Sober thinking under the chip manufacturing boom

The mobile Internet era that emerged with the smartphone revolution has injected tremendous vitality into my country's economic and social development, and has also led to a continuous increase in the country's attention to industries such as the Internet, semiconductors, and integrated circuits (chips). Integrated circuits are an important part of the semiconductor industry and are also the focus of most attention in the development of my country's semiconductor industry.

[[350211]]

Since the issuance of the "National Integrated Circuit Industry Development Promotion Outline" in 2014, the government's support policy for the integrated circuit industry has been elevated to the national strategic level, and a national industrial investment fund has been established. Especially after 2017-2018, my country began to promote the development of the integrated circuit industry with a determination to burn its boats, and governments at all levels have introduced support policies, and a wave of chip manufacturing has been set off across the country.

The chip-making dreamers who come one after another

In this wave of chip manufacturing, only a handful of companies have successfully emerged, which is insignificant compared to the huge base of chip companies. It can be said that "one general becomes famous and thousands of bones are broken." Among those companies that were eliminated, some even caused serious losses of tens of billions. The first was Wuhan Hongxin, which mortgaged its 7nm lithography machine.

In November 2017, the Hongxin Semiconductor Project was established in Wuhan, claiming to have independent research and development technology for 14nm process and to conquer 7nm. It also has the only 7nm lithography machine in mainland China and invited the former CTO of TSMC, Chiang Shangyi, to serve as CEO. However, three years later, the project was exposed to be unfinished and Hongxin was sued by multiple subcontracting companies.

According to relevant information disclosure, the total investment of Wuhan Hongxin project is 128 billion yuan, of which 15.3 billion yuan has been invested by the end of 2019, and 8.7 billion yuan is planned to be invested in 2020. Now the project is unfinished, the 7nm lithography machine has been mortgaged, and the economic losses caused are as high as tens of billions of yuan.

The second is Chengdu GlobalFoundries, which has joined hands with GlobalFoundries.

Chengdu GlobalFoundries was launched in May 2017, jointly established by GlobalFoundries, the world's third largest wafer foundry, and the Chengdu Municipal Government, with a planned investment of US$9.053 billion. At that time, it was known as GlobalFoundries' largest and most technologically advanced production base in the world.

The Chengdu government invested 7 billion yuan in GlobalFoundries' factory construction, responsible for the construction of factory buildings and supporting facilities, as well as the establishment of R&D, operations, and logistics teams. However, less than two years after the factory was built, GlobalFoundries stopped operations and issued a notice of suspension of work and business in May 2019, causing economic losses of nearly 10 billion yuan.

There is also Nanjing Decoma, which wants to be an IDM wafer factory.

Tacoma (Nanjing) Semiconductor was established in November 2015 with a total investment of approximately US$2.5 billion. It plans to produce power management chips, micro-electromechanical system chips, etc. In addition, the project will also build an 8-inch/12-inch wafer manufacturing plant, a packaging and testing plant, an equipment remanufacturing plant, a scientific research and design center, an IC design company and supporting living areas.

Decoma was once highly anticipated, but due to its lack of financing capabilities, it eventually failed and was announced as a dishonest debtor by the local government of Nanjing on November 5, 2019.

These chip-making projects established in different regions were all launched with great fanfare at the beginning, but all of them fell into unfinished projects, suspension of production, and labor disputes within just a few years, causing serious economic losses and shattering the chip-making dreams of many participants.

These companies initially used the banner of "domestic chips" and hyped up "chip autonomy", then hastily launched their projects, only to return home in a flash with failure, leaving behind a mess. What is disturbing is that among all the chip companies, it is likely that there are quite a few who are eager for quick success and instant benefits in the "chip manufacturing movement".

Chip manufacturing tests the long-term endurance of chip companies

In the current intensifying chip manufacturing craze, although we have seen a number of companies with initial scale grow up, such as SMIC, Yangtze Memory Technologies, and Cambrian, most of them are bubbles that are born and die in the driving effect of "policy subsidies" and the cry of "domestic substitution".

Perhaps this experience of "panning for gold in the sand" is also a "compulsory course" that cannot be avoided for industrial development, but the "tuition fees" currently paid by some regions and companies are really too high.

With the support of policies and capital, the "chip manufacturing fever" sweeping the country is still being pushed to new climaxes. According to Qichacha data, there are currently 46,300 chip-related companies in my country, with 5,800 new related companies added in 2019 and 12,800 new companies in the first three quarters of this year, of which 6,200 were newly registered in the third quarter, a year-on-year increase of 288.4% and a month-on-month increase of 34.8%.

However, integrated circuits are a capital and technology intensive industry, and only a few companies can stand out. In today's domestic chip-making movement, only a few participants will be able to achieve success, and the efforts of most regions and companies are destined to fail to reap the expected rewards.

After all, looking back at the history of the industry over the past few decades, only a few giants such as TSMC, Intel, and Micron have survived to this day. The development of domestic integrated circuits is also inseparable from the objective laws of the industry.

Chip autonomy requires more centralized layout

The development of the domestic integrated circuit industry faces a very complex internal and external environment. From the perspective of the external environment, it is necessary to accelerate the realization of chip independence; and from the perspective of the internal environment, all regions want to make achievements in the emerging semiconductor industry. In such an environment, the emergence of "chip manufacturing fever" is inevitable.

However, in fact, many losses can be avoided. The integrated circuit industry itself has a very high threshold for employment in terms of technology and talent. After getting rid of the "restless" and "lucky" mentality, it is not difficult to realize that in order to truly accelerate the development of integrated circuits, it is better to concentrate resources on developing a few companies instead of allowing industrial parks to spring up everywhere. The country has now stated that it will further strengthen planning and layout.

In addition, in the integrated circuit industry, my country only has obvious shortcomings in a few areas, and is not lagging behind the international advanced level in all aspects.

In the field of IC design, in addition to Huawei HiSilicon, companies such as Unisoc and ZTE also have the ability to design advanced process chips below 7nm; in the field of packaging and testing, China's Changdian Technology is one of the top three packaging and testing leaders in the world.

The IC manufacturing field is considered to be the weak link of my country's integrated circuit industry, and the more basic fields of semiconductor equipment and semiconductor materials also have a significant gap compared with the international advanced level.

To be more specific, the advanced process technology of my country's wafer manufacturing is the biggest shortcoming, and the key difficulty is that domestic lithography machines are far behind ASML. ASML's mass-produced EUV lithography machines can be used for wafer manufacturing below 5nm, and it is rumored that the next-generation EUV lithography machines for below 3nm will be launched in 2021.

If we want to accelerate the development of my country's integrated circuit industry and achieve chip independence, expanding the scale is certainly important, but the more critical thing is to solve difficult problems such as lithography machines as soon as possible. Only by solving difficult problems and filling in the gaps can chip independence be truly realized.

In short, compared with thousands of companies flocking to develop aimlessly, the integrated circuit industry needs to focus on developing a few key companies to overcome difficulties, make up for shortcomings, and then drive the rapid development of the entire industry.