Lecture Practice 4

The Microchip Crisis

TRANSCRIPT: Hello and welcome to this short lecture on the global microchip crisis. I’ll start with a bit of background on the microchip industry, then I’ll talk about how chips are manufactured, after that I’ll discuss the economics of microchips and finish with a summary on what needs to change.

Ok, so, let’s begin with some background. Microchips, which can be found in many of the devices we need for work, study, travel, and entertainment, are one of the most manufactured items in the world. In 2020, according to ASML, more than 932 billion chips were manufactured in a sector which continuously generates new products, transforms industries and is currently worth $484bn. However, as the production of semiconductors has steadily declined in the west, and now China, Japan, South Korea and Taiwan are responsible for 60% of global semiconductor sales, we are currently experiencing a huge shortage of these chips needed for almost every electronic device to work, for the most part due to our over-reliance on the manufacturing in Asia and our ever-increasing demand for the most up-to-date technology.

Ok, now let’s look at how microchips are manufactured. A modern microchip, also known as a semi-conductor memory chip or integrated circuit, is typically made from silicon, which comes from sand and is one of the Earth’s most ubiquitous minerals. Before the first silicon transistor was created in 1947, the mechanics of computing had been performed by slow and cumbersome vacuum tubes, but since the introduction of silicon, the electronic switches that run computations inside computers known as transistors can be made small enough to fit a billion of them on one chip. Another reason for using silicon is that it is a semiconductor, unlike many other metals used in electrical currents, meaning that when it is mixed with those other metals, it can switch an electrical current on or off. Making microchips, however, is extremely challenging. King et al. (2021) report that not only do the semiconductor processing plants take years to build, but once they are built, the rooms where the chips are produced must not only be as sterile as an operating theatre as one tiny dust particle or hair can impair the intricate systems, but must also manage a number of variables such as temperature, pressure, or magnetic fields, all of which could impact the manufacturing process. Although the majority of the operations involved in making chips, such as converting the silicon powder into discs known as silicon wafers, or printing the circuits onto each wafer are carried out by robots, the process to produce just one chip is rather lengthy and can involve anywhere between 1,000 and 2,000 steps over approximately three months. To some extent, this could perhaps explain why demand is currently outstripping supply.

So, now let’s move onto section three: the economics of the microchip industry. As the demand for microchips grows, so do the costs, and for an industry whose worth continues to proliferate year on year, the economics can be burdensome. On the one hand, with regard to the fabrication plants required to be in constant operation, it costs roughly $15bn to produce 50,000 wafers, or $20bn in the case of the more advanced facilities such as those found at Intel, Samsung and TSMC (the Taiwan Semiconductor Manufacturing Company) and that is only as a result of serious investment and perseverance of around three years to begin construction. On the other hand, according to Kaur (2021), the chips themselves become obsolete quickly and for companies to remain lucrative, only a yield above 90% and a profit of at least $3bn will suffice, which is extremely difficult in a market which is dominated by the three aforementioned companies with a joint revenue of $188bn in 2020, a total which is greater than the next twelve largest chipmakers combined. Therefore, it could be said we are presently in a catch-22 situation – as the current lack of microchips is slowing down industries and delaying products, there is pressure to meet these demands as quickly as possible. However, this would only encourage the largest companies to remain unchallenged, meaning that more and more chipmakers in other parts of the world would ultimately cease to exist.

Ok, let’s summarise and I’ll offer some fundamental solutions. Although the complex and time-consuming process of making microchips continues to struggle to keep up with the increasing global demand for them, it is clear that relying on a small number of manufacturers to produce the majority of the world’s chips is economically unviable. Consequently, governments across the globe must attempt to take more risks in order to become more self-sufficient. This, of course, takes time, money and patience, which suggests that the temporary shortage of microchips from 2020 may in turn, be a more permanent issue for many more years to come.Thank you.

Written by C. Wilson (AEUK) / Spoken by Sarah Jackson