In November last year Taiwan Semiconductor Manufacturing Company (TSMC) gained victory in an eight-year-long legal dispute with Chinese rival Semiconductor Manufacturing International Corporation (SMIC) over the alleged theft of trade secrets.

In an out-of-court settlement, the rival chip foundries put to rest a state of enmity that had lasted a decade. The subsequent resignation of SMIC founder and CEO Richard Chang set the stage for a more friendly state of competition for a share of China's NT$560 billion semiconductor market.

As part of the settlement, SMIC will pay TSMC US$200 million. On top of that, TSMC gets an eight-percent stake in SMIC in compensation and an option to buy another 2-percent stake within three years at a price of HK$1.3 per share. Should TSMC execute the option, it will hold a 10-percent stake in the Shanghai-based chipmaker, becoming its third largest shareholder, behind Shanghai Industrial Investment (Holdings) Co. and Datang Telecom Technology Co. The victory has gained considerable attention within Taiwan's high-tech industry, but the winner itself is far from basking in its glory.

"The very first thing we wanted to do is to protect our intellectual property and our business," says TSMC vice president and general counsel Richard Thurston, who spearheaded the litigation against SMIC.

"Everything I have done over the eight years since I came here has been using intellectual property not as a clout to beat people, but to give us a competitive advantage, because we had spent billions of dollars to develop our technologies and intellectual properties," he told CommonWealth Magazine. Throughout the interview Thurston kept reiterating that TSMC's main objective was not to squeeze out its competitor by suing for ruinous royalties as other high-tech giants such as Texas Instruments or IBM had in the past.

On the contrary, TSMC was careful not to cause SMIC to bleed to death from compensation payments, because once SMIC went bankrupt and its fabs were sold off, TSMC's process technology would also be leaked to the competition. TSMC was anxious to prevent a worst-case scenario: whichever company took over a cash-strapped SMIC would fight a price war and destroy the market order. Given that SMIC's process technology comes from TSMC, TSMC would have suffered most.

"TSMC put billions and billions into the technology. If other foundries in the market can freely get that, what distinguishes us from them? Everybody can claim that they are TSMC-like," notes Thurston in explaining what drove TSMC to defend its intellectual property in court.

"Once a war situation presents itself, you can only seek to make peace, you can't seek battle," says Yi-chia Chiu, associate professor at the Department of Business Administration of National Chung Hsing University, in analyzing TSMC's strategy.

Exactly how did TSMC use war to compel peace, and peacefully prevent a war? For the past eight years, what strategies has it pursued?

Eight years ago, Thurston had just been recruited by TSMC chairman Morris Chang to design a new intellectual property strategy for the chipmaker. At the time TSMC was in a conundrum, because it knew that the newly founded SMIC was poaching its employees, but was not sure whether commercial secrets had been stolen in the process too. As SMIC built China's first eight-inch wafer fab, developed as a model enterprise with massive support from the Chinese government, powerful TSMC could only watch helplessly, since the Taiwanese government restricted investment in chipmaking plants in China.

At the time Morris Chang was frequently visiting Suzhou, Nanjing and Shanghai to sound out the possibilities of constructing an eight-inch wafer fab. At the same time TSMC filed an injunction against former TSCM manager Katy Liu on allegations that Liu had been involved in leaking layout and design plans for a 12-inch wafer plant as well as wafer process recipes to SMIC.

Subsequently, Taiwanese prosecutors raided and seized Liu's computer at her home. Documents and some e-mails on Liu's hard drive indicated that Marco Mora, then SMIC chief operating officer, had asked Liu for information on TSMC's 12-inch fab, such as process and equipment lists. But SMIC continued to deny it had obtained any TSMC technology.

Litigation, Discovery, Getting the Upper Hand

In order to substantiate its concerns and suspicions over potential leaks, TSMC decided to take legal action in the United States to get its hands on hard evidence and determine what really happened. "People ask me why I sued in the US. One of the main reasons is that I can get discovery in the US. I can't get it here. And in the case of trade secrets, you can't win a trade secret case without discovery," explains Thurston.

(Here, "discovery" refers to a formal process of obtaining information, documents and evidence from one's opponent in the preparatory stage of a lawsuit.)

While TSMC was simulating the lawsuit by doing sand table exercises, it also began to gather parts that had been manufactured by SMIC and sold in the United States, and to analyze them through reverse engineering.

Seven years ago in the fall, TSMC finally had a major breakthrough with the analysis of one particular part.

The engineers came to Thurston to complain that they had been given the wrong part.

Thurston categorically assured them that he was "200 percent confident" that he had given them an SMIC part. "And they said it can't be, because it's identical. And I said thank you very much, that's what I wanted to hear."

Thurston's suspicions had been confirmed: SMIC had infringed on TSMC's patents and misappropriated its trade secrets.

In late 2003 TSMC eventually filed a lawsuit against SMIC in the U.S. district court of Northern California. At the time SMIC already ran three eight-inch fabs in Shanghai, had acquired a Motorola eight-inch fab in Tianjin and was building three 12-inch fabs in Beijing in what it called a diamond-shaped expansion strategy.

When a year later the United States International Trade Commission (ITC), where TSMC had also filed its case, came out with a comprehensive discovery schedule, SMIC, which had not yet handed over any documents, knew it would be required to come clean. As a result, SMIC was prompted to go for a settlement. But at the time of the first settlement in 2005, TSMC did not know that SMIC actually had 15,000 TSMC documents, amounting to 500,000 pages of information.

"Quite honestly, if we had known back in 2005 they had our stuff on 0.13 (micron), 90 nanometer, then, yeah, the dollar amount would have gone up in the original settlement. We wouldn't have been satisfied with a settlement of (US$) 175 million. We would have wanted more," Thurston admits in hindsight, shrugging his shoulders.

The company had no choice but to believe the assurances of SMIC CEO Richard Chang, who claimed that it all had been "independently developed."

But in the end developments at SMIC itself were playing into the hands of TSMC. Despite government backing, SMIC had been racking up losses for nine consecutive years, prompting major shareholders to make plans for Richard Chang's replacement. (Eventually, his seat was filled by David N.K. Wang, a former chairman at Chinese chipmaker Huahong NEC.)

The time was ripe for TSMC, which had long been patiently waiting, to use its strategy number three – instead of squashing its rival, keeping it alive and securing strategic influence on the inside.

A World-class Legal Team

Crucial for TSMC's victory was that it put together a world-class legal team in the five years after filing the first lawsuit against SMIC. The company has also established a comprehensive trade secret management system, and firewall management in particular was brought up to international standards.

Thurston urges other Taiwanese companies not to take IP protection lightly. "Small or medium-sized companies need to start with the right internal analysis, to protect your basic IP, get your technology roadmap, and figure out what you want to protect," he suggests, noting that protecting IP as a trade secret is more inexpensive than filing patent applications.

He also points out that support from the CEO and top management is crucial when fighting legal disputes.

"For any legal officer, certainly for general counsel, to succeed, they need the management team to commit and believe this is the right thing to do," says Thurston in describing the unconditional backing that he always enjoyed, even though Morris Chang often presented him with sharp questions. "Morris, you know, he's very tough. He will challenge me. He will challenge me on budget, on how we are going to win, that sort of thing, but there's never a question about support."

Having recruited a foreign general to command its front line, TSMC assembled a legal team of more than twenty, and scored a victory in court that gives the chipmaker a strategic advantage most Taiwanese companies could never imagine.

"As a result of this case, TSMC will be able to control China's contract chipmakers. Not only are they SMIC's shareholder, but they can also use patent assignments to control SMIC in the long term. It's like they can threaten SMIC, saying, 'If you don't behave, we will terminate our patent assignments and none of your products will be able to leave the fab'," concludes a veteran legal expert from the semiconductor industry.

Making Beijing a Winner

TSMC's strategy of settling out of court has another added bonus in that it makes the Chinese government look like a winner too.

Eager to develop a homegrown semiconductor industry, the governments of Beijing, Shanghai, Wuhan, Chengdu and Guangzhou had all dug deep into their own coffers to support SMIC. But frustrated with SMIC's lackluster performance, the major shareholders judged that Richard Chang's mission – laying the groundwork for a Chinese semiconductor industry and overseeing fab construction – had come to an end.

The next step for SMIC should be to accelerate R&D and increase innovation in IC design. The specific role of TSMC in this process remains to be seen, but the current sentiment within SMIC's board of directors is receptive to leadership from its new Taiwanese shareholder.

As one professional manager who used to be involved in SMIC's operations puts it: "As long as there are people within TSMC who are able to help SMIC take off, why not? There's nothing to lose."

In the past, TSMC disrupted the value chain in the chipmaking industry with its unique business model of a dedicated foundry. In the future, it will be able to realize its ambitions in China by nurturing Chinese IC design with an open innovation platform. In a tacit understanding, TSMC and the Chinese government fought a perfect intellectual property battle, both emerging as winners.

Translated from the Chinese by Susanne Ganz