10,000 chip companies "died" in 2023 

Xinchao IC

2024-04-16 09:10Published on the official account of "Xinchao IC" under Beijing Core Watch

"Core Affairs" is Tencent Technology's semiconductor industry research and planning, and this issue focuses on the crisis and opportunity of domestic chip entrepreneurship.

Author丨Xinchao IC Wang Yike

Editor丨Tencent Technology Su Yang, Xinchao IC A Niu

Chunjiang plumbing duck prophet, headhunters can best experience the cold of the semiconductor industry.

Zhao Cheng, a senior headhunter in the semiconductor industry, said that she was looking for a new career direction, "I disappeared with the chip company and have left the chip industry this year." ”

Zhao Cheng's circle of friends has not updated the chip industry trends for more than half a year, and the latest one is psychological counseling-related content. Her departure can partially reflect the real situation of the chip industry.

In the past year, many chip companies have been full of negative news, layoffs and disbandments, and many bankrupt and liquidated companies have quietly disappeared in no one's corner - Qichacha data shows that in 2023, 10,900 chip-related enterprises in China have been cancelled and revoked, which means that nearly 30 chip companies disappear every day on average. However, the other half of the sea is flames, on the other hand, there will be 65,700 newly registered chip-related companies in 2023, an increase of 9.5% year-on-year.

Old companies are leaving, new projects are coming, and the "metabolism" of this industry has never stopped.

01 Nearly 30 chip companies disappear every day

Layoffs, bankruptcies, liquidations and financings, listings, unicorns and other words are in stark contrast.

In 2023, there are 3,243 chip design companies in China (many of which benefit in part from incentives from local governments), more than half of which have annual revenues of less than 10 million yuan, according to Wei Shaojun, head of integrated circuit design at the China Semiconductor Industry Association and a professor at Tsinghua University. If these companies do not have the ability to make blood, I am afraid it will be difficult to survive this cold winter.

In the past few years, start-up chip companies, driven by capital and policies, have registered companies and formed teams in groups, and after several rounds of financing, they have become phenomenal unicorn companies.

"Those returnee doctors who draw a big pie, plan to go public within 5 years, and do not have a working background in international chip manufacturers, have become the objects of contention among investors. Zhong Lin, the founder of Sanwu Micro, recalled the phenomenon of "being born together" of domestic chip companies.

Zhong Lin is an entrepreneur in the field of RF chips, and has repeatedly issued industry risk reminders on his personal public account, "Many investors do not understand chips, and they borrow Internet investment logic and semiconductor entrepreneurship portraits to choose investment projects." ”

The rashness of some investors and the inertia of being keen on "storytelling" have laid the groundwork for the reshuffle of China's chip industry.

If there is still a threshold for investors to pull returnees to enter the bureau, then Xin Microelectronics, which is engaged in power semiconductor chips and devices, directly pulls the industry threshold to the lowest - according to its prospectus at that time, the company has 793 employees, and only 16 undergraduates, accounting for 2.02% of the total number of employees. In March 2023, Xin Microelectronics voluntarily withdrew its listing application and terminated its GEM IPO.

Core Microelectronics is not an exception, some chip startups themselves are in a state of "a bottle is less than half a bottle swaying", and the technical strength is not solid, but the PPT roadshow ability is very strong, and it can raise a lot of money during the industry dividend period, but unfortunately the technology has not precipitated.

There are also many chip companies that are investor's projects, looking for some returnees or experts and professors to start a business, but most of these people are not front-line engineers and do not know the way to make chips. When the company is established, it uses the financing money to poach people in large quantities, and then redoes the things of the big factories and puts them into the market, and finally comes out with a large number of homogeneous products, but it is difficult to achieve technological innovation.

In the upward cycle, the heat of the market covers up the negative effects of this brutal expansion, and when the market is cold, many companies cannot make profits independently, and the valuation is too high and it is difficult to promote financing, and finally the capital chain is broken, heading towards bankruptcy and liquidation.

Compared to the above extreme cases, the biggest dilemma for most companies today is a lack of money.

"Many domestic chip companies are behind closed doors, spending a lot of money to make a bunch of domestic alternative chips and having to sell them, and they will sell them at low prices if they are not competitive. The more volatile the price, the less money the distributor can make and will not act as an agent and promote, and eventually a vicious circle. Zhong Lin told Xinchao IC.

In 2023, OPPO will disband the Zheku team and give up core manufacturing, like the first domino to fall, quietly accelerating the speed of the entire chip market "clearing". According to incomplete statistics, more than 50 companies in the industry have laid off employees in 2023, and so far in the third quarter alone, the number of companies that have laid off employees has exceeded 20.

Chen Ming, who has been in the industry for 5 years, is the beneficiary of various high-paid poachers. Around 2020, he jumped from a chip company in Shanghai to a chip company in Hangzhou, and his salary doubled. But now, his communication with insiders is no longer where to change jobs, but who has laid off the proportion of employees, and whose family has broken the news of layoffs.

He said, "The company is relatively stable this year, but I heard from internal sources that we may cut salaries after the Spring Festival, and there are plans to lay off employees, but I don't know how much the proportion is." ”

Cheng Bing, a veteran of the semiconductor industry, said, "Industry bubbles and business closures are not entirely a bad thing."

He emphasized that innovation requires a certain price, but bankrupt companies can help the chip industry to try and make mistakes from the dimensions of business model, product application, and market, which is also valuable for industry exploration.

"In the past few years, the chip industry has received policy support and capital attention, driving resources to gather in the chip industry, and now the cycle changes in the industry can help high-quality enterprises stand out and cultivate high-quality talents. ”

Now, the "metabolism" continues. Only projects with money can live, and those who can't make money and can't raise money can only go bankrupt and go bankrupt.

02 It is difficult for people who make quick money to survive

There are three key nodes in China's chip entrepreneurship: the release of Document No. 18, the establishment of a large fund, and the establishment of the Science and Technology Innovation Board.

In 2000, the "Several Policies for Encouraging the Development of the Software Industry and the Integrated Circuit Industry" (referred to as Document No. 18) was promulgated, and a separate chapter was set aside to carry out the "Integrated Circuit Industry Policy", one of which reads: domestic and foreign enterprises are encouraged to set up joint ventures and wholly-owned integrated circuit production enterprises in China.

In addition, Document No. 18 focuses on the integrated circuit industry in tax refunds: "For general VAT taxpayers selling their self-produced integrated circuit products (including monocrystalline silicon wafers), VAT will be levied at a statutory rate of 17% before 2010, and the part of the actual tax burden exceeding 6% will be refunded immediately, which will be used by enterprises for research and development of new integrated circuits and expansion of reproduction." ”

While driving foreign-funded enterprises to build factories in China, it has also attracted a wave of overseas chip technology experts to return to China to start businesses, and companies such as Vimicro, Spreadtrum, and GigaDevice were founded in this wave.

In 1999, Deng Zhonghan founded Vimicro Electronics Co., Ltd. in a warehouse of more than 100 square meters, and successfully developed China's first million-gate ultra-large-scale digital image processing chip "Xingguang No. 1", and became the first chip design company listed on the NASDAQ in the United States.

In 2005, Zhu Yiming, who returned from overseas, returned to China with a patent called "ultra-high-speed static random storage technology" and founded "Core Technology Jiayi", the predecessor of GigaDevice. Founded decades ago, GigaDevice has become the largest NAND FLASH local design company in Chinese mainland for several consecutive years. At the beginning of 2020, GigaDevice's market value once exceeded 120 billion.

Zhong Lin, who is also an entrepreneur, sighed, "Some of these people who returned to China to start a business decided to return to China to start a business out of feelings." They have entrepreneurial feelings, as well as love for the industry and the motherland, coupled with China's entrepreneurial environment can help them realize their dreams, leaving a good foundation for China's chip industry. ”

In 2014, the National Integrated Circuit Industry Investment Fund (hereinafter referred to as the "Big Fund") was established, guiding social capital to begin to focus on the semiconductor field, and domestic chip entrepreneurship set off another wave of climax.

In 2019, the emergence of the Science and Technology Innovation Board completely activated the enthusiasm for chip entrepreneurship. The tide of chip entrepreneurship is blooming everywhere, not limited to chip design, but also beginning to turn to the upstream of chips, equipment, materials, GPU, EDA, autonomous driving chips and other fields.

During this period, chip companies entered a blowout growth. In 2020, the number of registered chip companies in China reached 21,700, a year-on-year increase of 216%, and in the first half of 2022, there were as many as 30,800 new chip-related enterprises.

Stepping on the tuyere of the times, every wave creates a group of heroes, and it will also leave a "myth of making wealth", and it is inevitable that there will be people who fish in troubled waters.

Cheng Bing shared a case of relying on "chips" to get free of wealth - a college student who entered the sales department of a semiconductor company after graduation, and after accumulating a certain number of contacts, founded a small chip design company when the chip was in the wind, but the company did not have full-time chip designers, and the core technology came from the "outsourcing team". Relying on the design drawings brought by a few part-time engineers, he successfully earned a net worth of hundreds of millions.

"He has become a rich man, and I am still working hard to develop my own technology, and I will have unwilling and depressed emotions in my heart. But Cheng Bing knows that the chip industry needs people who really understand technology and the market, and this kind of company will not last long after all. "Lately, it's been rare to hear from this company," he said. ”

People who make a living from crystal balls are destined to eat glass that shatters on the ground. Taking advantage of the high popularity of chips, a group of people can make a wave of quick money, but they do not have the ability to survive the winter. Leaving the market together, there are also a number of companies whose purpose is to "cheat subsidies" and "cheat financing".

In 2018, Yan Yan, the founding managing partner of SAIF Asia Investment Fund, once said that we are lagging behind others in chip technology. In terms of chip manufacturing, only a few companies are doing it because it requires a lot of investment. However, there are many companies that do chip design, and some of them are purely indiscriminate. There are not many chip design companies that can really run, and even fewer companies can produce good smart chips in the end.

According to the statistics of the China Semiconductor Industry Association, among the 108 chip design listed companies in 2023, 66 will be profitable companies and 42 loss-making companies will report semi-annual reports, with a loss rate of 38.9%, and the gross profit margin and net profit of profitable enterprises will decline significantly.

"The market influence and channel ability of listed chip design companies will be better than those of non-listed companies, and the price advantage of the supply chain is more obvious, but they still suffer serious losses," Zhong Lin said.

In 2023, the number of domestic chip design companies will be 3,451, and most chip design companies will focus on low-end chip products and price wars, rather than investing time and energy in research and development. Over time, the chip design industry will fall into homogeneous competition, listed companies are still struggling with the loss line, and unlisted startups may not be doing as well as they want.

Ye Maochang, who is also a chip entrepreneur, summed up the "rout" in the field of chip design, saying that in a large number of financing projects, the agreements signed between the chip startup team and investment institutions have repurchase and VAM clauses, "(Next) there will be a lot of stories that are poor." Investment institutions themselves are also involuntarily affected, and they are also under great pressure from LP redemptions, especially since many funds are funded by government funds. ”

The reshuffle of the chip industry continues, and wealth creation is only an added value of chip entrepreneurship, and the project is gone, and the story of wealth creation comes to an abrupt end.

03 Artificial intelligence blows away the "cold air"

Is there a sign of recovery in the semiconductor industry in 2024?

"If I had talked about this topic in the first half of 2023, I would definitely give a negative answer, but now my judgment has shifted. Cheng Bing analyzed.

From the perspective of the industry cycle, this round of semiconductor downward cycle will run until the end of 2023, and the bottom of the industry has been confirmed. The current consensus in the industry is that the semiconductor industry will come out of the bottom of the downward cycle in the first half of 2024. TSMC's corporate briefing, which is known as the barometer of the semiconductor industry, also revealed positive signals.

TSMC President Wei Zhejia said, "The global economic and political situation is uncertain, but we are still optimistic that the output value of the overall semiconductor industry will grow by 10% this year, and the wafer foundry is expected to increase by 20% year-on-year." ”

The high prosperity of the chip industry is driven by technological innovation and the application of the consumer market, and the explosion of the AI industry will contribute a lot to the recovery of the semiconductor industry. As can be seen from the skyrocketing stock price of Nvidia, AI has become a definitive opportunity.

Musk previously said at the BCW event held by the Bosch Group, "Artificial intelligence technology is advancing rapidly, and I have never seen any technology that can advance faster than it, which explains why Nvidia's market capitalization is so huge because they have the best neural network chips." ”

Although Musk did not propose a 7 trillion core manufacturing plan like Sam Altman, he gave his prediction of the development of the industry, "The chip boom is bigger than any previous gold rush." ”

As the second largest category of integrated circuits, memory chips occupy the core position of the industry, and also have an important role in the prosperity of the semiconductor and consumer electronics industries.

"AI has driven HBM's shipments (climbing), and major suppliers of ordinary DRAM have reduced production last year," said semiconductor analyst Guan Ji to Xinchao IC.

Guan Ji has long tracked the market trend of the semiconductor industry and is known for blowing cold air in the circle, but this year's Semicon China conference changed his mind, "At this exhibition, I heard a lot of good news in at least one aspect: the overall trend of the semiconductor industry is slowly picking up. ”

According to a data provided by him, the expected sales of global semiconductor devices in 2024 have exceeded 570 billion US dollars, gradually approaching the peak in the past 10 years, and the expected year-on-year growth rate is more than 11%.

"In February, there was a temporary pullback in memory and AI chips, but semiconductor devices as a whole are on the rise", and for the growth rate of more than 11% in 2024, his summary is that sales in 2023 are too low, in other words, "bottoming out".

If the data is rational, the popularity of Semicon China is a more emotional cognition.

"The scale of this event is probably the largest in history, with more than 1,100 exhibitors in 10 venues, and next year's venues are sold out in three days. The average is more than $600 per square meter. The popularity of the industry can be seen. ”

However, the problems of "low-end", "burning money" and "lack of profitability" in the domestic chip industry have not been completely solved.

SEMICON focuses on the upstream manufacturing of semiconductors, and the theme of the exhibition focuses on "stuck neck". In Ye Maochang's view, the hot semiconductor upstream equipment and materials reflect the neglected chip design industry. "I can boldly say that the big fund was established in 2014 and the brilliant science and technology innovation board was launched in 2019, and the first round of Kangbo's upward cycle belonging to the chip design industry has ended. ”

Ye Maochang believes that most of the fate of chip startups is to be merged, "If you can't make money from the market and can't return to normal business logic, then there is no need to survive." ”

Like Cheng Bing, he also thinks that such a reshuffle is a good thing. He said, "One is to reduce the disorderly involution competition in the industry, objectively increase production capacity, which is beneficial to truly valuable companies, and the other is to brew the next Compo cycle after a period of consolidation." ”

In other words, in the eyes of entrepreneurs, the road of small team entrepreneurship in the field of chip design has come to an end, and process manufacturing and advanced process semiconductor equipment are the breakthrough direction of the domestic chip industry.

"From the perspective of the completeness of the semiconductor design industry, the product line involved in the existing domestic design companies is not very perfect, and most companies are crowded in a certain outlet, pursuing a quick listing, and there are not many people who really want to solve the problem. Next, startups need to be clear that involution is not determined by market size, but by competition. Don't follow the trend, don't be impetuous, calm down to develop technology is the most important. Zhong Lin said.

Disclaimer:

1. The content of this article is the original of Xinchao IC, the content and views are for reference only, and do not constitute any investment advice, and the information cited in the article is from the market public information, and our company does not make any guarantee for the accuracy and completeness of the information cited.

2. This article shall not be reproduced, reproduced, published, published or quoted without permission. If you need to reprint, please contact us.

View original image 60K

•