For a few months now, computer gamers and car manufacturers have had to limit themselves. Many gamers are currently unable to buy a Playstation 5 because Sony simply doesn't have enough processors to equip the console. BMW or Audi, on the other hand, are painting entire shifts because they cannot get to the chips that control the increasingly complex electronics in their vehicles. These are two symptoms of the current lack of chips. Both have completely different reasons - and yet have a common root: the unimaginable complexity that microchips have achieved in recent years.

Their development has followed Moore's law for over half a century. In the 1970s, Intel founder Gordon Moore predicted that the complexity of integrated circuits would double about every two years. Back then, computer chips had several thousand transistors. These electrical switches are the physical basis of the "ones and zeros" with which every modern computer calculates. Today, chips like the M1 from Apple or the most modern graphics processors combine several billion transistors on the surface of a fingernail. The smallest structures on these commercial semiconductor components are five nanometers wide. The usual comparisons with the thickness of a human hair to illustrate the minuteness of spatial dimensions fail here.Such structures are a hundred times narrower than the wavelength of visible light. At best it can be compared to the DNA double helix. It's two nanometers wide - that's how big the structures on a chip that IBM presented in May are.

Moore's Law wields the whip

Moore's Law is not a law of nature.

It's a promise the semiconductor industry is trying to deliver.

Because smaller structures enable more computing power and less energy consumption.

But industry is paying an ever higher price for this race into the microcosm.

Compared to the 1970s, it takes 18 times more professionals to maintain the constant doubling in complexity, according to Stanford University researchers in the 2020

American Economic Review

have calculated. And today it takes months to manufacture a smartphone chip, for example, because the individual manufacturing steps are tedious. That makes production inflexible. The supply chain behind the computer chips has also become more and more susceptible to disruptions. Because only a few manufacturers have been able to keep up with Moore's Law over the years. Those who have not yet been left behind have to rely on highly specialized suppliers in many ways. Often these are companies that are the only ones to offer a specific machine or software. These neuralgic points are then reflected in the geographical distribution of the supplier companies and give individual regions great power over some parts of the chip industry.