In 2020, when the coronavirus caused a worldwide microchip shortage, you may have heard words like “semiconductor” and “transistor” for the first time, as tech companies scrambled to explain what was causing their orders to delay and stocks to fall. Even as most industries recovered, our reliance on advanced semiconductors will only increase as high compute products like AI become more popular.

But what’s actually been causing these shortages?

Why is this lucrative industry so hard to disrupt?

And what do policies like the CHIPS and Science Act hope to accomplish?

The Building Blocks of Technology

Have you ever wondered why some computers download videos faster than others, or what makes the new iPhone feel like such an upgrade? The secret lies in the microchips and transistors driving nearly every piece of technology we have.

Transistors are like tiny electronic switches in your computer – they have an off mode and an on mode. While we use languages similar to English to write software, computers operate in a language of 0's and 1's called binary. Transistors are a very convenient carrier for that language, with 0’s mapped to off and 1’s mapped to on.

So, the more transistors you have in your chip, the better you can speak the language, and the more things you can get done.

Moore’s Law

The number of transistors in a microchip has doubled every two years since its invention, a trend called Moore’s law. And after more than 50 years of development, we’re up to 19 billion in the new iPhone A17 Pro chip.

This has enabled advancements like smaller computers, faster processing, and the development of high compute systems like ChatGPT.

As the number of transistors per chip has exponentially increased, we’ve had to figure out new and extraordinary fabrication techniques.

Cutting-edge tech needs cutting-edge manufacturing

Extreme Ultraviolet (EUV) Lithography

The setup is incredibly complex – specialized mirrors and light sources, vacuum conditions, and facilities where the air is thousands of times cleaner than a hospital operating room. But there’s no other way to create the next generation of chips.


During EUV lithography, a drop of molten tin is blasted with a specialized laser. When the laser hits the tin droplet, it turns into plasma—a highly energized state of matter—that emits light in the extreme ultraviolet spectrum. This light is necessary to etch the extremely fine patterns that make up the transistor topology.


The laser must hit the tin drop with greater accuracy than was required to send Apollo missions to the moon. And it needs to do this 50,000 times per second. From the release of tin droplets to the firing of the laser to the generation of EUV light, the whole process is controlled by extremely precise timing systems that synchronize each moving part.


The price for all this manufacturing is high. While the number of transistors in a chip may double every two years, the cost of a manufacturing facility to build these chips has doubled every four years – this we call Moore’s Second Law (also called Rock’s Law). These costs have become a barrier to entry for all but the most deep-pocketed companies, further consolidating power in the hands of a few key players.

The Economics of Moore’s Laws

Now Moore’s two laws – governing principles behind the semiconductor industry – have come together to disrupt traditional business dynamics.

Moore's First Law is about operating expenses at scale
Moore’s Second Law is about the upfront capital needed to enable production.

As the cost of production goes up and price per chip goes down, it becomes harder and harder to compete, and players are simply forced to drop out. When Moore’s two laws are taken together, we see an industry built for winner-takes-all. And at the front of this last-man-standing competition is the Taiwan Semiconductor Manufacturing Company known as TSMC.

TSMC’s market share has risen to above 60%, with over 90% market share of advanced chip manufacturing, according to Nasdaq. As they’re share has grown, competition has dwindled. Peers like UMC and GlobalFoundries have scaled back their efforts, with GlobalFoundries in 2018 citing the unsustainable capital expenditure required to keep pace with TSMC's advancements in 7nm and beyond. Similarly, Intel, a historically dominant player in semiconductor manufacturing, has faced significant delays and challenges in transitioning to smaller process nodes, which led to its reliance on TSMC for certain chip production.

“We achieved technology leadership…I don’t think we’ll lose it.”
Morris Chang, the founder and former CEO of TSMC, as reported by the New York Times in 2023.


Because the scale of production is so extreme and requires such specialized knowledge and equipment, only a few firms like Samsung can afford to keep reinvesting in the newest tech. For example, in the shift from 10nm chips to 5nm chips, the cost to TSMC of building a new fabrication plant rose from $1.7 billion to $5.4 billion. This ever-widening gap between TSMC and the rest of semiconductor manufacturers has firmly established its dominance in the industry.

But the world cannot afford the consequences of a semiconductor industry dominated by just one company. We saw what happened during the COVID-19 pandemic, when access to these chips was restricted and annual worldwide production losses were estimated as $110 billion. With so many industries relying on these chips, managing this chokepoint a priority that goes even beyond standard pricing concerns.


So, in a winner-takes-all game, how do you play?

That’s the question the United States Congress tried to answer when they passed the CHIPS and Science Act in 2022. The CHIPS Act set aside $52 billion to revitalize domestic semiconductor manufacturing, research, and development. The goal? Reduce America's dependence on foreign production, especially in light of the geopolitical tensions surrounding Taiwan and the looming threat of further monopolization.



The CHIPS Act aims for geographical diversification by funding new fabrication facilities in locations that reduce risk, even funding TSMC to build a new plant in Arizona. By bringing production closer to home, the U.S. hopes to mitigate the chokehold that one company or one country can have on the global chip supply.

While TSMC and other incumbents do benefit from this legislation, the CHIPS Act isn't just handing cash to giants; it’s also supporting a broader range of players—from upstarts to legacy companies—that can contribute to innovation. This funding opens the door for new entrants to step up and for older companies to reinvest in their capacity, potentially leveling the playing field.



What’s happening in the semiconductor industry may seem distant from daily life, but the call you just placed was on a 5G network, your lane assist kicked in near that curb on the drive to work, and you got an extra five minutes back when ChatGPT wrote that email you’ve been avoiding.

Advanced chips power the tech we rely on, and if speed and innovation matter to you, so does this industry.