Who helped Samsung become chip giant

Through the 1980s, Japan pushed the United States — the birthplace and longtime leader of the semiconductor industry — out of the world memory chip market, and the damage ran deep enough that Intel gave up DRAM, the short-term memory that holds a computer's working data.

American chipmakers came to believe Japan could not be beaten in memory chips, and Washington and the industry began looking for ways to counter it. Two questions followed: how to contain Japan, and where to find a viable challenger to it.

The country that emerged as a memory chip powerhouse in the 1990s was Korea, and Intel had a hand in how Korea overtook Japan to dominate the world market.

Chip business nearly dropped

"Hey, drop it."

Samsung Group founder Lee Byung-chul shook his head in disbelief while he sat through a report on the semiconductor company his son had bought. The son was Lee Kun-hee, former Samsung chairman. Lee, then a director at Samsung's broadcasting branch TBC, had acquired Hankook Semiconductor in 1974 after the chipmaker went bankrupt due to financial strain.

Hankook Semiconductor, an early Korean chipmaker, was the country's first, founded by Kim Gyu-han, president of the telecommunications equipment importer Kemco, and Kang Ki-dong, a pioneer in the country's semiconductor industry.

It was the first company in Korea to process wafers — the discs of raw material on which chips are built — into transistors, which it sold to electronics manufacturers. A transistor works like a switch controlling electrical signals, and no electronic product can be built without one.

Hankook Semiconductor, however, could not survive the first oil shock, which struck after the 1973 Middle East war sent crude prices soaring.

Lee Kun-hee had recognized this dependency — no transistors, no electronics — and saw room for growth in the industry. While the oil shock had left the Samsung Group short of cash and in no position to enter a new line of business, he financed the purchase of half the shares with personal funds. He bought the remaining half in 1977 and renamed the company Samsung Semiconductor the following year.

Samsung Electronics' official corporate history records how little impact the purchase had. Samsung insiders did not know the acquisition had happened. The forgotten business resurfaced in 1979, when Samsung Electronics absorbed Samsung Semiconductor.

The report that reached Lee Byung-chul put monthly output at 100 million transistors, priced at 1 to 2 cents apiece. He shook his head. He took it for a low-margin business with limited revenue potential. Lee Kun-hee and Samsung Electronics' management persisted, arguing that a component the electronics industry could not function without would only grow more important. The founder eventually relented.

Only chips to feed Samsung

Lee Byung-chul, who had taken a dim view of chipmaking, changed his mind in 1980. Stopping in Tokyo, he met Hidezo Inaba, a longtime acquaintance. Inaba had studied economics at the University of Tokyo, served as director of the National Economic Research Institute and as president of the Sankei Shimbun, and took part in shaping Japanese economic policy after World War II.

According to "Hoam Jajeon," Lee Byung-chul's autobiography, Inaba answered his question — what Japanese industry had to do to survive — by pointing to semiconductors.

"Japan's way to survive depends on advanced technology industries that are lightweight, thin and compact," he said. "In particular, it must concentrate on semiconductors. The Japanese government too is nurturing this as a strategic industry."

Inspired by Inaba's insight, Samsung's founder watched the industry closely. What sharpened his conviction was his 1982 visit to the United States. In San Francisco, he took in the tall buildings and the cars crowding the streets and felt the abundance of an economic superpower, then was startled to see that even this America was losing its own market to Japanese chips. In the autobiography, he called it "a strange phenomenon."

Visits to Hewlett-Packard and IBM plants in the United States solidified his decision. In "Chip War: The Fight for the World's Most Critical Technology," Chris Miller, a professor of international history at Tufts University, describes an employee at HP telling Lee that the reason a company, which started in a garage, had become a giant was "thanks to semiconductors."

IBM showed him its chip production facilities. No company readily opens plants it treats as confidential. Puzzled, he asked why they were willing.

"Just by looking, you can't copy it," an IBM employee told him. The remark reportedly spurred him to act.

Back home, in October 1982, Lee assembled a semiconductor and computer business team. In March the following year, he called Hong Jin-ki, former chairman of the JoongAng Ilbo, then Samsung's media branch, told him the company would invest on a large scale in memory chips, and declared the expansion publicly through the outlet.

Intel turns to Samsung

In expanding the business, Samsung drew help from both the United States and Japan.

According to "Hoam Jajeon," it brought in 64K DRAM technology from Micron Technology in the United States and 16K SRAM technology — the faster memory a processor draws on for immediate work — from Japanese electronics maker Sharp. Engineers who had worked at Intel, IBM and Zilog joined the company. Samsung set up a U.S. subsidiary in Santa Clara, California, and researched chip technology in Korea and the United States at the same time.

Up to that point, Intel was uneasy about an American company partnering with Samsung. Gordon Moore, Intel's co-founder, worried that Micron's technology transfer meant core American chip technology leaking out.

The Korean government then granted special zoning approval for the Giheung plant site, which it had designated for another purpose. Employees gave up the Lunar New Year holiday, the country's biggest, and worked without a day off, and six months of the construction brought the Giheung plant online in 1984.

Financing followed — what Lee's autobiography calls "the special understanding and cooperation of financial institutions." In short, Samsung's semiconductor business rested on the sacrifice of its workers and on government support: the plant site and the money.

Samsung went on to develop a 64K DRAM and put it into production, the third in the world after the United States and Japan. Korea, which had trailed the two by more than a decade in chip technology, cut the gap to four years.

The same year, Samsung Electronics also developed a 256K DRAM with far greater storage capacity, and said nothing about it publicly. According to the founder's autobiography, the government asked Samsung to keep it secret for fear of provoking foreign companies.

Around that time, the United States and Japan were locked in a chip war. Japan was raising its chip yields while aggressively cutting prices, squeezing American chipmakers.

If word of Korea's 256K DRAM had come out, Samsung would most likely have faced pressure from both American and Japanese competitors, and it could have collapsed before it reached the market. There was also the risk that the 64K DRAM it had just put into mass production would be made obsolete by the 256K. Samsung had no reason to bring the 256K to market early, and so no reason to announce that it existed.

Intel, under pressure as its DRAM revenue plunged amid intense Japanese competition, exited the DRAM business in 1986. It concentrated instead on the central processing unit, the brain of the computer. At the same time, Samsung Electronics began producing EPROM chips for Intel — erasable programmable read-only memory, a read-only chip that can still store and erase data — and Intel sold them under its own brand.

The arrangement suited both sides. Intel could lower costs by having EPROMs made in Korea, where labor was cheap, and Samsung secured a stable buyer.

By industry accounts and Miller's book, Samsung Electronics' chips at the time lagged Japanese ones in quality. Intel judged all the same that Samsung was a useful counterweight to Japanese chipmakers.

A former Intel official said Samsung Electronics' semiconductors were lower in quality than their Japanese counterparts at the time, but were reasonably priced and good enough to take market share from Japanese companies. The official said this led Intel to see Samsung Electronics as a good partner.

Intel as partner, rival

U.S. policy aimed at checking Japan's chip industry helped Korea's as well. In 1986, the United States concluded a semiconductor agreement with Japan. At its core was a quota system that limited exports of Japanese DRAM to the American market, together with a price floor that prevented dumping and weakened Japanese price competitiveness. By restricting both volume and price, the United States constrained Japanese suppliers' trade practices. For Korea it was a chance to build.

This article from the Hankook Ilbo, the sister publication of The Korea Times, is translated by a generative AI system and edited by The Korea Times.