How did Japan lose the chip war over 30 years ago?

Only a little over 30 years have passed, and few remember the chip war that erupted between Japan and the United States.

In this war, the Japanese lost completely, falling from a peak of nearly 80% of the global DRAM market share to zero today. This chip war perfectly illustrates what is meant by international political economics; Adam Smith's theory of free market competition is just a beautiful fairy tale in the context of major power industrial competition.

The first five years of the 1980s were a high point for Japanese semiconductor chip companies.

Silicon Valley tech startups like Intel and AMD were relentlessly pursued, overtaken, and dethroned in the semiconductor memory sector, which (primarily semiconductor memory at the time) became a Japanese corporate back garden.

American tech companies failed in their model.

The Silicon Valley model involved venture capital injecting funds into startups, which, after receiving funding, achieved market success through continuous technological innovation, improved company valuation, then went public, enabling venture capitalists to sell their stock and exit with profits.

This market-oriented model is efficient but small-scale. Resource integration is difficult among companies, as they are all competitors vying for the same resources.

The Japanese approach was drastically different: focusing resources on major undertakings. Kyushu Island was known as Japan's "Silicon Island," once home to the majority of Japan's semiconductor chip companies.

In 1974, the Japanese government approved the "Very Large Scale Integrated Circuit" plan, a semiconductor chip initiative aimed at surpassing American integrated circuit technology.

Subsequently, Japan's Ministry of International Trade and Industry (MITI) organized five companies—Hitachi, NEC, Fujitsu, Mitsubishi, and Toshiba—mandating the integration of Japan's industry-academia-research semiconductor talent, breaking down corporate barriers, fostering collaboration, and improving Japan's semiconductor chip technology.

Japan's plan nearly faltered; companies were suspicious of each other and undermined each other's efforts, and government funding was slow to materialize. At a critical juncture, Yasuo Tsurui, the pioneer of Japanese semiconductor research, stepped forward. He used his prestige to unite the diverse stakeholders.

Tsurui's argument was simple and clear: only by working together could they overcome Japan's backwardness in fundamental chip technology. Once research results were obtained, each company could conduct its own product development. This was the only way to reverse the predicament of Japanese companies fighting alone in international competition.

Four years into the plan, Japan had secured thousands of patents, dramatically closing the technological gap with the United States. Then, the Japanese government introduced loans and tax incentives, bolstering companies like Hitachi, NEC, and Fujitsu.

Modern semiconductor memory chip manufacturing plants sprouted up across Japan. As production lines ran day and night, the Japanese launched a saturation attack.

The Americans' nightmare began. In 1980, Japan captured 30% of the semiconductor memory market; five years later, Japan's share exceeded 50%, leaving the United States behind.

Silicon Valley's high-tech companies could not tolerate the plummeting market share and repeatedly sent people across the Pacific to Japan for reconnaissance, with dispiriting results.

Andy Grove, then Intel's production chief, lamented, "Those returning from visits to Japan painted a grim picture." If Grove had visited Japan, he would have been terrified: one Japanese company dedicated an entire building to chip research; the first floor was dedicated to 16KB capacity development, the second to 64KB, and the third to 256KB.

This Japanese research pace was like a legendary three-pronged attack, leaving Silicon Valley companies accustomed to single-handed combat completely unprepared.

What suffocated Americans was that Japanese memory chips were not only high in volume but also of superior quality. In the 1980s, the Semiconductor Industry Association (SIA) conducted quality tests on US and Japanese memory chips hoping to find a weakness in their competitor's products. They found that the highest quality American memory chip was inferior to the lowest quality Japanese chip.

Furthermore, the Japanese confidently assured customers that their memory chips were guaranteed for 25 years!

Under Japan's aggressive offensive, American chip companies suffered a devastating defeat; their financial data resembled melting ice cream—a complete mess.

In 1981, AMD's net profit fell by two-thirds, and National Semiconductor lost $11 million, compared to a profit of $52 million the previous year.

The following year, Intel was forced to lay off 2,000 employees. The Japanese continued to expand their gains, while the Americans continued to lament. In 1985, Intel surrendered, announcing its withdrawal from the DRAM memory business.

This war cost Intel $173 million, its first-ever loss. If it hadn't been for IBM's intervention at Intel's most critical moment—purchasing 12% of its bonds to secure cash flow—this chip giant might have collapsed or been acquired, potentially rewriting the history of the American information industry.

Intel founder Robert Noyce (left) joined forces with other Silicon Valley companies to form SIA, aiming to counter the competition from Japanese semiconductor companies.

Robert Noyce lamented that America had entered a period of "imperial decline." He predicted that if this situation continued, Silicon Valley would become a wasteland.

Even more intolerable to the Americans was Fujitsu's plan to acquire 80% of Fairchild Semiconductor. Fairchild Semiconductor was a Silicon Valley relic, as most Silicon Valley tech company founders (including Intel and AMD) had worked at Fairchild.

In the hearts of Silicon Valley people, Fairchild Semiconductor was a divine entity. Now the Japanese were going to buy their "god." Wasn't this a shame? One American newspaper wrote, "This deal sent a message that we are already far behind. The important thing is how we respond to this."

Years earlier, Silicon Valley tech companies formed the Semiconductor Industry Association (SIA) to counter the Japanese offensive. After years of lobbying, the results were:

The capital gains tax rate was lowered from 49% to 28%, and efforts were made to encourage pension funds to invest in venture capital. However, the government was unwilling to offer direct assistance.

By June 1985, the SIA finally formulated a view that made Washington uneasy, turning the tide.

The SIA's viewpoint was that the weakening of the American semiconductor industry would pose a significant risk to national security.

Isn't Japan an ally of the United States? With Japan's semiconductor industry rising and America's declining, it looks like a simple case of money changing pockets. How could this threaten US national security?

SIA's logic chain is as follows:

Because superweapon technology is inseparable from super electronics technology, and super electronics technology is inseparable from the latest semiconductor technology (this is correct);

If US semiconductor technology lags behind, the US military will be forced to use foreign products, including Japanese products, for key electronic components (interesting);

Foreign sources are unreliable. During wartime, they would cut off supplies to the US, and during peacetime, they would supply goods to US rivals like the Soviet Union (the Japanese will probably cry when they see this);

Therefore, allowing Japan to dominate the semiconductor chip field is equivalent to sacrificing national security... (the Japanese hearts are bleeding now).

Previously, SIA lobbied for 7 years, and the government's response was always: the US is a free market, and government power should not interfere in business operations.

This time, with SIA's "national security argument," the US government had an epiphany, going from sluggish to rapid action with astonishing efficiency:

In the spring of 1986, Japan was found guilty of dumping ROMs; in September, the US-Japan Semiconductor Agreement was signed, requiring Japan to open its semiconductor market and guarantee that foreign companies would obtain a 20% market share within five years; shortly after, a 100% punitive tariff was imposed on $300 million worth of chips exported to Japan; the acquisition of Fairchild Semiconductor by Fujitsu was vetoed.

This series of actions by the Americans set at least two precedents: the first time the economic interests of an ally were attacked globally; and the first time that national security was used to transform a trade dispute from an economic issue into a political and economic one.

Clyde Prestowitz, the chief US trade representative in Asia responsible for negotiations with Japan, criticized Japan's semiconductor industry policy as unreasonable while simultaneously praising it, "So I told the US government that we should also adopt the same policies and measures as Japan."

Regarding this double standard, Yuji Kaminaga, who had worked for many years in research and development at Hitachi and Elpida, angrily stated in his book: "This person is simply outrageous!"

With the signing of the US-Japan Semiconductor Agreement, Japan's semiconductor chip industry, which was at the peak of the wave, turned sharply downward.

Japan's semiconductor chip industry fell from a high of 40% in 1986 to 15% in 2011, losing more than half of its market share. DRAM was hit hardest, falling from a high of nearly 80% global market share to a low of 10% (2010), losing nearly 70%.

It can be said that in this battle with the Americans, the Japanese basically lost all their accumulated capital, and the hard work of the entire nation for eleven years (from 1975 to 1986) was wiped out overnight.

However, the meat that the Japanese lost did not end up in the mouths of the Americans, because more than 70% of Silicon Valley technology companies cut their DRAM businesses (including Intel and AMD). After 1986, the market share curve of the Americans was a flat, dead worm, hovering around 20%.

Intel, unable to compete with Japanese companies, eventually abandoned the semiconductor memory business and entered the microprocessor field.

So, who swallowed up this huge 70% market? The answer is South Korea.

Around 1986, when Japan was being pummeled by the US, South Korea's DRAM industry took the opportunity to start, but its size was like a toddler, with no presence in the global semiconductor chip industry. Compared to Japan, South Korean semiconductor chip companies, represented by Samsung, were completely outmatched:

They couldn't break into the high-end market dominated by the Japanese and could only survive in the low-end market by relying on low prices; in terms of market size, the difference was like that between an ant and an elephant.

But Samsung understood that all trade frictions fall under the category of political economy, seizing the opportunity to overthrow the Japanese elephant.

Lee Kun-hee seized the perfect opportunity presented by the US-Japan chip war to lead Samsung semiconductors to a successful counterattack and rise to the top.

In the 1990s, Samsung faced anti-dumping lawsuits from the US, but its chairman, Lee Kun-hee, cleverly used the opportunity the Americans were using to suppress the Japanese semiconductor industry by sending a powerful public relations team to lobby the Clinton administration: "If Samsung cannot manufacture chips normally, the trend of Japanese companies dominating the market will become even more pronounced, and the reduction in competitors will further increase the price of chips purchased by American companies, which will be even more detrimental to American companies."

As a result, the Americans only imposed an anti-dumping tax of 0.74% on Samsung, while Japan was charged the maximum 100%. This is blatant favoritism.

Samsung's embrace of the US was like a stab in the back to Japan, completely eliminating them from the competition.

If Samsung hadn't delivered the coup de grace, the Japanese semiconductor chip industry might have had a chance to recover.

The Americans used the US-Japan Semiconductor Agreement to restrain the Japanese and beat them with the anti-dumping stick, but the Japanese semiconductor memory chip industry only suffered superficial injuries because more than 70% of Silicon Valley companies withdrew from the semiconductor memory chip industry, and the market was still firmly in the hands of the Japanese. After weathering the storm, they would be a formidable force again. After all, Japan remains an irreplaceable force in the global semiconductor chip industry chain.

After Samsung joined the fray and actively sided with the US, the once irreplaceable Japanese suddenly became dispensable, and the Koreans became the new favorites.

Subsequently, Samsung's DRAM "bidirectional data selection scheme" was recognized by the US Semiconductor Standardization Committee as compatible memory for microprocessors, while Japan was excluded. This allowed Samsung to smoothly ride the personal computer era driven by microprocessors, surpassing Japanese companies.

The semiconductor chip share lost by Japan almost entirely ended up in the hands of Korean companies led by Samsung.

From the DRAM share chart above, it can be seen that Japan's share plummeted, while South Korea's is a steep upward curve. The two lines, one up and one down, form a giant pair of scissors, cutting off the future of Japanese semiconductor chips.

Since then, even with the Japanese government's intensive implementation of semiconductor industry support policies and massive investment, it has been unable to turn the tide. The fate of Japan's semiconductor chip industry was sealed.

Toshiba's sale of its semiconductor division marked the complete exit of Japan from the semiconductor chip market.

Even today, some believe that the rise of South Korean semiconductor chips and the decline of Japanese semiconductor chips are the result of industrial transfer. This is inaccurate, because industrial transfer involves the relocation of production lines and factories from high-labor-cost regions to low-labor-cost regions. Japanese semiconductor companies did not transfer their production lines to South Korea; instead, they were directly replaced.

The Americans, in effect, collaborated with South Korea to restructure the global semiconductor industry supply chain, removing the Japanese from the supply chain and causing a seemingly indispensable industrial force to vanish completely.

Throughout the US-Japan chip war, the ability to restructure the global industrial chain was the key to victory in the trade war. The volume of market share did not constitute a major strength; this is one of the key reasons why Japan lost the chip war.