SMIC: Huawei’s most critical friend

Domestic chip manufacturing leader SMIC international return A moment comes. On July 16, SMIC Integrated Circuit Manufacturing Co., Ltd. officially landed on the board, stock code 688981. Becoming the first company in China to simultaneously realize the “A-H” company, it will also become a leading A-share technology company and one of the highest-capsemiconductor companies. By midday, SMIC was trading at 87.98 yuan, with a market value of 627.863 billion yuan.

Reporter: Peng Xin Lu Keyan

SMIC: Huawei's most critical friend

SMIC, established in 2000, is one of the most special in the history of China’s semiconductor industry.

At that time, China in the field of semiconductor manufacturing has been hit twice, the government led the “908” and “909” projects to break through the bottleneck of semiconductor production, but under the western technology blockade, both failed.

SMIC, founded by Zhang Weijing, has quickly become a leader in the semiconductor industry in the Yangtze River Delta region, becoming the world’s fourth-largest chip contracting company.

Two decades on, under the background of Sino-US trade conflict, Huawei’s chip-off event fermentation, the return of hard technology, the pursuit of capital, a collection of factors, the company has once again become the focus of public opinion.

On May 5, SMIC announced an IPO at CST. On June 1st, the Shanghai Stock Exchange formally accepted its application for a listing, and then completed the first round of inquiries in just one week with lightning speed, taking just 18 days from acceptance to the meeting, an unprecedented 18 days.

Since its establishment, the chip foundry enterprise has experienced a series of losses, patent litigation, founder out, control struggle and a series of changes. To be able to have today’s status, behind zhang Weijing, Jiang Shangzhou, Wang Ningguo and many other semiconductor people’s grass-making struggle, hard work.

The company is destined to play a leading role in the torrent of the great times.

Tough grass-creation

In January 1948, Zhang Was born in Nanjing, and went to Taiwan with his parents before he was over the age of one. His father, who worked in metallurgy in his early years, left the mainland with more than 200 technicians who built steel mills to work at Kaohsiung’s “60 Arms Factory”, the soul of Taiwan’s early smelting technology.

Zhang Weijing’s mother was involved in the founding of Jinling Women. She attaches great importance to the cultivation and education of children, from an early age requires her son to study hard, learn the ability to serve the country. In Taiwan, Zhang went from elementary school to college.

In 1970, after graduating from the Department of Mechanical Engineering of Taiwan University, Zhang studied in the United States, and in 1974 graduated from the State University of New York, Buffalo, usa in a master’s degree in engineering, to work for Texas Instruments. At the time, Texas Instruments was the world’s largest chipmaker, representing the highest standards in the chip industry.

Zhang Weijing began his chip career at that time. When he moved to Texas Instruments, TSMC founder Zhang Zhongmou was already a senior vice president of the company. He is responsible for the management of the consumer products division, under the control of more than 1000 people, Zhang Weijing is one of them.

During his time at Texas Instruments, Zhang Had a lot to gain, and his boss, the research and development boss, was Jack Kilby, the big-name inventor of integrated circuits, and Zhang Weijing spent four years in Kilby’s lab, learning a lot of ways to live.

Another teacher who had a great influence on Zhang Weijing was Dr. Zhai Zifan, who was the world’s top chip factory construction master at the time, and when he was the director of the factory, he would have come back to life from the fab, which was on the verge of collapse, and had always been a model in Zhang’s mind.

Following with Zhai Zifan, Zhang Weijing participated in the construction of nearly ten chip factories, becoming recognized in the industry as a “master of building factories.” I worked in Deyi for 20 years. During this period, he has been involved in the construction and management of more than 10 wafer factories in the United States, Japan, Singapore, Italy and other places, the construction and management of IC enterprises has accumulated a wealth of practical experience.

According to Texas Instruments, Zhang, who has been in work for 20 years, can choose to retire early. In November 1997, he first participated in the renovation of the 6-inch fab in Huajing, Wuxi, and then because the investors’ bosses wanted them to set up a factory in Taiwan first, and then to the mainland.

Under this arrangement, Mr. Zhang returned to Taiwan in February 1998 to become general manager of World Semiconductor. By then, he had also put forward a clear plan to invest in the mainland, with the first and second plants in Taiwan, followed by the third plant to the tenth plant in mainland China to invest in construction.

Zhang is responsible for setting up chip manufacturing plants around the world, but he has not set up a factory in Chinese mainland. On his deathbed, the old father “When did you come to the mainland to build a factory?” “This question has been lingering in Zhang’s heart.

At that time, TSMC and United China Electronics (United) in the industry has formed a competitive scale advantage. Every four or five years, the semiconductor boom cycle, like the waves, the size of the enterprise is qualified to survive, which has even become the new entrants difficult to get rid of the spell.

In 2000, Zhang Weijing founded SMIC in Shanghai, as one of the sponsors of SMIC, Zhang Weijing employee identification card number is E000001.

At that time, a group of overseas Chinese experts engaged in integrated circuits to the national leadership of China’s development of integrated circuits, has been strongly supported. In 2000, several experts further arranged a mission to explore the city where the then more advanced first 8-inch plant should be built. Jiang Shangzhou was the permanent deputy director of the Shanghai Economic Commission, under Jiang’s strong invitation and efforts, finally SMIC settled in Shanghai Zhangjiang High-Tech Park.

At that time, the semiconductor plant in Shanghai, the cost advantage is unique. The cost of building a factory is 38% cheaper than Taiwan’s, and the price of land is one-tenth that of Taiwan. Because the building materials are cheap, so SMIC capping the plant when the steel bars are more than taiwan’s use of large, quality is better. Cement is cast in a 90-degree way, and 135 degrees. The latter is better, but labor costs, Zhang Weijing will choose the latter.

In addition to trying to save money, SMIC has to find money. In order to raise funds, Zhang Weijing thought of a lot of alternative sashimi. The most typical operating idea is to make up for the lack of capital investment by using the enthusiasm and investment of local government to develop the industry. For example, a joint venture with the local government in Chengdu is to set up a factory. In Wuhan, sMIC provides hosting services with financial support from local governments.

Through self-construction, local government co-management and joint ventures and other models, Zhang Weijing’s SMIC in Beijing, Tianjin, Shanghai, Wuhan, Chengdu, Shenzhen and other places to establish 6 8 inch fabs and 5 12-inch fabs, which is known in the industry as the “diamond layout.”

The semiconductor manufacturing industry has also been dubbed by the industry as an industry that will “fall” half way. The main reason is that this is an industry that requires huge capital investment and a large cluster of high-tech talent to work. Huge amounts of capital and the demand for a large number of high-level talents, forming a huge barrier to entry in this industry. Money and talent are indispensable.

In 2002, SMIC began mass production in search of potential customers.

At this time, as far away as Shenzhen’s Huawei, chip design has been a small achievement. Huawei executives have repeatedly expressed a desire to work with mainland foundry, but with mainland manufacturers just getting started, the technology level and quality of the stream are low. In order to ensure quality, Huawei had to look abroad at the time.

In Zhang’s planning, SMIC’s development is to learn from the advanced western experience and follow the model of market economy. As a result, In204, CSCO was listed in New York and Hong Kong as a public company.

SMIC ushered in the high-gloss moment of this year, Hess Semiconductor has just been established, Huawei self-research chip business from then on into the deep water area.

Deep water has always been ferocious. After Heath’s founding, he lost seven years in a row in the midst of burning money and struggling.

Huawei has long suffered from being stuck in the neck of the United States, so it is under pressure to be firmly committed. Otherwise, HiSilicon will like some old chip companies, give up, take their lives, submerged in the flood of the times.

TSMC has not yet enjoyed the high-gloss moment, it ushered in a wave. From 2003 to 2006, TSMC sued SMIC twice, directed at Mr. Zhang’s Taiwan team, alleging that they had infringed TSMC’s technology patents and stolen trade secrets.

The fact that some of the engineers made mistakes made things irretrievable. Mr. Zhang tried to save the situation by filing a counterclaim against TSMC in Beijing High Court and California, but to no avail.

In the early winter of 2009, a US court ruled against SMIC, and four days later, Mr. Zhang flew to Hong Kong with then-chairman Jiang Shangzhou to negotiate with TSMC.

But under pressure from TSMC, compromise is the only solution. The price was not heavy, with bad news coming in succession: Mr Zhang announced his resignation and left SMIC, which was founded, SMIC, which paid tsMC hundreds of millions of dollars in huge compensation, and TSMC, which took a stake in SMIC, holding about 10 per cent of the shares.

Reversing the tide

After Zhang’s bleak departure, Wang Ningguo began to govern SMIC.

Looking at Wang Ningguo’s resume, it is not difficult to find that this is a well-trained and experienced professional manager. With a Ph.D. in Materials Science and Engineering from the University of California, Berkeley, Wang has worked in the Lawrence Radiological Laboratory and Bell Labs, and is well-known in the semiconductor industry worldwide.

In the 1980s, Wang Ningguo entered the Applied Materials Company, leading a number of technology development, with more than 100 patents. In the early years, he first proposed a single chip assembly process technology, solvethe the problem of traditional fabs, the semiconductor industry into a more sophisticated field. Applied Materials called it the most successful product of all time and its inventor was Wang Ningguo. Prior to joining SMIC, Wang Ningguo was the chairman of Huahong NEC, a subsidiary of Huahong Group.

But Wang Ningguo took over a heavy burden of enterprises. The fourth-quarter 2009 results, released shortly after he joined SMIC, showed a significant loss of $482.3 million and a $299.7 million settlement with TSMC.

In addition, Wang Ningguo took office, had to first stabilize the team and stabilize customer relations. Soon after, Wang Ning guoguo an “organizational streamlining plan”, intended to promote the change of the company’s organizational structure, the core is to adjust the upper management team and the grass-roots team, the integration of technology, production, market and other resources, improve operational efficiency. Wang Ningguo took office at the beginning of the “three fires” although moderate, but basically stable military heart.

Mr. Wang has also personally hired a team of executives with international experience, including Yang Shining, the future chief operating officer who is opposed to him. In terms of cost control, he streamlined the organization, which led to the departure of several management staff, but controlled the company’s spending.

Wang Ningguo has set two strategies for SMIC: avoiding positive competition with TSMC, and striving to make SMIC the “best alternative” for customers, that is, to make TSMC spare tires, and to improve the local industrial chain to help local semiconductor equipment, materials and chip design companies develop.

Fortunately, the good news boosted the semiconductor industry as a whole in 2010, thanks to the global semiconductor industry boom, particularly the record growth in the foundry market. At the same time, the profit-oriented internal adjustment also began to work, with the company’s gross margin rising from 7.6% in the fourth quarter of 2009 to 23.9% in the fourth quarter of 2010.

At this point, however, the crisis is emerging. At an internal board meeting, the majority shareholder, Datang Telecom, proposed promoting Yang Shining as executive vice president, the highest-ranking position in the president. Until then, other members of the board were unaware of the proposal, the sources said. Within the company, rumors of a disharmony between COO Yang Shining and CEO Wang Ningguo began to emerge.

The battle for control

In March 2011, at the Kerry Center in Pudong, Jiang Shangzhou, then chairman of SMIC, held a large-scale press conference. In his speech, he revealed a grand plan: he wants to use three years or so to catch up with the world’s advanced manufacturing processes, and five years or so to bring SMIC’s annual sales revenue to $5 billion.

However, the company fell into a circle of interests vying for control after the sudden death of The River, which was so painstakingly funded and supported by SMIC.

SMIC’s shareholder background is complex. Sovereign funds and local governments, Hong Kong and Taiwan capital and foreign investors from Europe and the United States gathered together, the demands of all parties are not the same, the differences continue. To ensure the normal operation of the company, there must be a person running in the parties to maintain the balance of interests. Since Zhang’s departure, the role has been jiangzhou.

After the death of the main bone, SMIC was mired in infighting. Customers trust them greatly reduced, some of the major international customers even directly withdraw orders. In the advanced process process, the original plan to catch up was interrupted, but also hindered the company’s development and breakthrough in the field of technology.

In the end, the new chairman, Zhang Wenyi, recommended Qiu Ciyun, a former “old man” of SMIC, as SMIC’s CEO. Qiu was Mr. Zhang’s first deputy and served as senior vice president of sMIC international operations, knowing the company well. This return to fire, zhang Wenyi described as “the best person to lead the company towards a bright future.”

Faced with SMIC’s embarrassment, Qiu needs to get the company back on the right track as soon as possible.

Mr Qiu recognises that SMIC is far behind TSMC in terms of sales and investment, so he must acknowledge SMIC’s weakness. In his view, as a contract enterprise, it is important to be in the right market position. On the one hand, we need to realize our fundamental position in the information industry. On the other hand, we should also seek sustained development within the correct industrial chain.

His strategy, therefore, is to target international giants as they seek to differentiate and diversify their businesses without confronting competitors directly.

The reason is two, one is that with the continuous development of the process, the process is getting closer to the physical limit of semiconductors, coupled with capital investment and research and development costs are huge. One figure is that SMIC invested $3.5 billion to grow by $1 billion, but another giant achieved $2 billion in growth and invested as much as $20 billion, meaning Moore’s Law could also face a “ceiling”;

While seeking differentiation, SMIC has not dropped a more refined process of development. Under the leadership of Qiu Ciyun, SMIC successfully achieved 28 nm mass production in 2014. While there is still a two-generation technology gap with the industry’s cutting edge, the breakthrough also helped SMIC get its first mobile chip order, the SnapDragon 410 from Qualcomm.

The arrival of the technocrade

On May 10, 2017, Qiu Ciyun resigned from SMIC’s core management for personal reasons.

After Qiu’s departure, SMIC announced a big announcement: Liang Mengsong, head of research and development for The Front Office, and Samsung wafer foundry, will join SMIC as joint CEO and executive director.

Mr. Leung has a legendary experience, having worked at AMD, a processor factory in the United States, after graduation, and joined TSMC at the age of 40 (1992). What really made him famous in 2003 was TSMC’s “copper process” of 130 nanometers. In 2003, TSMC beat IBM with its own technology and made it famous worldwide with a 130nm “copper process”.

The importance of Liang Mengsong is also clearly stated in TSMC’s files, saying that he is “responsible for or participating in the advanced technology of tsMC’s craftsmanship for every generation.” Liang Mengsong is also the inventor of tsMC’s nearly 500 patents, far more than any other supervisor, and a member of the “New Process Equipment Selection Committee”.

It can be seen that Liang Mengsong for TSMC advanced technology master the breadth and depth, as well as from research and development to manufacturing integration of the familiarity, in the company less people can and can.

In 2017, SMIC threw an olive branch to the top talent, and his arrival has been described by the media as “China’s semiconductor industry has entered the Liang Mengsong era”.

The technical support of the top semiconductor talent scored a leap forward for SMIC, and in a year and a half, Liang Mengsong will be SMIC from the 28nm era to enter the 14nm era. What’s more, SMIC raised the yield of 14nm process chips from 3% to 95% in less than 300 days.

Huawei was in a state of crisis in early 2019, shortly after SMIC’s 14nm news.

In May, the U.S. added Huawei to its “entity list” to limit the supply of key components to Huawei by U.S. companies. Heath had its toughest moment, and all the spare tyres were announced overnight.

The hurricane is coming. At this point, Huawei realized that the path of chip contracting would be cut off sooner or later if the U.S. continued to crack down.

At this time, just broke through the 14nm process of SMIC, once again into Huawei’s line of sight. Although still not enough to meet Huawei’s high-end mobile phone chip 7nm process requirements, but the domestic chip foundry technology breakthrough for China is undoubtedly good news.

Orders for Huawei’s mobile phone chips are starting to tilt toward SMIC. Earlier this year, orders for the 14nm Kirin 710A, originally in TSMC’s hands, were also transferred to SMIC.

According tomedia reports, at the end of 2019, HiSilicon began to instruct some engineers to design chips for SMIC, and gradually shifted the production of the chip from TSMC to SMIC, and research and development resources began to tilt toward SMIC.

This is the ideal winter solution for Huawei, which has access to more advanced technology and a more stable source of orders. Up to 20 per cent of SMIC’s revenue comes from Hess Semiconductor, according to Bernstein Research, a market research firm.

The fate of the two companies is tied together.

There’s no way back.

To this day, the worst has happened.

Huawei is in the midst of an indefinite food shortage crisis after the U.S. slammed off Huawei’s partnership with TSMC’s chip foundry. The worst rumor is that some of Huawei’s chip stocks will only be maintained until early next year.

After TSMC cut off its supply, Huawei desperately needed a friendly army. SMIC is seen as one of the most critical saviours in the domestic reserve.

Chip manufacturing is a technology-intensive industry, the gap between the head manufacturers and the chasers. In contrast, TSMC has begun planning a 5nm production line, and SMIC has just achieved 14nm mass production.

Although 14nm chips still have a lot of demand in AI, IoT and other industries, but the international chip manufacturing leader has already entered the 7nm era. In the smartphone industry alone, the major manufacturers’ flagship and mid-range machines are also basically equipped with 7nm chips. If it stays at 14nano00, TheSea chips will soon lose their competitiveness.

But advanced process upgrades often come at the cost of huge costs. For each future generation, the cost will increase accordingly.

This has led to the creation of high financial and technical barriers, which are almost seamlessly accessible to later entrants. More and more players are choosing to get out.

Today, manufacturers with advanced processes of 7nano00andlow and below are left with only TSMC, Samsung and Intel, and the new U.S. rules have largely cut off the companies’ deals with Huawei.

In China, SMIC is the closest player to a 7nm process, but is still in a technology catch-up phase. To speed up the catch-up on the stage of power generation, the need for high-end semiconductor equipment and materials from upstream, the core is to provide lithography process, which is the largest domestic chip manufacturing industry short board.

Globally, EUV lithography is the most advanced lithography system available today and the only device capable of producing processes below 7 nm. EUV, or ultraviolet lithography technology, is mainly used in high-end chip manufacturing, and Asme (ASML) is the only manufacturer of lithography machines that can produce EUV.

In 2018, SMIC spent a full year of profits on ordering an extreme ULTRA-UV EUV lithography machine from ASML for the manufacture of the 7nm process. But the device has not yet arrived due to a series of restrictions in the United States. With a u.S. order, it remains a mystery when SMIC will be able to use the device.

Without a top-notch lithography machine, SMIC’s desire to rush into the 7nm process in a short period of time is not easy.

But SMIC also has its own “ammunition” reserve, a new process code-named “N1”, which SMIC says is 20 percent better than 14nm performance, 57 percent lower power consumption and a 63 percent reduction in logic area. Industry insiders generally believe that the “N1” target is TSMC’s 7nm, 5nm advanced process.

The key friend of Huawei has a new historical mission on his shoulders.