The modern computer likely represents the most technologically sophisticated electronics product available to individual consumers. Reflecting this technological sophistication, the supply chains of modern computers are also some of the most complex in the world. Therefore, it probably should not be a surprise that we can find in the electronics industry a microcosm of the global economy, with its tangle of suppliers and retailers that make the entire system possible.
Even with this knowledge, I was surprised by the extent to which the supply chains for computer components spread across different sectors. I was also unprepared for just how specialized some manufacturing processes are. When I started researching which components I should be buying for my new DIY computer, I had expected a fairly simple connection between the brands we see on the shelves, and their suppliers located in East Asia. However, as is the case with real life, things are rarely so simple.
What I found instead is a complex supply chain with contractors and subcontractors each responsible for only a very small part of the entire process. Sometimes, entire factories in the production process have only one purpose: to process tens or even hundreds of thousands of units of a component per day. In contrast to this increasingly divided labor process is the consolidation of different manufacturing companies in the industry. Thanks to the economics of scale and global trade, only a few companies are now major players in the system.
But as COVID-19 swept the world in 2020, this integrated system of specialized producers broke down. This has devastated prices for consumers. As the disruptive effects of the COVID-19 pandemic recede, the future of the electronics industry is at a crossroads. COVID-19 has shown the vulnerabilities of relying on a globally integrated supply chain. From a geopolitical standpoint, nationalism, and protectionism, have also reasserted themselves in the 21st century.
The global technology industry will likely be increasingly divided. Countries are rushing to subsidize semiconductor factories within their borders to reduce their dependence on foreign suppliers. It is difficult to say whether such developments will be good for consumers. I hope that in the long run, such divisions will foster innovation both in manufacturing and computer efficiency. However, in the short term, the fragmentation of the global technology supply chain will be painful for consumers who are used to cheap electronics.
Global Integration
Geographic Labor Specialization
It probably should not come as a surprise to many people that the electronics industry is dependent on specialized labor spread across the globe to produce and deliver its products to buyers. It is somewhat common knowledge that there is a global division of labor when it comes to electronics manufacturing. For instance, the high-value-added stages of the production process often take place in the developed economies of North America and Europe. Many of the well-known consumer chip-makers Intel, AMD, and NVIDIA all have their R&D laboratories in the advanced economies.
In contrast to the advanced economies, the less “developed” (although that term is now highly debatable) economies of East Asia specialize in manufacturing the components that go into a computer. The term “less developed” primarily refers to the economies of Taiwan, South Korea, and to a certain extent Japan and the People’s Republic of China as well. For example, two of the world’s largest silicon fabs, TSMC and Samsung, are located in this region. These semiconductor fabs are commissioned by the likes of Intel and Nvidia to produce the processors in accordance with their designs. East Asian brands in the personal computer industry, such as Gigabyte and MSI, design and manufacture circuit boards that host the computing chips. Meanwhile, the manufacturing of the less technologically sophisticated but more labor-intensive electrical components, such as fans and computer enclosures, takes place in the relatively poorer countries of the region with even lower labor costs, such as the PRC and Vietnam.
The Global Supply Chain
While my understanding of the broad geographic division of labor was correct, I was still surprised by just how specialized many of the manufacturing processes for computer components are. Until I started my research, I was still under the assumption that a factory is responsible for the entire end-to-end production process. For example, I believed that a factory that produces heat sinks can turn a block of aluminum into a functional heat spreader all within its factory walls. But in this regard, I was dead wrong.
In fact, the manufacturing process for almost every component involves extremely specialized factories working on very specific tasks. Different computer components require different painting procedures. For example, the process of painting a simple heat sink aluminum sheet requires an entire factory onto itself. In this case, a relatively cheap process called electrophoretic deposition is used. Other components may require different processes and their own separate painting factories. With modern production methods, entire factories are dedicated to just one task and one task only.
Such super-specialized supply chains developed in response to the explosion in the demand for technology products. As Adam Smith’s well-known theory on specialization explains, the specialization of labor allows for factories to produce more total output than if each worker is responsible for more steps in the production process. But such specialized factories for such specific purposes must be very well integrated into the global supply chain to be economically viable. Factories must have a reliable supply of unfinished parts coming in from their suppliers and a steady output of semifinished components moving forward to businesses in the next stage of the process. Without this well-lubricated supply chain, the current division of labor would make little sense economically or practically.
Original Equipment Manufacturers
On the opposite end of the division of labor, is the surprising consolidation of electronics manufacturing into only a few companies at each stage of production. It was a revelation to me that many competing brands source their products from the same factories. For many of the computer components, only a handful of Original Equipment Manufacturers (OEMs) actually make them. The rest of the companies active in an industry are mostly about branding rather than having their own factories and manufacturing processes. Just to give some examples. Many of the computer power supply unit brands that consumers see in stores, such as Corsair, actually source their products from OEMs such as Liton or Great Wall. Similarly, many of the AIO brands on shelves actually all source their products from Asetek, because Asetek has a patent on putting a pump on the liquid block.
The end consumer rarely sees the brands that actually manufacture their components. Instead, the brands that consumers purchase from in the retail stores are the end result of subcontracting chains. At most, many of the brand names in the computer component industry design their products but do not make them. This revelation did make me wonder, how significant brands actually are when it comes to competition. If competing brands are sourcing their products from the same manufacturers, does a brand’s reputation actually matter? Can people say that multiple brands are competing against each other when they are all really buying from the same suppliers?
As previously mentioned, the economic reason for this concentration of manufacturing into a few companies is to allow manufacturers to take advantage of the economics of scale. Being able to take advantage of scale is especially important for products that have low-profit margins for each unit. As a result, manufacturers must compensate by selling more units. However, this phenomenon of concentration is also prone to charges of monopoly and collusion. As long as the market continues to offer cheap goods for consumers, these charges of monopoly usually fail to gain traction. However, 2021 was far from a normal year. Consumer frustration with consolidations within the electronics industry has led to more than a few worries about companies leveraging their near-monopolistic power to the detriment of the consumers.
Monopoly
If the consumer electronics market has one defining feature in 2021, it was the shortage of almost everything. The most standout component that was was constantly out of stock was the computer chip. However, the great chip shortage of 2021 did reflect an interesting fact about the global economy. This is how much of the world’s economy is dependant on semiconductors. And semiconductor manufacturing is handled by only a handful of companies. For instance, Intel and AMD design most of the world’s computer and server CPUs. Meanwhile, TSMC and Samsung account for 71% of the global semiconductor output. As result, one or two companies often dominate many of the links in the computer supply chain. While smaller producers may exist, they are almost insignificant in terms of market share.
People may point out that there are other computer chip manufacturers out there. Qualcomm, VIA, and other manufacturers of the ARM family of computer chips come to mind. However, these companies do not design or manufacture chips for computers or servers. As interesting as it would be if we are heading towards a future where ARM-powered computers are commonplace; currently AMD and Intel still dominate the computer market.
However, the emergence of the ARM model on licensing may fundamentally change the computer CPU landscape, and break the Intel-AMD duopoly. Rather than designing and manufacturing their own chips, the ARM corporation licenses out their CPU designs to other manufacturers who can modify them for specific use cases. This is the model that had allowed lower-level framework software, such as Windows or Android, to set the standard in the Operating System market yet still maintain competition in the application space. ARM chips have come to dominate the smartphone industry, driving its competitor Intel out of that growing market entirely. One wonders if such a model is possible in the computer and server sectors.
But even if ARM-based chip design companies such as Qualcomm are willing to manufacture computer chips for the retail market, they still depend on chip plants located in East Asia. The big names here are TSMC and Samsung, with other semiconductor companies falling significantly behind. This bottleneck is not likely to be overcome soon, as manufacturing plants are large sunk costs and it takes years to get economically viable production lines set up. But as 2021 draws to a close, the global geopolitical situation just might overturn this economic logic. As COVID-19 continue to severely disrupting the global supply chain and the rise of protectionism all over the world, the globe-spanning, oligopoly-like global semiconductor supply chain may be coming to an end.
The End of the Global Supply Chain?
Countries are moving to secure chips for their own national industries by investing in their own chip plants. The COVID-19 pandemic and the disruption to the global supply chain it caused sped up a process that was already in motion for many due to the China-U.S. geopolitical rivalry. The largest economic blocs of the world, which are the EU, US, and PRC, are all investing in chip manufacturing capabilities to lessen their dependence on the global supply chain.
Like all new ventures, however, the products coming out of these plants will be initially rough around the edges. It takes a long time to build new manufacturing plants. Furthermore, the products produced in these new semiconductor plants will be using older technologies with worse performance than what is achievable in the Taiwanese or South Korean plants. For instance, the new Borsche plant planned to open in Dresden will manufacture 65nm computer chips for the automotive industry. In contrast, the latest computer processors from TSMC are down to 7nm nodes. However, if the will to break away from the established supply chains holds, the end of the globalized semiconductor supply chain is a very real possibility.
For consumers, whether this is development is good or bad price-wise is hard to say. The fragmentation of the global semiconductor supply chain may lead to the development of more diversified manufacturing bases in the future. This emergence of independent semiconductor manufacturing centers, in turn, might lead to more competition and innovation between companies that will be located in different jurisdictions. But the fragmentation of the global supply chain can also lead to the loss of economics of scale, the development of incompatible standards, and the breakdown of exchange in technical know-how. Furthermore, if protectionism really does come back in full force in the future, then technology enthusiasts and early adopters will be disappointed for a long time. This is because for protectionism to work, governments will shut down any foreign competition by lowering imports and supporting local substitutes wherever it can. Therefore, technically “better” products may no longer be obtainable for many people living in jurisdictions where their governments are adamant about building an independent national or regional semiconductor supply chain.
Ultimately, it is almost certain that consumers must face higher prices in the short term. Many of the plans for new factories will not be realized until many years into the future. Meanwhile, new uses for computers are developed and implemented almost every day. From the Internet of Things to cryptocurrency mining to mass migration to the cloud, there seems no end to demands for more computing capacity. And in standard economics, higher demand but fixed supply inevitably leads to higher market prices.
Rising Prices and Anger
One common theme for system builders in 2021 was the seemingly never-ending price increase of computers’ key components. As one can expect, the PC building community has been less than happy about this development. Accusations were rife. Reports and keyboard warriors have blamed scalpers, cryptocurrency miners, retailers, and even the manufacturers themselves. Accusations of “greed”, market manipulation, or that the manufacturers “do not care about their customers” were easy to find.
It was perhaps a symptom of the echo chamber effect that social media has, that camps on all sides were increasingly doubling down on their arguments rather than trying to reach any sort of understanding. For example, one only needs to visit a few of the social media sites for cryptocurrency mining to see threads encouraging miners to buy larger amounts of computing equipment to mine more coins. On the other side of the argument, complaints regarding cryptocurrency mining on PC hobbyist forums feed into a self-sustaining cycle of hatred by members of that community.
But is it of any surprise that people who make money off of their computers are willing to pay higher prices than gamers and technology enthusiasts? For cryptocurrency miners, GPUs are productive assets rather than just something they can enjoy during their leisure time. Large manufacturers and the tech industry are also willing to pay for more to be at the head of the queue for new computer chips. This is because their business profit depends on continued access to computer chips.
At the end of the day, the only true “obligation” of corporations is to generate returns as high as possible for their executives and shareholders. It may be tempting for some consumers to think their favorite corporations care about their sentiments. However, from the investor and owner perspective, the best corporations are those that can generate the most profits within the boundaries set out by the law. To sideline a group of potential buyers who buy in bulk, as many in the gaming community want to happen to the cryptocurrency miners, is detrimental to the manufacturer’s bottom line. Therefore, such actions by corporations are unlikely to happen. As many have pointed out, it is not in the interests of the manufacturers to see product prices fall soon.
Conclusion: the End of Cheap Electronics?
So what does this all mean for consumer electronics going forward? In the short term, it is almost certain that prices for electronics will remain high and will keep rising, at least for the average retail consumer who buys computers for personal use.
The disruption to the global supply chain brought about by COVID-19 and the global geopolitical situation will likely result in a more fragmented global supply chain. A more fragmented global supply chain means less economics of scale and division of labor. This in turn likely means a higher cost of production for the same electronics.
Additionally, demand for computers has exploded over the past decade. Computer chips are now embedded into every conceivable appliance. People are also demanding ever-faster processing speeds for both their personal computers and from cloud services. It is a perfect storm of rising demand but inelastic supply. Most manufacturers simply cannot adjust their production capacity fast enough to meet the challenges facing the global economy.
The prices for electronics will probably decline only in the very long term. The fragmentation of the global supply chain may bring in some much-needed competition to the oligopolies and monopolies that currently dominate the electronics supply chain. As countries and trading blocks attempt to create secure and reliable access to computing equipment for themselves, the current stranglehold that the few large companies have over the entire industry might finally be broken. Over the long term, this fragmentation may encourage more innovation and competition, which will hopefully offset the price increases brought about by the loss of economics of scale.
As a consumer, I hope the increased competition will bring about more innovation that will significantly reduce both the cost as well as the ecological footprint that semiconductor production has. However, as the world scrambles to adapt to the increasing importance of computers in a global economy wrecked by COVID-19 and geopolitical rivalries, I am pessimistic that the situation is unlikely to get any better soon.