Morris Chang: Foundry Father

Even in the very anomalous category of tech entrepreneurs who've become pop stars, Morris Chang is an anomaly. He's 79 years old, and unlike Steve Jobs, who makes stylish consumer gear, or Mark Zuckerberg, who runs the world's largest social network and is the subject of a major motion picture, Morris Chang runs a semiconductor foundry. It's a big one, to be sure: Taiwan Semiconductor Manufacturing Co. (TSMC), in Hsinchu, is by far the world's largest foundry company, with annual revenues last year of about US $13.3 billion. And yet, the foundry business isn't what most people would regard as the glamour sector of the semiconductor industry—a foundry is basically a fab-for-hire. Companies create their own designs and pay the foundry to manufacture the chips.

Chang is credited with pioneering the foundry concept, in 1987, when he founded TSMC with backing from the Taiwanese government and Netherlands-based Philips Electronics. For that achievement, Chang will receive the 2011 IEEE Medal of Honor.

Like many radical ideas, the foundry concept did not seem so brilliant at first. But the business took off, with TSMC dominating it in a way that few companies have ever dominated any industry. Today TSMC's revenues account for almost half those of the entire foundry industry. And its operating profits—about $5.3 billion last year—were an astounding 90 percent of the whole global foundry industry's.

As TSMC's revenues soared, so too did Morris Chang's stature in Taiwan, where he is a national hero unlike any other. His smiling face appears on billboards endorsing consumer gadgets, real estate, and other goods. Even in his own company, work stops and people stare when he walks down a hall.

Checking out of my hotel in Hsinchu, I decide to test Chang's fame. "Is this your first visit to Taiwan?" the desk clerk asks politely. She looks to be about 25 years old. "Yes," I say. "I was here to meet with Morris Chang." Her voice jumps nearly an octave. "Oh, really? How exciting!" she yelps. "That must have been so, so…." Her English fails her. I feel sorry that I hadn't mentioned it sooner; I could have gotten his autograph for her and really made her day.

Necessity is the mother of invention, as the old saying goes. Chang wasn't trying to reinvent the semiconductor industry when he started TSMC. He did it because government officials in Taiwan wanted him to start a semiconductor business, and with Taiwan weak in both design and marketing, he didn't see any other way.

Chang had no idea that he was launching a technological revolution, the likes of which the semiconductor industry had not seen since, arguably, the rollout of user-programmable logic devices during the previous decade. Suddenly, entrepreneurs could create semiconductor businesses around chips without the huge expenditure of cash and effort it takes to open a semiconductor fabrication facility; all they had to do was design a chip and then market it. Thus was born the "fabless" semiconductor industry, which racked up $73.6 billion in revenues last year. A surprising number of today's hottest high-tech companies got their start in TSMC's factories—telecommunications pioneers Broadcom and Qualcomm, graphics powerhouses Nvidia and ATI, mobile device innovator Marvell, programmable logic creator Altera—and their products continue to roll off TSMC's manufacturing lines today.

"Morris Chang completely changed the landscape of the semiconductor industry," says James Plummer, dean of the school of engineering at Stanford. "He enabled start-ups to start with a few million dollars rather than a few hundred million. That makes a huge difference."

---

"Morris's foundry model unleashed an army of engineers and made people's dreams come true." --ROAWEN CHEN, vice president of manufacturing, Marvell Technology Group

---

Chang aspired to be a writer as a child. Born in Ningbo, just outside of Shanghai, Chang moved several times as a youngster: to Nanjing, for his father's work in banking; to Canton (now Guangzhou); to Hong Kong, to escape the Japanese bombing during the second World War; and eventually to Chongqing in southwestern China. He thought he'd be a novelist, maybe, or a journalist, but perhaps it was "just a young man's fancy," he says. Then, in 1949, Chang, with the help of an uncle in Boston, was accepted at Harvard University.

"My reaction entering Harvard was sheer ecstasy, almost disbelief," he recalls. "What a country! The United States was at its peak in its moral leadership, in its political leadership in terms of democracy—and it was the richest country in the world." It was then he gave up his childhood dream of becoming a writer. "In the early '50s in the United States, there were Chinese laundrymen, Chinese restaurateurs, Chinese engineers, and Chinese professors," he says. "Those were the only respectable professions for Chinese—no lawyers, no accountants, no politicians"—and certainly no writers. Chang decided he'd be an engineer, declaring "applied physics and engineering sciences" as his major.

After a year at Harvard, he realized that the school's general technical education wasn't going to land him a top engineering job, so he transferred to MIT. He majored in mechanical engineering because it seemed the most general of the specialties. After failing his Ph.D. qualifying exam—he simply didn't study hard enough, he says—he went out looking for jobs. Ford Motor Co. offered him $479 a month; Sylvania Electric Products' then tiny semiconductor division offered $480. Chang asked Ford to beat the Sylvania offer. It refused. He took the job with Sylvania. It was 1955.

His first assignment was to work on germanium transistors, first on improving manufacturing yields and later on developing devices. Chang started working toward yield improvement by looking at bonding, the process by which electrical contacts are attached to the transistors. At the time, a technician on the assembly line connected those electrodes by soldering on a wire, and the heat often damaged the transistor. Chang devised a way to connect the wire using indirect heat instead.

It was a basic mechanical engineering problem; he didn't have to understand semiconductor theory to figure it out. But he started studying semiconductors anyway, using as his text Electrons and Holes in Semiconductors with Applications to Transistor Electronics (1950) by William Shockley.

Parts of the book baffled him. But he soon discovered that a senior engineer at Sylvania spent hours every evening drinking heavily at the bar in the same hotel where Chang was living, in Ipswich, Mass. So Chang's typical evening in those days went like this: He'd spend a couple of hours holed up in his room poring over a section of the book and trying to solve the problems on his own. Then he took his questions down to his friend in the bar. "He didn't solve all my problems, but he solved enough so I could move ahead. He was my main teacher," Chang says.

Chang eventually moved on to designing new transistors. He attended technical conferences, served on a standards committee. By 1958 he'd grown tired of developing what he saw as good transistors that went nowhere because of Sylvania's lack of marketing prowess, so he accepted an invitation to visit Texas Instruments, in Dallas.

"My future boss actually met me at the airport," he recalls. Then 27 years old, Chang was struck by the relative youth and exuberance of TI's corporate culture. The majority of the people in the semiconductor business at Sylvania had moved over from vacuum tubes and were in their 50s. He took a job at TI in production, working on four transistors being manufactured for IBM, for which IBM had provided the design and process. In essence, it was an early foundry deal.

One of the transistors, a particularly difficult one to produce, was manufactured by both TI and IBM, which operated a pilot line elsewhere just for that device. Chang took charge of TI's production line for that troublesome transistor. At IBM, at best 10 percent of the transistors produced on the pilot line worked, while at TI, the yield on Chang's line was pretty much zero, occasionally getting as high as 2 to 3 percent. Months passed while Chang tinkered with the process.

"The supervisor was concerned. The operators were concerned. Everybody was concerned," Chang recalls, but he says he wasn't too worried, because he knew that getting a 10 percent yield was at least possible.

Finally, about four months into the project, Chang began testing yet another combination of process parameters, changing the temperature and the time in the furnace and experimenting with different alloys and diffusion materials. The yield jumped to more than 20 percent.

Suddenly, even TI president Pat Haggerty knew his name. IBM thought Chang had just gotten lucky, but when the company sent engineers to talk to him, he described the theories he'd been testing and explained why his experimental process worked. The achievement propelled him into his first management job, creating a germanium transistor development department with 20-plus engineers.

Chang loved management. He liked being able to do bigger things than he could as an individual, and frankly, he aspired to a higher standard of living—what he calls a "reasonably comfortable life." He was 28 years old and entertaining dreams, and not unrealistic ones, of being vice president of R&D someday. His boss at the time told him that if he aspired to that job, he'd need to get a Ph.D.—not because his superiors cared but because the R&D teams at TI definitely would. So, in 1961, at age 30, Chang went to Stanford, with TI paying his tuition and expenses on top of his full salary while he pursued his degree full time.

By spring of 1964 Chang was back in Texas with his Ph.D. in electrical engineering. He quickly moved up, eventually having all 3000 people in the germanium transistor department under him. He had a lot to learn—product marketing and finance, for example, weren't taught in engineering school.

For the next 19 years Chang moved through various management jobs at TI, working his way up. He pioneered the then controversial idea of pricing semiconductors ahead on the cost curve—that is, sacrificing profits early on to grab market share to achieve manufacturing yields that in turn would yield greater long-term profits. By 1972 he was running the company's entire semiconductor business—and aspiring to take over as CEO one day.

But in the mid-'70s, TI decided to move more heavily into the consumer electronics business, with calculators and wristwatches. In 1978 the company picked Chang to run the consumer business. He was proud of Speak & Spell, the breakthrough speech-synthesis toy that came out that year, which used the world's first single-chip speech synthesizer [see "25 Microchips That Shook the World," IEEE Spectrum, May 2009]. But overall, he says, the move was a mistake, for the company and for him personally.

Mark Shepherd, then chairman and CEO of TI, agreed with the prevailing wisdom at the time that a good manager could manage anything. In this case, Chang says, "I think he was wrong. I found the consumer business to be very different. The customer set—completely different. The market—completely different. And what you need to get ahead in that business is different, too. In the semiconductor business, it's just technology and costs; in consumer, technology helps, but it's also the appeal to consumers, which is a nebulous thing."

In 1983, with TI's consumer business struggling and Chang sidelined as head of "quality and people effectiveness," he knew his path at TI no longer led to the executive suite. He decided to leave.

He had no real plans, but he was confident he'd be a hot job prospect. And so he was; after his departure hit the news, his phone started ringing, he says, and didn't seem to stop for days. Eventually, he had two offers he seriously considered—one from a venture capital firm and one from General Instrument, an East Coast manufacturer of semiconductors and cable set-top boxes.

He joined GI in 1984 as president and chief operating officer and moved to New York City; he'd once dreamed of living there and walking to work. He moved to an apartment on Fifth Avenue, just a few blocks away from his office in the General Motors building. And he had, he thought, free rein to build up the company's R&D operation.

He didn't. The company, he says, wasn't really serious about R&D, preferring to buy businesses rather than grow them internally. And the pleasure of walking to work paled quickly in the rain and winter slush. Meanwhile, he had separated from his then wife, Christine.

He left GI after just a year. Once again, he quit without a plan. "After these two setbacks, at TI and at GI," he says, "I did not think that my aspiration to be the CEO of a major U.S. company was in the cards." He contemplated going to a venture capital firm, but, he says, "I didn't think I would be very good at it."

Then a government official in Taiwan, K.T. Li, contacted him. Chang had met Li several years previously, when TI was negotiating to build an assembly plant in Taiwan. Li was looking for someone to manage the decade-old Industrial Technology Research Institute (ITRI), turning it into the Bell Labs of Taiwan.

Just about everybody he knew told him not to do it. His soon-to-be ex-wife thought he'd lost his mind for even seriously considering the offer. But Chang, thinking he had little to lose, decided to give it a try. He took over as president of ITRI in 1985 and immediately began shaking things up. For starters, he broke the "iron rice bowl," challenging the assumption that once you got a job in a government organization, you'd be set for life. Instead, Chang instituted a review process that put the lowest 2 percent of performers on probation, with the possibility of dismissal looming if they didn't improve. He also began transitioning ITRI toward partial funding by industrial contracts, not just by government subsidy. Not surprisingly, these changes didn't go over well with the staff. Hate mail began to pour in to Taiwanese legislators, complaining about Chang and his policies. "Back 25 years ago," he says, "they considered me a foreigner who suddenly became their boss. They were scared of me."

"He's known not to tolerate fools," says Chenming Hu, a professor emeritus at the University of California, Berkeley, and former chief technology officer for TSMC. "And that engenders both respect and fear."

The situation at ITRI had only just started to settle down when Li came back with another proposal. The government, Li told him, wanted Taiwan to have a semiconductor company and thought that if anyone could get it off the ground, Chang could.

Chang knew it wouldn't be easy. TI was still the dominant company in the semiconductor business, with Intel rising fast and National Semiconductor Corp. and Advanced Micro Devices also strong players. Just to survive among these giants would be a challenge.

And he would be starting at a huge technical disadvantage. In 1975, ITRI had licensed from RCA Corp. a 7-micrometer semiconductor processing technology that at the time was already out of date. By 1985, research at ITRI had advanced to a 3-µm process; however, the rest of the industry had moved forward even faster—about two generations ahead, with advanced products being manufactured in 1.5-µm technology.

But Chang said yes anyway. "It's like in the movie The Godfather," he says, "an offer you can't refuse," explaining that to say no would have tagged him as someone completely without ambition and would likely have dimmed his future in Taiwan.

Taking on the semiconductor behemoths head-to-head was out of the question. Chang considered Taiwan's strengths and weaknesses as well as his own. After weeks of contemplation he came up with what he calls the "pure play" foundry model.

In the mid-'80s there were approximately 50 companies in the world that were what we now call fabless semiconductor companies. The special-purpose chips they designed were fabricated by big semiconductor companies like Fujitsu, IBM, NEC, TI, or Toshiba. Those big firms drove tough bargains, often insisting that the design be transferred as part of the contract; if a product proved successful, the big company could then come out with competing chips under its own label. And the smaller firms were always second-class citizens; their chips ran only when the dominant companies had excess capacity.

But, Chang thought, what if these small design firms could contract with a manufacturer that didn't make any of its own chips—meaning that it wouldn't compete with smaller firms or bump them to the back of the line? And he realized that this pure-play foundry would mean that Taiwan's weaknesses in design and marketing wouldn't matter, while its traditional strengths in manufacturing would give it an edge.

TSMC opened for business in February 1987 with $220 million in capital—half from the government, half raised from outside investors. Its first customers were big companies like Intel, Motorola, and TI, which were happy to hand over to TSMC the manufacture of products that used out-of-date technology but were still in some demand. That way, the companies wouldn't have to take up their own valuable fab capacity making these chips and would face little harm to their reputation or overall business if TSMC somehow failed to deliver.

Soon, Chang says, start-up companies, which would live or die depending on TSMC's manufacturing runs, signed on. Early customers like Broadcom, Marvell, Nvidia, and Qualcomm, Chang says, "started with us when they were small," and being able to tap into TSMC's manufacturing prowess was a big reason for their success.

"Not having us would have slowed down innovation in the industry," Chang maintains, pointing out that those little companies would likely have had to invest in manufacturing capacity instead of R&D or share their intellectual property with a TI or an IBM.

Jen-Hsun Huang, cofounder, president, and CEO of Nvidia, says that TSMC's birth enabled all sorts of creative ideas—in areas like networking, consumer electronics, computers, and automotive technology—to be turned into successful companies, because "the barriers to getting your chips built, to realizing your imagination, disappeared."

Huang started Nvidia assuming that the company's chips would have to be made in partnership with a traditional semiconductor company; Nvidia initially partnered with SGS Thomson Microelectronics (now STMicroelectronics). As those things go, he says, it was a good enough partnership, but there was no denying that the two businesses had different interests.

"When I learned about Morris and TSMC," Huang says, "I was thrilled. I wrote him letters—the only way I had of contacting him. And one day he called me." The two met soon after that phone call and, for Huang, the two companies were a perfect match.

"I loved that TSMC's intentions were pure, that their success only came with our success. Morris and I were both building our companies; Nvidia had to move fast to keep up with the competition, and TSMC kept up with our needs," he says. Today Nvidia has revenues of nearly $4 billion a year, and TSMC still does virtually all of its manufacturing.

TSMC was already well established when fabless semiconductor firm Marvell Technology Group launched in 1995. Says Roawen Chen, vice president of manufacturing at Marvell, which brought in $2.8 billion in 2010 revenues: "It would have been impossible for Marvell to start without foundries—it would have needed $100 million to build factories, the entry barrier was so high."

"Morris's foundry model unleashed an army of engineers," says Chen, "and made people's dreams come true."

Today TSMC has a market capitalization of nearly $70 billion. Chang himself received no stock for founding the company. Though he created one of the biggest companies in Taiwan, he doesn't even make the country's Forbes 40-richest list. "I like to measure a person's wealth as a portion of the shareholder wealth he created," Chang says. "On that basis, I made out really, really poorly." Still, he isn't complaining; he is well compensated, he says. And several years after TSMC's start, he invested his own savings in the company, profiting as TSMC boomed.

Life is good, to put it mildly. During the week, he lives with his second wife in an apartment in Hsinchu, about 20 minutes from TSMC's huge headquarters complex of factories and offices. On weekends, they stay at their 460-square-meter duplex penthouse in Taipei.

He is also famous, though he shrugs it off. "I don't give a great deal of stock to this thing of being recognized," Chang says. "After all, Taiwan is a pretty small place."

And at nearly 80 years of age, Morris Chang is still running TSMC full time, as CEO and chairman. He starts his days at 6:30 a.m. to check e-mails at his apartment, arriving in the office around 8:30 and working until about 6:30 every evening. He doesn't have any plans to stop. He does contemplate stepping back as CEO at some point while remaining chairman. He smiles, warming to the idea. In Taiwan, he notes, the chairman's role can be defined just about any way someone wants to.

And if Chang is good at anything, it's defining things. After all, he defined an entire industry.

This article originally appeared in print as "Foundry Father."

This story was corrected on  20 April 2011.