Why Govt Must Get TSMC, Intel, Or Samsung To ‘Make In India’

The coronavirus pandemic has had an unintended consequence. While people struggle for oxygen, medicines, and other medical equipment from the United States to Australia, the tech giants are struggling with a steady supply of semiconductors.

The demand-supply equation for the world’s most important technology commodity has gone for a toss, exposing an import dependency of numerous countries, including China.

What makes semiconductors so critical is their indispensability to most devices we use every day, from Sony PlayStation to cars, and from computers to mobile phones. As the world moved to a work-from-home regime, the demands for various such devices went up, creating a bottleneck of semiconductor supply.

The bottleneck in supply led to a longer waiting time, which in turn impacted the production times for devices, resulting in an overall delay in manufacturing.

Earlier this year, in February, the duration for the order fulfilment stretched to 15 weeks from 13 in December 2020. In January 2020, it used to take 12 weeks to complete an order for a semiconductor. By March, it had touched 16 weeks.

Many companies faced the brunt of this supply bottleneck. In February, General Motors had to shut down three plants in North America for a week due to a shortage of microchips.

Sony, the makers of Playstation, already stated that the console would be in short supply across 2022. Ford Motors has already stated that it's estimating a decline in its earnings by $2.5 billion in 2021, and BMW is struggling to get its two plants going in Europe because of the chip shortage.

While Volkswagen predicts the supply to stabilise by the end of the year, reports have been coming in of their production being impacted in the midst of the pandemic.

It’s not about the complete product alone, for the production of the parts and components is now also under stress. For instance, Nissan was forced to leave out the navigation systems from its cars because of the chip shortages.

Ram Trucks, an American brand, left out a routine ‘intelligent’ rearview mirror from its trucks that is helpful with blind spots. Renault has been struggling with digital screen offerings for its SUV.

Honda Motors has been impacted by the chip shortage as well.

For the auto industry alone, the chip shortage could result in a $61 billion hole this year.

Unlike 20 years ago, cars today rely more on electronics for various automated systems and functions. Today, 40 per cent of the total car cost is made up of automotive electronics, from merely 18 per cent in 2000. By 2030, this percentage will touch 45 per cent.

So, why don’t companies simply fund the creation of more semiconductor manufacturing units?

Firstly, it is a very specialised process, perhaps more specialised than making a vaccine for Covid-19.

The plants are running 24 hours a day, across the year, and the returns on the investment are not immediate. While companies producing chips for automakers can get a plant going for as little as $4 billion, it usually takes anywhere between $15 billion to $20 billion to build a factory that can produce 50,000 wafers, a thin slice of the semiconductor.

It is not only about the cost, for the chip production requires talent on the human resources front, huge infrastructure, dust-free rooms, equipment that costs billions of dollars, and other rare minerals.

In 2020, the market for equipment that goes into the making of semiconductors crossed $60 billion, doubling in the last 10 years.

To understand the elaborate and intricate process, some imagination is warranted.

A chip can be imagined as a small rectangular component, merely 2.342cm in length and 2.6829cm in width. In this area, the objective for chip companies is to pack as many transistors as possible.

It is made by transforming wafers or a thin slice of silicon, which is extracted from plain sand, into a network of billions of transistors.

Fifty years ago, when Intel released its first microprocessor, the chip contained only 2,300 transistors with a node size of 10 microns or 10 millionths of a metre.

However, after four decades of leadership, Intel was displaced by Taiwan Semiconductor Manufacturing Co. (TSMC) and Samsung. Today, we have transistors with a dimension as little as 5 nanometres or 5 billionths of a metre.

To put things in perspective, an average human hair is 20,000 times the size of a transistor being used by TSMC or Samsung, and a transistor is many times smaller than a virus.

The increase in the number of transistors within the semiconductor reflects the intricacy of the entire process and consequently more efficient and powerful devices, thus adding to the computing prowess.

Fifty years ago, in 1971, when Intel came out with its first chip, it had 2,300 transistors. Intel 8086, which came out in 1979, had 29,000 transistors. Pentium in 1993 had 3.1 million transistors. Pentium 3 in 1999 had more than 9 million transistors.

By 2000, TSMC was manufacturing chips with 20 million transistors. By 2006, TSMC’s chips had 177 million transistors. By 2010, TMC was making chips like the GeForce GTX 580 which had 3 billion transistors and were being designed by NVIDIA.

Today, the GeForce RTX 3090, designed by NVIDIA and manufactured by Samsung has 28.1 billion transistors. Apple’s new M1 chip, being made by the TSMC, and released in 2020, has 16 billion transistors.

To work with components of this complexity, the infrastructure includes giant rooms without a speck of dust. The entire process requires constant filtration of the air. The water requirements are quite high too.

For instance, in 2019, one of TSMC’s plants in Taiwan exhausted 156,000 tonnes of water every day. To sum it up, the entire process works at a very atomic level.

Perhaps, for that very reason, not many companies in the world have been able to invest in semiconductor facilities, even after the backing of the state in the form of investment, tax breaks, and infrastructure support.

At present, there are the big three — Intel, Samsung, and TSMC that make up for revenue of close to $200 billion from total revenue of around $450 billion worldwide for the semiconductor market.

The market consolidation is such that the next 12 manufacturers, some including Qualcomm, Broadcom, Micron, NVIDIA, Texas Instruments, AMD, and Sony, make as much as the first three.

A lot of semiconductor orders for automakers were delayed or diverted to industries because other industries like wireless communications and computing devices make for a majority of the global semiconductor market.

A market that was a mere $300 billion in 2010 is now breaching the $475 billion mark in 2020. Consumer electronics, industrial, aerospace, and military needs make up for a significant chunk of the chip demand.

The consolidation is skewed within the industry itself. For instance, TSMC is a supplier of chips for Qualcomm that further supplies majorly to Apple, Samsung, and Xiaomi.

It also supplies to Broadcom which supplies majorly to Apple, Samsung, Amazon, and Dell. NVIDIA and AMD are also critical clients for TSMC and they further supply most computing industries.

To put it mildly, it’s a very interconnected, yet small industry, and most of it is based in a few countries, and this is where the problem begins.

In 2020, the demand for these communication and computing devices went up as more people began working for home.

What complicated matters were the drastic cuts implemented by auto and electronics manufacturers in the early days of the pandemic, anticipating lower demand, and then a sudden spike in demand within months, leaving chip manufacturers with more demand for the supply.

Suddenly, the United States and China realised how dependent they had become on Taiwan. The island nation with a mere population of 23 million people is regarded as a province by China.

In the past, Beijing has convinced many nations to not establish any diplomatic ties with Taiwan.

Being the central point in a conflict between the United States and China, Taiwan is not only a ticking diplomatic timebomb but also critical enough to take down the entire global supply chain.

The United States was quick to address this problem. During the days of the Trump administration, sanctions were ordered to cut off China’s access to any US chip technology including design. This disrupted the supply of chips to Huawei even, one of the biggest 5G companies in the world.

The US government has also negotiated the setting up of a $12 billion chip fabrication plant with TSMC in Arizona. The production in this plant is expected to start from 2024, with a focus on 5-nanometer fabs, and 3-nanometer fabs in the future to cater to Apple’s and Intel’s growing requirements. With Samsung, it has negotiated for a $10 billion facility in Texas.

In the last week of February 2021, President Joe Biden signed an executive order directing the administration to address the worrying shortfall in semiconductor production that impacted operations at some auto plants.

Even China is playing the catchup game. Earlier this year, in March, China also voiced its ambitions to cut its dependence on chip imports, annually worth $300 billion.

Beijing has also laid out plans to develop infrastructure which will include national laboratories and innovation centres under their Sci-Tech Innovation 2030 Agenda.

There is also a push being made to usher the ‘technology immigration system’ to get talent from Taiwan and the United States. In its five-year plan presented in October last year, China has committed $1.4 trillion to develop the chip industry and other critical technologies and industries.

India, if not on the same scale, is also attempting to get the likes of TSMC to come and build in India. While offering $1 billion in cash to each semiconductor company, the government aims to strengthen the supply chain for the companies in India manufacturing computing devices, industrial devices, and smartphone segments.

However, the government needs to do more.

While it is not possible for the Modi regime to invest on the scale at which China does, it can offer more incentives in the form of infrastructure support across the first five or 10 years, a cash handout of $3 billion to $5 billion across the first five or 10 years, which may constitute 20-25 per cent of the initial investment that any incoming company makes.

The government must realise the potential of this industry as well. To begin with, it must start with reaching out to TSMC and Samsung using the government back channels. While the Production Linked Incentive Schemes, with all their right intent, can appeal to other industries, the semiconductor sector needs a bit more.

India is still light years away when it comes to an indigenous manufacturing capacity for semiconductors, but it all starts with getting someone to ‘Make in India’.