Semiconductor Wars in Second-tier Cities

If counties and prefectures are well governed, chips and semiconductors will be established.

In the business invitation map of local governments, chip industry chain companies are becoming the most popular guests, which were once lonely but are now the most welcomed. Against the backdrop of slowing economic growth, unsustainable land finance, and increasingly fierce economic competition among regions, large chip semiconductor projects are becoming the hot cakes that local governments are competing for.

From Wuhan, Hefei, Xi'an, to Chengdu, Chongqing, Xiamen, Zhangzhou and Quanzhou, these second-tier cities have successively realized the huge gold mine of the chip semiconductor industry. Although they do not have the technical R&D talents and strength that can match super cities such as Beijing and Shanghai, with the opportunity of chip semiconductor manufacturers to expand the market, these second-tier cities have used their land, policies and market dividends to start digging for gold in the chip semiconductor industry chain.

However, it is not easy for second-tier cities to negotiate "chips". The matching and implementation of fab factories, IDM factories and local governments is like an expensive love affair.

Wuhan XMC is the cornerstone and soul of this city's chip industry.

Wuhan Han River Night View

In 2006, Wuhan XMC started construction and in 2008, it started mass production. Its investment scale was over 10 billion yuan. It was the only 12-inch chip production plant in Central China at that time. When it was first established, Wuhan XMC was managed by SMIC because Hubei Province and Wuhan City had neither talent nor technology.

From 2006 to 2012, Wuhan Xinxin almost replicated all the twists and turns of Shanghai's domestic chip manufacturers, taking all the necessary steps and taking all the necessary risks.

First, XMC tried to enter the international market with DRAM as its main product, but the price of DRAM plummeted just after mass production, and the company had to switch to flash memory. In comparison, Shanghai Huahong was luckier than XMC in that Huahong caught the tail of the DRAM upswing cycle and achieved profitability in the year of mass production. However, XMC started construction in 2006 and started production in 2008, when the financial crisis and the semiconductor cycle took a sharp turn for the worse, and it did not even catch the tail of the cycle.

Afterwards, Wuhan Xinxin teamed up with the American flash memory manufacturer Spansion Semiconductor, trying to survive by transferring Spansion's technology to the existing market. Unfortunately, this last straw in 2008 was also difficult to save itself. At that time, Spansion was on the verge of bankruptcy and Wuhan Xinxin had almost no orders.

The overall semiconductor cycle is down, and both DRAM and flash memory are facing a tragic situation of both quantity and price being cut. Under the condition that foreign manufacturers are operating at a loss and cannot operate production lines at full capacity, domestic manufacturers will inevitably suffer. In a recession, only the landlords have surplus food. Wuhan Xinxin is crying out for food, but it has attracted the attention of vultures.

At that time, foreign acquirers such as TSMC and Micron all took a fancy to Wuhan Xinxin. At that time, Wuhan Xinxin had already been mass-produced and was managed by SMIC. The acquisition of Wuhan Xinxin by foreign chip manufacturers was equivalent to adding another mature production base in the Chinese market, and it could also curb competitors in the Chinese market, which was a two-birds-with-one-stone strategy.

Fortunately, in 2011, SMIC reached an agreement with the Wuhan Municipal Government to establish a joint venture.

Jiang Shangzhou, then chairman of SMIC, played a key role in this incident. Unfortunately, Jiang Shangzhou passed away in 2011, and SMIC was far from recovering from internal and external troubles. SMIC and Xinxin became brothers in distress, and it was not easy for either company to stay together.

Later, by chance, in August 2012, Yang Shining, the former COO of SMIC, joined Wuhan Xinxin. In 2013, Wuhan Xinxin became independent from SMIC. After winning a large order for memory chips from GigaDevice and working with Phison to push the NAND memory process from 55nm to 32nm, Wuhan Xinxin won the favor of the big fund and won the competition for the national memory base. In 2016, the national memory base was launched in Wuhan, and Yangtze Memory was established on the basis of Wuhan Xinxin.

In the long history of Wuhan Xinxin's development, the company achieved a turnaround only after government subsidies met the right business leaders and excellent timing. The instant transformation from an ugly duckling to a white swan cannot avoid the helplessness of Wuhan Xinxin in its early operations.

Local governments cooperate with foreign capital to introduce technology, but the key problem is that the semiconductor cycle is volatile and does not change with people's subjective will. Once they lose strong support, where can local governments find the real money to invest? Therefore, the smart approach is to balance the domestic and foreign markets and walk on two legs.

A senior person in the chip industry said that the success of Wuhan Xinxin was due to the ruthlessness of the people of Hubei and the coincidence of timing. If it were in other places, such a project might have failed.

As a local government chip project, Wuhan Xinxin's early difficulties were that it lacked, or had no ability and resources, to be classified into a national chip market segment. At that time, there was probably no market segment that the government was capable of leading. At the same time, local governments have inherent shortcomings in fully market-oriented operations, so Wuhan Xinxin's six or seven years of painful period was inevitable.

Such expensive tuition fees are really painful to pay. However, Hefei, also located in the central region, has found an extremely ingenious and hidden shortcut. While Wuhan is still paying tuition fees, Hefei has already made a lot of money in the high-tech industry, and has spent a lot of money in the storage industry in recent years.

Why is Hefei so confident?

At the end of 2017, before BOE's 10.5-generation production line in Hefei rolled off the production line, the Hefei News Broadcast produced three consecutive episodes of the news broadcast titled "There is a BOE in Hefei".

In an interview with the media, Zhang Yu, vice president of BOE Technology Group and general manager of Hefei region, said: Hefei is a blessed land for BOE.

Ten years ago, China was "lacking screens and chips", but today, the phrase has become "lacking chips and soul". While semiconductor integrated circuits (commonly known as chips) are still struggling, the semiconductor display industry has emerged as a new force.

In 1992, when the young factory manager Wang Dongsheng took office at the Beijing Electron Tube Factory, he faced a difficult mess, just like many electron tube factories in China at that time. How could a high-tech enterprise with full marketization and international competitiveness be established on the site of an old and backward electron tube factory?

In retrospect, those national enterprises that later became legends and supported China's semiconductor industry all chose the path of technological independence.

On the remains of countless state-owned electron tube factories, world-class enterprises such as Jiangyin Changdian and Beijing BOE have grown up. Although both companies were originally "electron tube" factories, under the keen market insight and drastic reforms of strong corporate leaders, they both chose to invest in the most difficult and most promising cutting-edge technologies at the bottom of the industry cycle.

From the establishment of BOE in 1993, its listing on the A-share market in 2000, its acquisition of South Korea's Hyundai's panel business in 2003, to its two consecutive years of ST status due to heavy losses from the screen cycle, BOE's performance and stock price have been as thrilling as a roller coaster. However, none of these can stop local governments from flocking to BOE's screen industry.

BOE started building a 5th generation line in Beijing, which generated billions of yuan in tax revenue each year. The Beijing municipal government also made a lot of money from the rising stock prices due to the booming screen cycle. This shows the multiplier effect of high-end manufacturing.

Then in 2008, the Hefei Municipal Government won the competition with the Shenzhen Municipal Government and recruited BOE's 6th generation production line to Hefei, with a total investment of 17.5 billion yuan. However, the total fiscal funds used for development in Hefei in 2007 were only about 3 billion yuan. In addition to the 8.5 billion yuan bank loan, who would fill the remaining 6 billion yuan capital gap?

The government's credit endorsement laid the foundation for market-based financing. After seeing the government's money, institutional investors and retail investors in A-shares flocked in.

This operation has been done more than once and has been proven to be successful. In the news broadcast of "There is a BOE in Hefei", the official media of Hefei said that the operation of the Hefei Municipal Government has been included in the textbooks of business schools. Whenever BOE has a new production line to invest in, the Hefei Municipal Government platform and BOE's A-share listing platform will join forces to open up a bright road for BOE's financing.

Hefei Municipal Government and BOE are a match made in heaven.

The benefits that BOE has brought to the Hefei municipal government are obvious: first, in terms of reputation, the commissioning of the 6th generation line in 2010 ended the situation in which China's large-size LCD panels were completely dependent on imports; second, in terms of affordability, BOE has invested not only in production lines in Hefei, but also in projects such as smart manufacturing factories and digital hospitals. As of the end of 2017, BOE's investment in Hefei has exceeded 100 billion. Even with a 10-fold multiplier effect, the total GDP that BOE can drive has exceeded Hefei's total GDP in one year, and this is not calculated based on the 100-fold multiplier effect of the semiconductor industry.

It is no exaggeration to say that BOE supports the entire economy of Hefei. And, most importantly, the high-tech ecology of Hefei has become "chickens lay eggs, eggs lay chickens, and both chickens and eggs stay in Hefei and don't want to leave."

In the news broadcast "There is a BOE in Hefei", Ren Haidong, the factory director of Hefei Sanlipu Optoelectronics Technology, a supporting supplier of BOE, spoke out and said that the company had hesitated between choosing Wuhan or Hefei as its location, but finally decided on Hefei. Now Sanlipu Optoelectronics can achieve zero inventory when supplying BOE.

This shows that Hefei and Wuhan are determined to compete for supremacy.

Wang Kai (pseudonym) has experience in chip entrepreneurship and is currently the head of the chip semiconductor field of a well-known domestic investment institution. Wang Kai told CV Intelligence that if we only count the BOE shares held by the Hefei Municipal Government platform, the floating profit at its peak was over 10 billion.

Hefei has made money from BOE and developed a successful model. Now it is investing heavily in the chip industry to make up for the shortcomings in the industry. In addition, Hefei's development of the chip industry is "the same desire from top to bottom", and it is highly recognized at the national (big fund, Ministry of Industry and Information Technology, etc.), provincial, municipal and district levels. Long-term experience has given the grassroots practical experience, which is an incomparable advantage of Hefei compared to other places.

In the chip industry layout that Hefei City started around 2017, two projects are most eye-catching.

The first is Jinghe Integrated Circuit, a cooperation between Hefei Construction Investment and Taiwan-based wafer foundry Powerchip, which aims to solve the supply problem of supporting panel driver (LCD driver) chips for BOE. In fact, it is still within BOE's "egg" ecosystem; the second is Hefei Changxin, which cooperates with GigaDevice. As a "provincial team", Hefei Changxin has been in the limelight recently, but for the Hefei Municipal Government, Hefei Changxin's memory chip business is a brand-new attempt. Independent of BOE's "egg" ecosystem, it remains to be seen whether Hefei Changxin can carve out a path.

According to Wang Kai's observation, since 2017, Hefei's first-level investment promotion bureau staff have been frequently seen in Shanghai, Shenzhen and other places. Hefei is competing with cities such as Nanjing for projects in the upstream and downstream of the chip industry chain, and taking over the industrial spillover from first-tier chip centers. Similar to the strategies of many cities in Jiangsu Province, Hefei is also taking the path of giving equal weight to both domestic and foreign chip companies, and is actively introducing both domestic and foreign chip companies.

This year, Wang Kai felt a significant change: Hefei's investment promotion staff and industry insiders emphasized that Hefei no longer belongs to the central region, but is part of the Yangtze River Delta "three provinces and one city" (Anhui, Jiangsu, Zhejiang, and Shanghai). From the perspective of the district-level investment promotion bureau, Wang Kai believes that the styles of Anhui, Jiangsu, Zhejiang, and Shanghai are very similar.

"If you have dealings with the investment promotion bureaus of various cities in Jiangsu, you will find that the directors of their investment promotion bureaus are very humble. They are all sales-oriented talents and will recruit chip companies whenever they meet them. Anhui also has this style, which is obviously very different from the investment promotion officials in the central and western regions," said Wang Kai.

Looking at the development history of the chip semiconductor industry chain in Hefei, Hefei’s pioneering efforts are indeed “daring to be the first in the world.”

In the current industrial policy and financial system, the Hefei municipal government made the most of all resources to give unprecedented leverage to BOE, the core enterprise in the region. The success of Hefei's big gamble was inseparable from the judgment and determination of local officials. Of course, it was also inseparable from BOE's high R&D investment for decades.

From this point of view, Hefei Municipal Government and BOE are in the same boat. Such an extremely successful case requires close cooperation among the government, industry, finance, academia and research. If there is a disconnect in the middle, it will be a blow to the high-tech industry. The formation of such a strict operation system of the science and technology industry depends on "the right time, the right place and the right people" and "the same desire from top to bottom".

In 2012, Samsung, a giant in the chip field, launched its storage project in Xi'an. As a purely foreign-funded project, Samsung's operation in Xi'an was very successful.

Wang Kai told CV Intelligence that Samsung's landing in Xi'an was the result of a coordinated plan at the national level, and the final location was chosen. This project has had a very obvious impact on the local social economy. Once the Koreans arrived, Korean-opened restaurants, clubs, and schools all sprang up. As for the cascading effects such as land appreciation, taxation, and employment, it goes without saying.

"But for purely foreign-funded factories, we can only calculate the economic accounts, not the technical accounts," said Wang Kai.

How to calculate the technology account? Because China owes too much in chip semiconductors, and also has to consider the source of patents, many local governments have chosen to cooperate with foreign manufacturers, providing money and land in exchange for technology transfer from foreign factories.

But the process of technology transfer is obviously full of twists and turns. The GlobalFoundries project in Chengdu was previously unsuccessful in Chongqing and also ended in failure in Chengdu. According to Jiwei.com, no company dared to take over the built factory because of engineering quality issues; the Fujian Jinhua project, a cooperation between Fujian Province and UMC, was suspended due to a patent dispute with Micron.

Wang Kai believes that local government heads will play a vital role in the development of the region's chip semiconductor industry. As long as the officials in charge are from a semiconductor background, they will never miss the opportunity to revitalize the local economy through the chip industry, because they are too aware of the huge driving effect of this industry.

However, because the chip semiconductor industry is so difficult to understand, the threshold is very high, and the circle is relatively small, after the officials in charge leave, there is often a question mark as to whether the chip projects that have been launched can continue to operate. If they fail, the money is basically wasted.

At the same time, Wang Kai told CV Intelligence that the initiative of district (county) level officials, especially investment promotion officials, is crucial to the development of the regional chip industry ecosystem. Officials at the provincial and municipal levels are transferred here and there, but once a project is settled, they will not leave. The subsequent development depends on how the district and county level officials operate and build this industrial ecosystem.

In Xi'an, Chengdu-Chongqing, Wuhan, Jinjiang and other places, the initiative of district and county officials is quite different from that in the Yangtze River Delta.

Wang Kai believes that the above-mentioned places all have the problem of insufficient enthusiasm for investment promotion in developing chip semiconductor industries. Samsung projects have been located in Xi'an for many years, and apart from the upstream and downstream brought by Samsung itself, the district investment promotion bureau has brought relatively few projects. In contrast, Nanjing has attracted TSMC, and its officials are more active. They go to Shanghai every day to recruit chip companies and enrich the upstream and downstream of chips around TSMC.

Wuhan, Chengdu and Xi'an are in similar situations. Xinxin has been developing in Wuhan for nearly a decade, but its driving effect is limited. Now with Yangtze Memory, the situation may change after mass production. The business environment in Chengdu is similar to that in Wuhan. GlobalFoundries was also a large project, but it failed.

As for Fujian Jinhua, it may be that Fujian Province was previously dominated by industries such as real estate and shoemaking, and suddenly the chip semiconductor industry came along, so local officials needed a longer time to learn.

As local grassroots officials’ learning ability, willingness and experience in high-tech industries are not up to par, any changes in external factors of the project may cause major shocks and catch local investment plans off guard.

Take the GlobalFoundries project as an example. The project focuses on low-power FD-SOI specialty technology. The domestic market is not enough to support this project. At the same time, the FD-SOI industry cluster is concentrated in Shanghai. Coupled with the high-level changes at GlobalFoundries in early 2018, the GlobalFoundries Chengdu project was substantially aborted.

All the factors mentioned above cannot be reversed overnight by local governments, but the engineering quality issues of wafer foundries are at a relatively elementary level. If local governments fail to do their basic homework, they may end up with a pile of abandoned scrap metal after spending money, not to mention the multiplier effect of the chip industry.

On the premise of doing the basic homework, partners also need to make prudent judgments.

Lin Zonghui, a senior media person in Taiwan and founder of TechEdge, told CV Intelligence that some chip semiconductor companies in Taiwan have recently cooperated with many local governments. Because they may have certain types of technologies, in order to enter the mainland market, they can transfer some technologies that are in demand in the mainland after review by the Taiwanese authorities.

However, it should be noted that some of these companies may have a bad reputation in the past, the source of technology may be unclear, or the technology level may be exaggerated, although it is in line with the needs of local governments. However, the management of these Taiwanese companies is not pursuing long-term development, but tends to "kill the chicken to get the eggs". In the case that the source of some technology patents is unreliable, partners need to carefully investigate.

Only those with a wealthy family and outstanding people can make chips.

Whether it is Wuhan, Hefei, or Xi'an and Chengdu in the west, they are all provincial capitals built with the efforts of the whole province. In the PK of the chip war, only those cities with money, people, determination, strategy and wisdom can become the best in the competition among second-tier cities.

Local governments that realized the importance of high-tech industries such as chips and semiconductors early on took the lead in the initial industrial competition.

As major core projects are implemented one after another, the upstream and downstream of the city's science and technology industries are gradually taking shape. The resulting agglomeration effect helps the city further consolidate its existing advantages. At the same time, the positive externalities brought about by industrial agglomeration (including taxes, employment, talents, housing prices, etc.) released through the multiplier effect will strengthen the positive feedback of local government policies and ultimately enable the local government's industrial policies to be continued.

From this perspective, a large chip project, or even a core technology company, is complementary to the growth of a city. In the chip semiconductor industry chain, the core fab and IDM plants are crucial. As long as a city can build a successful fab and IDM plant, the entire industry chain can be fully driven.

The failure of a large chip semiconductor project is a huge blow to a city, because it means that the city has lost the ability to build an ecosystem. Take Chengdu as an example. Although the city has accumulated some strength in areas such as power chips and small and medium-sized chip companies are still okay, the successive setbacks of SMIC Chengdu and GlobalFoundries have caused the city to miss two excellent development opportunities.

After more than a decade of hard work in the technology industry, Wuhan and Hefei, two central cities, have begun to rise to the top of technology. Some cities have rushed to start technology without learning from the historical lessons of technology introduction. Obviously, some have paid a lot of tuition fees. For the chip semiconductor industry, we might as well be more patient and less impatient. After all, money can't buy technology.

The strength and weakness pattern of the chip semiconductor industry in second-tier cities reflects a problem: local governments face a very steep learning curve when they develop chip semiconductor projects themselves.

If it is a relatively strong foreign manufacturer with a strong driving force, such as Samsung, TSMC, etc., the local government needs to learn less, because foreign manufacturers do not give you technology, they solve all problems by themselves.

However, if the technology is introduced in cooperation with foreign-funded factories, it will be very difficult. Due to the interaction between the roles of enterprises and the government, local governments may need to pay a very high learning cost. Because of the two roles at the local level, local governments essentially have to go through two learning curves of trial and error. One is how to formulate industrial policies, serve the industry well, and form an industrial system; the other is how to run a market-oriented enterprise, and the latter is often difficult to achieve.

The road to semiconductors is long and difficult, and second-tier companies need to think carefully before pursuing "cores".

References:

1. Xinshi, Xie Zhifeng and Chen Daming;

2. Hefei News Broadcast;

3. GlobalFoundries’ transformation and pain in Chengdu, Jiwei.com;

4. Hefei Emerging Industry Capital Bureau: State-owned assets strongly support the implementation of 100 billion yuan investment, participate in the fixed increase and make a profit of 10 billion yuan, China Business News