Just as an airplane needs an airport to take off, the semiconductor industry needs a market to thrive.

Fortunately, our market is large enough.

Unfortunately, our talent and technology aren’t quite ready to take the stage.

On August 9, the U.S. CHIPS and Science Act was enacted, an event that has arguably shaped the trajectory of the entire semiconductor industry.

Under the bill, the U.S. will provide approximately $52.7 billion in direct subsidies to semiconductor manufacturing and R&D companies, and offer a 25% tax credit to companies investing in the construction of semiconductor factories in the U.S.

The bill stipulates that to qualify for these massive subsidies, companies must not build factories in China for 10 years.

This is clearly an attempt to block China’s high-end industries.

Discussions on “how to counteract” this have sparked a wide range of opinions.

Given recent well-known events in the South China Sea, one proposal has gained significant traction:

"Seize TSMC."

It seems that simply by occupying TSMC’s headquarters, one could gain control of the most cutting-edge chip manufacturing technology and produce the most advanced chips.

This idea is bold and highly effective at stirring patriotic fervor; while the intentions are good, it is not realistic.

Starting from scratch, high hopes followed by disappointment

In the semiconductor sector, we lack both talent and technology.

Talent drives R&D, technology leads to equipment, and equipment, through application, “nurtures” talent—this is a closed loop. The reason we lack advanced technology is that the cultivation of talent has been interrupted.

In 1956, the Institute of Applied Physics at the Chinese Academy of Sciences produced the first crystal transistor, nine years behind the United States.

In 1965, we successfully developed China’s first integrated circuit product, seven years behind the United States.

In the same year, the Type 65 photolithography machine was developed, five years behind the United States.

In 1977, the JGK-3 semi-automatic lithography machine was developed.

In 1982, the KHA-75-1 photolithography machine was introduced, trailing Japan’s Canon—the most advanced at the time—by only four years.

We had the technology and the talent. Back then, although we started from scratch, we were only behind the United States and were essentially on par with Japan. TSMC hadn’t even been established yet.

And then, nothing happened.

After 20 years of silence, when TSMC partnered with ASML to develop the now-mainstream immersion lithography machines, our dry lithography project had only just been initiated.

It’s not that we didn’t recognize the importance of semiconductors, nor that we hadn’t cultivated semiconductor talent—it’s just that there was a generational gap.

Things were going well—how did it come to an abrupt halt?

A Reluctant “Halfway Abandonment”

In the early days of reform and opening-up, China was so backward it was heartbreaking.

While Americans were traveling the globe on Boeing 747s and Japanese commuters were riding the Shinkansen, we were still using steam locomotives, and many daily necessities were rationed.

Poverty became the greatest obstacle to scientific research and innovation, and it also served as the leaf that obscured our vision.

TSMC’s success stems partly from favorable external conditions and partly from its own hard work and perseverance.

For reasons well understood by all, TSMC had easier access to advanced technologies from Europe and the United States and was more readily recognized by them. It was only through the transfer of American technology and industry that the company was able to gain a firm foothold.

Consequently, Taiwan continuously increased its investment, and educational policies promptly shifted to prioritize the cultivation of semiconductor talent.

With access to advanced technology, free from elaborate blockades and embargoes, and a strong emphasis on talent development, TSMC’s rise was inevitable.

Our situation is different.

First and foremost, we must acknowledge that chips are not atomic bombs.

An atomic bomb is a knife hanging over our necks. If we cannot produce atomic bombs, we will constantly face nuclear blackmail and will be unable to develop our economy in a relatively safe environment—it is a matter of life and death.

Chips, on the other hand, are like credit card bills: although we will have to face them sooner or later, we can still live our lives normally until the due date arrives.

On one side lies the future of the tech industry; on the other, an imminent economic crisis. On one side is a potential chip crisis that may not materialize for a long time; on the other, 100 million citizens struggling in hardship…

Any rational decision-maker would make the same choice.

We have far too many pressing problems right now. When it comes to economic and livelihood issues, we absolutely cannot approach chip production the same way we approach building an atomic bomb.

“Trade-Industry-Technology” vs. “Technology-Industry-Trade”

“Trade-Industry-Technology” and “Technology-Industry-Trade” represent two distinct development paths.

The former prioritizes “trade-processing-technology,” first accumulating capital through agency sales and processing, then investing in R&D. The latter follows the exact opposite sequence.

Now that we have money, we only want advanced technology.

But if we could go back in time, we would realize that what was most needed back then was actually capital.

In May 1973, Wang Shouwu, then Deputy Director of the Institute of Semiconductors at the Chinese Academy of Sciences, led a 13-member delegation on a study tour to Japan.

This study tour yielded an unexpected result: NEC was willing to transfer a complete 3-inch chip production line, presenting an excellent opportunity to introduce advanced integrated circuit production technology.

The quoted price was: $30 million for one process technology and equipment, $40 million for two, and $50 million for three.

However, at the time, we could only come up with a maximum of $15 million.

Therefore, back then we had no choice but to adopt the “trade-industry-technology” development path: first becoming the world’s factory, then the world’s engine.

Many people believe that the “trade-industry-technology” approach neglects technology, which is why China’s advanced technology has stagnated, but this is not the case.

These two paths lead to the same destination.

Technology is the ultimate goal of the “trade-industry-technology” model and the initial driving force behind the “technology-industry-trade” model; neither path denies the importance of technology.

Take Huawei as an example: this company, whose telecommunications technology is so advanced that it has drawn U.S. sanctions, chose a classic “trade-industry-technology” path.

Huawei initially focused solely on selling switches, operating as a pure trading company. It wasn’t until the early 21st century that it began developing core technologies and entered the mobile phone market at the right time—which is how today’s Huawei came to be.

However, it is true that a large number of “trade-industry-technology” enterprises have strayed from the right path.

While trade is merely a means to accumulate initial capital, many companies have become addicted to quick profits and cannot extricate themselves.

How comfortable it is to make money without lifting a finger—why bother with R&D?

At the time, policies also required research institutes and factories to “operate on a self-supporting basis,” leading to a scenario that now seems utterly surreal:

Nuclear plants producing ice cream, radar factories manufacturing televisions, and the Wuhan Radio Component Factory No. 3—which was originally capable of producing lithography machines—was forced to switch to selling snacks, only to eventually go bankrupt…

▲ The ice cream produced by China National Nuclear Corporation’s Plant 504 defined the summers of generations of Lanzhou residents and shaped the childhood memories of countless people from that era.

When you’re bent over, rushing down the road, you don’t have time to look up at the sky.

The gap between us and developed nations in the chip industry widened from four years to twenty.

The Overwhelming "Sugar-Coated Bullets"

More terrifying than our own deviation from the “trade-industry-technology” path are the “sugar-coated bullets” delivered right to our doorstep from abroad.

There are many examples of this, such as EDA.

As is well known, the circuits inside chips have become so complex that they cannot be drawn by hand; computer-aided design is essential. This is known as Electronic Design Automation, or EDA.

This technology was included in the list of items subject to the COCOM embargo against us back then, which speaks volumes about its importance.

In 1991, the domestically developed Panda System was born, giving us our own EDA. Not only was it powerful, but it also won two international awards.

But before long, the three major U.S. EDA giants came crawling back at lightning speed. Not only did they quickly lift the embargo, but they also aggressively seized market share by slashing prices and offering free access.

"Panda" was truly reduced to an endangered species by these "sugar-coated bullets."

More importantly, the dominance of Western software cost us far more than just a domestic brand. From then on, businesses used Western software, and schools taught Western software.

Consequently, we’ve been forced to compete with the West under rules they set, with our necks constantly at their mercy.

That is why the industry reacted so strongly when the U.S. struck again on August 13, imposing export controls on EDA software.

Moreover, as domestic companies in this sector continue to wither away, opportunities to commercialize research achievements are dwindling—while numerous papers are published, very few technologies actually make it to market.

According to estimates in the *China Integrated Circuit Industry Talent Development Report (2020–2021 Edition)*, by 2023, China will need 760,000 professionals in the integrated circuit sector, yet the current workforce across the entire industry stands at only 540,000, leaving a shortfall of 200,000.

It takes ten years to grow a tree, but a hundred years to cultivate a person.

When it comes to talent development, forcing growth is highly inadvisable. Having missed the optimal window for development, we now have no choice but to calm down and diligently “catch up.”

In the context of talent development, this may actually be a stroke of luck.

Poverty is the original sin

The first wave of the chip industry boom was driven by personal computers.

You probably still remember a time when everyone around you was talking about computer chips—Pentium, Core, dual-core, and the like.

▲ Do you remember that line at the end of the ad: “Lights! Wait for the lights, wait for the lights”?

In reality, the biggest driving force behind chip development is the market. Without the sustained support of consumers, how could R&D costs—often running into the billions of dollars—be recouped?

However, in the 1980s and 1990s, we were still poor and couldn’t afford to spend much; personal computers remained a rarity until the year 2000.

Given our purchasing power at the time, we simply couldn’t support those money-guzzling behemoths that required billions of dollars in annual investment.

Even if we had insisted on independent R&D back then, and even if domestic chip companies had flourished in a competitive landscape, who would have footed the bill for their research and development?

So, returning to the original question—

Would taking over TSMC allow us to acquire mature semiconductor talent?

No. TSMC has a strong global atmosphere, with employees hailing from all over the world. Seizing the factory isn’t difficult, but how would we retain those employees from around the globe?

Would taking over TSMC grant us access to advanced semiconductor technology?

No. Chip production involves hundreds of processes, thousands of pieces of equipment, and countless patents.

TSMC is merely a foundry; its upstream and downstream industries are scattered across the globe—from manufacturers of chip materials to developers and producers of semiconductor equipment, to companies responsible for packaging and testing. Without access to the advanced technologies of these upstream and downstream sectors, seizing TSMC would be of no use.

Given the current situation, we have only two paths to save the domestic chip industry:

First, wait for the U.S. to lift the export ban.

Second, Chinese companies must establish a complete industrial chain, taking control of everything from silicon wafer materials to semiconductor equipment.

The first option depends on the U.S. and would leave us at their mercy.

The latter is like trying to catch up to an airplane on a hand-guided tractor—though as difficult as climbing to the heavens, it embodies the spirit of revolutionary heroism.

Fortunately, we have a sufficiently large market, which is a massive advantage.

Years of experience have proven that the more the West blocks and restricts our sectors, the more we are able to muster the determination to surpass them.

The U.S. government’s aggressive stance is, in fact, a boon for domestic companies committed to independent R&D.

In this environment, they are bound to attract the attention of capital markets and receive more policy benefits, making a stunning comeback.

Even Chinese companies that were previously uninterested in domestic chips and EDA software will begin to procure them out of concerns for technological security.

Looking back on this period decades from now, we may find ourselves involuntarily remarking: We must award the United States a gleaming gold medal.