Shulou(Shulou.com)11/24 Report--

On the way for TSMC to grow into a global chip manufacturing giant, there is one person who can be called the soul of the company. He is founder Zhang Zhongmou.

Zhang Zhongmou was born in 1931 in Ningbo, Zhejiang province, in Chinese mainland. At that time, the mainland was on the brink of war. Eighteen years ago, he had lived in six cities and changed schools 10 times. He also experienced Japanese bombing of Guangzhou and Chongqing, fleeing Japanese-occupied Shanghai with his family during World War II and crossing the front line.

In 1949, Zhang Zhongmou went to the United States to study at Harvard University and then transferred to the Massachusetts Institute of Technology to study mechanical engineering. In 1955, after he twice failed the MIT doctorate exam, he decided to try the job market.

"looking back many years later, I thought it was the luckiest thing in my life that I didn't get admitted to MIT's PhD program at that time." Zhang Zhongmou wrote in his autobiography.

At the time, he had two good jobs to choose from: Ford and Sylvania, a less well-known electronics company.

Ford offered Zhang Zhongmou a job at the Detroit R & D center with a monthly salary of $479. Although Ford recruiters enthusiastically tried to persuade Zhang Zhongmou to accept the job, Zhang Zhongmou was surprised to find that the monthly salary offered by Sylvania was $480, $1 more than Ford.

So he called Ford and asked it to match Sylvania's monthly salary. However, Ford recruiters, who had been friendly before, became hostile and told Zhang Zhongmou not to get any more money.

In the end, Zhang Zhongmou chose the engineering position of Sylvania. There, he learned about transistors, which are the most basic components of microchips.

"that was the beginning of my semiconductor career," he said. "in retrospect, it was a great thing."

His three years in Sylvania opened the door to the semiconductor world and consolidated his enthusiasm for semiconductors. But Sylvania then struggled, which taught him a lesson and affected his later business with TSMC.

"the semiconductor industry has been a fast-paced and ruthless industry from the beginning," Zhang Zhongmou said in his autobiography of the eventual collapse of Sylvania. "once you fall behind, it becomes quite difficult to catch up."

Zhang Zhongmou worked for Texas Instruments in 1958 and moved to Texas Instruments, a new semiconductor company. The Dallas-based company is "young and energetic", with many employees working more than 50 hours a week and sleeping in the office at night.

At that time, Texas Instruments established a world-class semiconductor business, and Zhang Zhongmou became the backbone of this business. At that time, Texas Instruments continued to innovate. In the 1970s, Texas Instruments produced a chip that could synthesize human voices, leading to the famous Speak & Spell toy, a handheld device that helps children spell and pronounce.

"it was a glorious period, but it didn't last long." Zhang Zhongmou recalled.

In the late 1970s, Texas Instruments shifted its focus to booming emerging markets such as calculators, electronic watches and home computers. Zhang Zhongmou, who was in charge of semiconductor business at the time, realised that his career was heading for a "dead end".

So he decided it was time to do something different.

Three pieces of "jigsaw puzzle" Zhang Zhongmou has always kept an old book called "introduction to VLSI Systems." This is a graduate textbook describing the complexity of computer chip design.

Zhang Zhongmou, 92, held it up respectfully. "I want to show you the publication date of this book, 1980." He added that the timing was important because it was the "earliest piece" of his TSMC jigsaw puzzle, changing not only his career but also the course of the global electronics industry.

The insight he got from this textbook seems simple: a microchip is the brain of a computer that can be designed in one place but made in another. This idea ran counter to the standard practice of the semiconductor industry at that time.

If the first jigsaw puzzle in which Zhang Zhongmou founded TSMC was this textbook, then the second jigsaw puzzle was an experience when he was about to leave Texas Instruments.

In the early 1980s, Texas Instruments opened a chip factory in Japan. Three months after the production line began mass production of chips, the "yield" of the factory was twice that of the Texas plant in the United States. Yield is a key indicator of how many available chips are produced in production.

As a result, Zhang Zhongmou was sent to Japan to understand why the yield of the new factory was so high. As a result, he found that the key lies in the employees. In Japan, the turnover of highly skilled workers is astonishingly low.

TSMC's silicon wafers on display in Nanjing, but in the United States, no matter how hard Texas Instruments tries, it can't find the same level of technical personnel. At an American factory at Texas Instruments, the most suitable candidate for the executive position has a degree in French literature but no engineering background. The future of advanced manufacturing seems to lie in Asia.

In 1984, Zhang Zhongmou joined General Instruments, another chip company. There, his third jigsaw puzzle surfaced. He met an entrepreneur who later started a company that only designed chips and did not produce them, which was unusual at the time.

Eventually, he found a trend that later proved to have long-term endurance: most semiconductor companies design chips but outsource manufacturing.

The emergence of this last jigsaw puzzle coincides with the transformation of Taiwan from a labor-intensive and heavy industrial economy to a high-tech economy. At that time, Taiwan officials began to focus on developing the semiconductor industry, inviting Zhang Zhongmou, already known as a chip expert, to lead a research institute to promote innovation.

In 1985, Zhang Zhongmou, then 54, left the United States for a place he didn't know until after several visits to the Texas instrument factory.

"I certainly didn't intend to stay in Taiwan for so long," recalls Zhang Zhongmou. "I thought I might go back in a few years. I really had no plans to set up TSMC or any companies in Taiwan."

Just weeks after Zhang Zhongmou arrived in Taiwan, Li Guoding, an official known as the godfather of science and technology in Taiwan, gave him the task of making government-led chip projects commercially viable.

When he was assessing Taiwan's strengths and weaknesses, he sensed an opportunity. "(combined with my experience at the Texas Instruments Japanese factory), I have concluded that compared with the United States, the situation in Taiwan is more like Japan." He said.

Founded TSMC in 1987, Zhang Zhongmou formally established TSMC. He has a clear business model in mind: TSMC will make chips for other companies, not design them. That means he only needs to win the support of industry insiders and then focus on what he is good at: manufacturing.

From the beginning, he planned to let TSMC enter the global market. He introduced a professional management system that is not common in Taiwan. In order to create an international environment, TSMC uses English for internal communications.

Zhang Zhongmou's foresight has been verified. As semiconductors become more complex and more expensive to produce, only a few companies have the ability to try to make them. The manufacturing process involves hundreds of steps that require advanced lasers and chemical processing to create tiny channels for electronic signals and perform the most basic calculations for computers. The cost is huge.

After years of fierce fighting, others dropped out, but Zhang Zhongmou persevered. If TSMC can take advantage of economies of scale to attract enough customers, it has a chance to beat Intel and Samsung.

In 1997, Zhang Zhongmou hired Jiang Shangyi, a new head of research and development, who asked him to target Intel, the industry leader at the time.

"our goal is to be number one, which is second to none." Zhang Zhongmou said.

Jiang Shangyi was surprised by this goal. "to be number one, you have to spend three times more than your nearest competitor." He replied at the time, implying that becoming an industry leader would be an overly noble and expensive goal.

"it may cost three times as much, but I do want to invest enough money to make us number one." Zhang Zhongmou said. He is prepared to be patient, and even after he resigned as chairman of TSMC CEO in 2005, his goals remained the same.

In April 2009, TSMC employees angry at their dismissal protested by setting up a protest camp on a tree-lined playground in Taipei's quiet Dazhi residential area. the camp is near the upscale apartment building where Zhang Zhongmou lives.

As night fell, protesters spread sleeping bags next to slides and climbing shelves, covered with a big sign that read "TSMC lie, lie". In the company's more than 20-year history, TSMC has never laid off staff. However, after the 2008 financial crisis, Mr Zhang's successor, Cai Lixing, began to fire employees.

Zhang Zhongmou, then 77, decided he could no longer stand idly by. As a result, he returned to the mountain, rehired the talent Mr Tsai fired and more than doubled TSMC's spending.

At that time, the semiconductor industry was in a difficult time, and Zhang Zhongmou's move was not appreciated by investors. "when I heard the news, I felt like hitting my head against a wall," recalls Sun Youwen, a former head of investor relations at TSMC. "

But Zhang Zhongmou's gamble paid off. In 2010, he received a phone call. This phone is important enough to accelerate TSMC's growth and consolidate the company's lead over Samsung and Intel.

Cook clinked glasses with Zhang Zhongmou on the phone from Jeff Williams, senior vice president of Apple. Williams contacted Zhang Zhongmou through his wife, Zhang Shufen. Zhang Shufen is a relative of Terry Gou, founder of Foxconn, Apple's largest contract manufacturer.

After the call, the four men had dinner on a Sunday and started negotiations the next day. Apple has worked with Samsung to make microchips for the iPhone, but the company is looking for new partners, in part because Samsung has become Apple's main competitor in smartphones. By contrast, TSMC does not compete with customers, so it has a good chance of winning the contract.

The negotiations lasted several months. "the contract itself is very complicated," Zhang Zhongmou said. "this is the first time we have encountered this kind of situation."

At one point, Apple suspended negotiations for two months. Zhang Zhongmou heard at the time that Intel might have intervened.

So a worried Zhang Zhongmou flew to San Francisco to meet with Apple CEO Tim Tim Cook, who reassured him. In 2013, Paul Paul Otellini, then Intel's CEO, said in an interview that he turned down the opportunity to make chips for the iPhone because Apple didn't give enough money.

Zhang Zhongmou would not have made the same mistake. Although Apple asked TSMC to offer better terms and lower prices than other customers in the contract, he understood that the size of Apple's contract would help TSMC quickly surpass its competitors. Zhang Zhongmou learned this from Bill Bill Bain, who founded Bain, a consulting firm of the same name, when he worked at Texas Instruments.

Bain, then a consultant at the Boston Consulting Group, had worked in the office next to Zhang Zhongmou for nearly two years. He analyzed Texas Instruments' production and sales data and concluded that the more the company produces, the better its performance will be.

When the TSMC deal with Apple was completed, Zhang Zhongmou borrowed $7 billion to build production capacity to produce millions of chips for iPhone.

In the years that followed, Apple briefly switched back to Samsung to produce iPhone chips, but TSMC has become its main chipmaker. Apple is now TSMC's biggest customer, accounting for about 20% of TSMC's revenue.

Even now, Zhang Zhongmou is still cautious about disclosing TSMC customer information. After his office started telling Apple's story, he wondered if he had said too much. "I don't think this content has exceeded Apple's limits." He said.

Mr Williams, who is now Apple's COO, said in a statement that Mr Zhang had "pushed the semiconductor industry to a new level".

In 2018, 86-year-old Zhang Zhongmou retired again. By that time, TSMC had succeeded in an area where other companies lagged behind: mass-producing chips the size of an electronic channel the size of an DNA double helix. This makes Zhang Zhongmou believe that he has realized a key concept of TSMC: technology leadership.

Huang Renxun's dignitaries in Zhang Zhongmou's Taipei office are covered with awards and photos of him with world leaders. It includes a framed cartoon presented by Jensen Huang, the founder of Nvidia, depicting their intimate relationship.

If Apple promoted the development of TSMC, Zhang Zhongmou helped Yingwei achieve the most important artificial intelligence chip design company in the world.

There is another story behind the cartoon of Zhang Zhongmou and Huang Renxun. In the mid-1990s, when Nvidia was a startup, Huang Renxun wrote a letter to Zhang Zhongmou asking if TSMC was willing to make chips for it. After a phone call with Huang Renxun, Zhang Zhongmou agreed to produce chips for Nvidia.

"I like him." Zhang Zhongmou said of Huang Renxun.

Taking advantage of this cooperation opportunity, Zhang Zhongmou helped promote the artificial intelligence revolution in the United States. Through TSMC's production, Yingwei has reached a market capitalization of US $1 trillion for the world's most important artificial intelligence chip design company. Breakthroughs such as generative artificial intelligence rely on a large number of Nvidia chips to find patterns from huge amounts of data.

In 2018, Huang Renxun delivered a speech at Zhang Zhongmou's retirement party. He said that without TSMC, Nvidia would not exist. Huang Renxun wrote in a cartoon given to Zhang Zhongmou: "your career is a masterpiece-Beethoven's Ninth Symphony."

For Zhang Zhongmou, the last notes of this masterpiece have not yet been played. As an old man in his 90s, he is still healthy, although he can no longer smoke a pipe after he implanted a stent in his heart a few years ago. Smoking a pipe used to be a symbol in his picture.

Zhang Zhongmou still keeps a Bloomberg terminal in his office. He also makes regular public appearances across Taiwan to discuss global politics and economy.

If there is any regret, it is that he is no longer able to lead TSMC to meet new challenges. But he believes that it is reasonable for him to retire in 2018, which is driven by technology rather than other factors.

"I was really convinced that we had achieved technological leadership," he said of retiring in 2018. "I don't think we're going to lose it."

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.