imposition of restrictions by the US government on its purchase of semi-conductors from its main supplier, the Taiwan Semiconductor Manufacturing Company (TSMC). These sanctions could be applied since much of the technology used in the manufacture of semiconductors is American. The US also placed Semiconductor Manufacturing International Corporation (SMIC), China’s biggest semiconductor producer, on the entity list in September 2020, preventing it from accessing US software and chip making hardware, citing possible end-use by the Chinese military.
Even before the blow-up in US-China relations, the Chinese government has been ramping up its efforts to increase chip production within the country, under the Make-in-China 202 initiative. While huge amounts of money have been invested in the endeavour, the Chinese companies have been hamstrung by the lack of technical know-how. Chinese manufacturing is at least two generations behind the latest chips available today.
For other countries that have tried to fashion a path to semiconductor independence, it is both technical know how as well as financial constraints that have so far presented insurmountable obstacles. The average semiconductor production unit requires an investment of at least $10-12 billion and a long gestation period. Coupled with rapidly changing technology and requirements of the client companies, a factory or foundry set up to produce a certain kind of chip could very well find its product is not in demand after coming on-stream or face price pressures from competitors. This has led to a situation where there are only a handful of suppliers in a $500 billion industry.
The geo-political tussle over micro-chips has been further exacerbated by the sudden shortage of semiconductors, which is partially an outcome of the tussle and partially of the pandemic. Huawei began to stock up on chips anticipating the US sanctions at the same time that many other purchasers cancelled orders on the back of reduced demand for their products. Semi-conductor foundries, in turn, reduced production and diverted their sales to sectors where demand increased, like laptops, webcams and gaming systems.
The shortage is leading to huge losses for companies across sectors, but the most affected has been the car industry. Semiconductors control critical areas of the vehicle like air-conditioning, airbags, driver information, audio/video entertainment, sensors, transmission and gear, navigation and collision detection systems.
The Biden administration has gone all out in its efforts to both restrain China from accessing technology as well as in bringing the US back as a leading manufacturer of semiconductors. The president has initiated a review of the semi-conductor supply chain and urged Congress to provide $37 billion of funding to spur semiconductor manufacturing in the US. In January, the CHIPS for America Act was introduced in the US Congress providing investment incentives and tax credits to manufacturers.
Though India consumes as much as 5 per cent of the global production of semiconductors, it produces only a miniscule amount, mainly in the in-house foundries of the Indian Space Research Organisation (ISRO) and the Defence Research and Development Organisation (DRDO). However, as much as 45 per cent of the research and design of semiconductors is carried out in India, by a talent pool of over 20,000 engineers. Many start-ups are also engaged in various aspects of the semiconductor ecosystem, such as testing and developing niche solutions.
The government, recognising the strategic value of semi-conductors, has made efforts over the years to have fabrication units (fabs) established in India, but those efforts have floundered because of the high levels of investment needed, the low levels of financial support from the government, coupled with very stringent conditions laid down in the proposal documents.
There have also been divided voices on whether setting up a fab would be a viable proposition. In an article in 2015, Vinod Dham, the father of the Pentium chip, posed the question, “Does India really need a $5 billion semiconductor unit?”1 He gave a qualified response, noting that China had been unsuccessful in its endeavour despite spending billions, as well as the fact that Moore’s law dictated that the technology required to produce semiconductors changed every few years, thus making the technology obsolete by the time the factories came up. He instead suggested that global majors like Samsung be encouraged to set up factories in India.
The government seems to be following this playbook by inviting expressions of interest (EoI) towards the end of 20203, and actively following up with manufacturers like Taiwan Semiconductor Manufacturing Company, VIA Technologies Inc., United Microelectronics Corporation, Intel, Micron Technology, Inc., NXP Semiconductors, and Texas Instruments, Fuji Electric Co. Panasonic, and Infineon.
As Dham himself noted recently, the context of 2021 is very different from that of 2015. Electronic manufacturing has begun in a big way, on the back of the Production Linked Incentives (PLI) Scheme, techno-nationalism is on the rise — which makes it imperative for India to establish supply of strategically important products such as semiconductors, and the imminent end of Moore’s law, has provided greater impetus to the current endeavour.
That said, it is still not an easy road given the complicated ecosystem of semi-conductors, with its myriad types, the many different parts of the process — ranging from design to production to testing, and the infrastructure required — from dust-free environments to the availability of water. In 2013, the government had got to the stage where approval was granted to two consortiums led by Jaypee and HSMC to set up fabs with an investment of Rs 63,000 crores but financial closure could not be achieved despite the government offering a 40 per cent subsidy.
With the current semiconductor demand of $40 billion expected to rise to $100 billion by 2025 for just the electronics industry, the time is indeed right to begin production of semiconductors within the country. Though the EoI covers all the bases from setting up facilities in India to acquiring assets outside India, additional alternatives should include expanding the scope of the existing government-owned production facilities in ISRO/DRDO.
This article first appeared in www.idsa.in and it belongs to them. The author is a research associate with IDSA.