When we tried to trace the 30-year trajectory of China's integrated circuit industry, we found that there was an insignificant but important turning point in 1985.

In this article in the Economic Observer, the author explains this point in detail.

Here is the full translated version of the article:

As one of the main builders of the industrial ecosystem of the nearly 80 billion yuan listed company, Gao Feng looks very relaxed.

It was a sunny afternoon in mid-January, and I was with this investor in an office building in Shanghai.

He graduated from Tsinghua University's Radio Department, Class 85 (Radio Department is now Electronic Engineering Department, Class 85 students call themselves "EE85"). The hotel lobby was too noisy for a prior appointment, and there just happened to be a company nearby that his fund invested in.

Behind him were flourishing fortune trees, half-walled patent plates, and various certificates of honor. At the time, China's semiconductor-listed sector was continuing to rise and was about to hit a collective record high.

Even not so long ago, under the U.S. tech blockade, some of China's leading tech companies were on the verge of disaster. China's semiconductor industry is not as fragile as glass.

The industrial map was clearly being reconstructed, and Gao Feng was still in a happy mood.

"The semiconductor is even more so, sooner or later it will be the world of China." He believes that China's market is large, there are more talents, and China's technology industry is in urgent need of upgrading. It was one of the big reasons he left the world's most famous semiconductor company and returned to China.

From licensed semiconductors to TSMC to Huahong NEC, Gao Feng's previous career covers the key chains of semiconductor production.GigaDevice founder, Zhu Yiming, simply invited Gao Feng to join Shixi Capital in 2017 to build GigaDevice's industry chain ecology, Shixi Capital is the fund that GigaDevice has funded.

Just as Gao Feng had just entered Tsinghua in 1985, Moore's Law had been issued for 20 years, in 1 square centimeter of the silicon wafer can be integrated 35 million transistors, the semiconductor industry has developed to the stage of ultra-large-scale integrated circuits.

At that time, Intel introduced a new generation of microprocessors 80386, the U.S. AMD began to venture into the CPU field, Silicon Valley venture capital was in full swing.

But at this time in China, the semiconductor industry is just starting, Tsinghua University Microelectronics was founded only five years ago, decades behind the leading level of foreign industries.

At the same time, the Department of Radio Studies of Tsinghua University welcomed a group of freshmen such as Zhao Weiguo, Yu Renrong, Shu Qingming, Zhao Lixin, Ren Zhijun, Zhao Lidong, Feng Chenhui, Gao Feng, Liu Weidong, Lv Huang, Yu Qunhui.

Now 35 years later, these names, as well as their classmates and alumni, in the past few years to support the semiconductor industry in China's half of the river.

In the vast and unpredictable wave of industry, it is difficult to define the role of a group of people to promote the development of an industry, not to mention the face of world-class giants, high-end chips almost all rely on the import of the Chinese semiconductor industry.

When we attempted to trace the 30-year trajectory of China's integrated circuit industry, we found that there was a small, insignificant oddity that emerged in 1985.

It is not entirely statistical that at least half of the company's core executives and team members have graduated from Tsinghua University during the rapid growth of China's semiconductor industry over the past decade or so.

Among them, from EE85, the main training department of Tsinghua semiconductor talent is the largest number of graduates in the industry so far.

Will Semiconductor, GigaDevice, Galaxycore, Enflame Technology, Henghui Electronics, and a host of other companies in various semiconductor chains were born from their hands.

This group of people involved in the creation of semiconductor companies has been listed, with a total market capitalization of more than RMB 3,700 billion, accounting for more than 20% of the total market capitalization of semiconductor companies listed in China.

In China's A-share semiconductor listed companies with a total market capitalization of the top 20, they occupy five seats, covering the semiconductor industry chain in the field of IC design, manufacturing, packaging, terminal equipment, etc.

Overall, 1985 was a delicate year, and it seemed like a gamble for some of the students who chose the semiconductor direction at the time.

By the time they graduated in 1990, almost all semiconductor graduates were unable to find jobs in the country that were suitable for them.

In the 35 years since then, this generation has gone through almost every important stage of the development of China's semiconductor industry, they have experienced the story of the past 30 years of development, is a vivid condensed history of China's semiconductor development.

Thirty-five years after 1985, China has not changed the reality that high-end chips are almost entirely dependent on imports, even though its market size is the largest in the world.

At a time when China's chips are in desperate need of a breakthrough, it is of special significance to look back on this history to find the footprint of their struggle.

Typical curve

In 1990, it was clear to Gao Feng and his fellow students who had chosen one of the semiconductors directions that the opportunity was abroad: the major they were studying was a hit in the global industry.

When it was, Japan Semiconductor like a day, half of the world's semiconductor products have been manufactured by Japanese companies, in order to maintain dominance, the United States began to support South Korean companies and confrontation, the semiconductor industry global gaming is in full swing, the global Internet era just opened.

Zhao Lidong went to the United States to study, and now he has returned to his home country and is at the top of the global semiconductor industry.

The Enflame Technology, founded by Zhao Lidong, launched the AI training chip "Suisi DTU" less than two years ago. The product is China's first independently developed artificial intelligence high-end training chip after Google and NVIDIA.

Sitting across from the reporter, Zhao Lidong, his hair unkemptly combed, the collar of his polo shirt ironed flat, speaks with a habit of carrying English.

In his words, when I first went to the United States, Silicon Valley was called "Silicon Valley", after 2000, the semiconductor boom in Silicon Valley faded and the Internet was in full swing, it was renamed "Software Valley".

From its roots in Silicon Valley in 1994 to the founding of AMD's China R&D Center in 2007, Zhao Lidong did not expect to be so closely tied to China's semiconductor industry.

"By the time we graduated, the domestic semiconductor industry was so poor compared to foreign levels." Even in 2007, when Zhao Lidong returned to China to found AMD's China R&D center, the domestic situation remained the same. But as Zhao Lidong says, today's industrial boom is the result of a combination of factors.

"As if you were a cook, but had nothing in your hand, and still couldn't make a very good banquet, the same thing." Zhao Lidong said the accumulation of knowledge and talent over the years has laid the foundation for the development of China's semiconductor industry.

AMD, for example, is in a unique position in the history of global semiconductor development. After acquiring ATI, it became the only company in the world to combine X86 CPU and GPU technology.

AMD China R&D Center started with less than 200 people and grew to more than 2,000 people five years later.

At this international manufacturer, we have developed and grown a large number of local IC talents who have been involved in the design, validation, and mass production of cutting-edge GPU and CPU chips, and have learned advanced process technologies and management processes by doing.

The "AMD phenomenon" is a microcosm of China's semiconductor industry accumulation.

Here, there is the establishment of a complete team, advanced process management, have real-world experience, all do is high-end chips.

Enflame Technology is not the only company that has benefited from the long-term development of the industry, and the entire Chinese semiconductor industry has also achieved rapid development in the last decade.

In fact, Zhao Lidong is better known to the outside world as the former vice president of Tsinghua Unigroup. Interestingly, the location he chose to meet with the reporter was the Unigroup Cafe outside the east gate of Tsinghua.

From here, to the south, is the Qidian Technology Building, and to the north is the Unigroup Building, which is also inextricably linked to Tsinghua University, witnessing the trajectory of China's semiconductor industry from development to leap forward.

Unigroup's sprint

On September 24, 2014, two major events occurred in the semiconductor industry. The National Integrated Circuit Industry Investment Fund was formally established; the world's chip giant Intel also announced on this day that it had invested about $1.5 billion in Unigroup's Spreadtrum Communications and Ryderco Microelectronics' holding company.

At the time, Chairman Zhao Weiguo had already set a goal for Tsinghua Unigroup to become a world-class chip giant, and M&A was one of the important paths to achieving that goal in his eyes. Zhao Weiguo's speech at a seminar in Yizhuang, Beijing in 2015 shocked many people. "Capital invests in us, we take the money to make the industry, and capital goes to the capital markets to arbitrage," he said.

Zhao Weiguo is exploring the development of integrated circuits in China in a "refreshing" way: do your own R&D if you can, or acquire from overseas if you can't. The targets of Tsinghua Unigroup's acquisition are, without exception, from the most powerful companies in the chip industry chain.

Integration after a successful M&A is also an issue. Previously, the Intel-funded holding company, behind which the communications baseband chip companies Spreadtrum and Ryderco, have been competitors. The fear of being "eaten up" by Spreadtrum has caused a stir in people's hearts. The task of rebuilding Tridonic's management and core technology team fell to Ren Zhijun, Zhao Weiguo's hometown, and university classmate.

Ren Zhijun, born in Kulle, Xinjiang, is thin, tall, quite scholarly, and his classmates affectionately called him "Ren". It was then that Ren Zhijun received a call from the depressed Zhao Lidong from across the ocean. Plans to move one of AMD's product lines to Zhao Lidong in Beijing were hampered by American colleagues.

"You hurry back, chances are good." In December 2014, after meeting with Zhao Weiguo, Zhao Lidong officially came to work at Unigroup and was appointed as the general manager of Tsinghua Unigroup three months later. Zhao Lidong spent more than two months to reorganize the products of Tsinghua Unigroup, and after rebuilding the management of Tsinghua Unigroup, Tsinghua Unigroup basically came out of the trough. In March 2017, Zhao Lidong was promoted to vice president of Tsinghua Unigroup.

However, in a chain such as a highly competitive industry, success is not easy. The semiconductor industry is highly segmented and the globalization of the division of labor is very detailed, originally this is one of the best chains of the industry, but with the recent U.S. continues to impose a blocking line on China's high-tech industry, making this industry at risk of decoupling.

"When I first graduated, the technology was all overseas, and making high-end chips back then was empty talk. But for so many years, the government, industry investment, multinational or domestic companies are cultivating talent, coupled with the international situation's push back on autonomous and controllable, this is the best point in time." The golden age of semiconductor development in Zhao Lidong's eyes is just beginning in China.

Similarly, in Ren Zhijun's view of the development of China's integrated circuit industry, "The big fund came and lit a fire, Zhao Weiguo came and lit another fire, and a few plants burned exceptionally well and finally had a blazing momentum." Now the latest big fund to support semiconductors, is also in the second phase of investment.

Today, Unigroup has completed the "core to cloud" high-tech industry chain layout. The chip part is dominated by Cheung Kong Storage, Unisoc, and Guoxin Micro, while the network part is dominated by Xinhua San and Unigroup shares.

Unigroup's strength in the chip sector and the rise of China's IoT industry have opened many eyes to entrepreneurial opportunities, and Zhao Lidong jumped on the AI chip bandwagon: "High-end training chips are a monopoly market and are exactly what China needs."

Winning for phones

When Zhao Lidong decided to start a business, Ren Zhijun also decided to start a company from scratch. The largest investment in his business came from his classmate, Yu Renrong, the founder of Will Semiconductor.

In the eyes of the students, Lao Yu is an "extremely intelligent person". After playing mahjong overnight the day before yesterday, she went to the all-school freshman math competition the next day and got first prize.

Yu Renrong is also one of the richer people in the eyes of his classmates, in fact, he will take Haidian's rolls to Baoding to sell them to "make a little money".

Yu Renrong, who is now at the helm of Will Semiconductor, spends most of his time visiting customers because "customers are the most important part of business logic."

In 2017, domestic semiconductor companies went public intensively, and Will Semiconductor, located in Shanghai's Zhangjiang River, is one of them, which now brings together some of the most important innovative forces in the Yangtze River Delta.

On the day of the interview, Yu Renrong had other arrangements, so the assignment was given to Han Jie, a classmate of Will's for many years.

After the interview, Han Jie took the reporter to his office to visit Yu Renrong, and from his office, looking out, the Chuanyang River meets the Xigou Port, and 1.4 kilometers along Zhang Heng Road is the Innovation River.

The public company founder, who made the 2020 Forbes list of global billionaires, looked down-to-earth, wearing a black down vest and a plaid shirt with rimless glasses and a smile that narrowed his eyes into a line.

He doesn't speak fast enough to make you guess what he's going to say next, stressing the importance of customers and the market, saying that the semiconductor industry is constantly changing, always changing, which is the most interesting thing about this industry.

In August 2018, Will threw 13 billion yuan to acquire Beijing Howe, one of the world's top three companies in CIS and hooked up with a world-class first-tier brand manufacturer.

In terminal devices, CIS is equivalent to the "eye" in semiconductors, and cameras, iris recognition, and face recognition are all used. It has the largest share of the mobile phone market in downstream applications and the first share of the image sensor market in the security and automotive sectors.

With the acquisition of Howe, Will has broadened its market reach in the security and automotive sectors, bringing Howe's products to Will's customers and connecting Will with world-class first-tier brand manufacturers through Howe.

But compared to the development path of Unigroup's M&A, distribution startup Yu Renrong has previously taken more of a market pushback product path: what the market needs, and design the product in turn.

After 2003, the Chinese mobile phone market began a round of rapid development, and Yu Renrong saw the surge coming.

Before the company was founded in 2004, Yu Renrong did not intend to make the product first but rather cut into the mobile phone market by signing the agency line.

At that time, a group of people from DXN Wireless and ZTE came out to do IDH (Independent Design House).

IDH is a bridge between upstream IC manufacturers and downstream OEMs and is closer to the market than IC designers.

It is for this reason that Han Jie recalls that the group of IDH design firms that were founded that year, Dragon Flag and Wentec, had a great influence on Will's founding.

Yu Renrong decided to abandon all his previous business when he was planning to make mobile phones when Will's Shanghai company still had a large part of the market for electric bicycles, which everyone didn't quite understand at the time of the chop.

At that time, mobile phones were foreign brands, and people could not see any hope for the growth of domestic brands.

But coinciding with Yu Renrong's view is old classmate Zhao Lixin - founder of Galaxycore, "Semiconductor companies when phones win, they all win, when phones lose, they all lose."

In 2003, Zhao Lixin returned from the United States to found Galaxycore, and the year before that, Ningbo Bird's annual sales of 6.785 million units, among the top three mobile phone market in 2002, with a market share of 10.4%.

Before returning to China, Zhao Lixin spent three years as a licensed semiconductor manufacturer in Singapore and one year as a CIS designer at ESS in the U.S. before moving to UTStarcom where he gained extensive manufacturing and design experience.

Before Zhao Lixin returned to China in 2003 to found Galaxycore, Ningbo Bird's 2002 sales reached 6.785 million units, ranking among the top three mobile phone market sales in 2002, with a market share of 10.4%.

But in China's mobile phone industry chain at that time, cameras were still uncharted territory. Despite the large market space, there were no design companies in China, so Zhao Lixin chose to start with a camera sensor.

Wei Jun, an old school friend who joined Galaxycore in 2005, recalls that Zhao Lixin visited Vimicro Corporation after the first product was introduced in July 2005.

That year, Vimicro Corporation was the first Chinese chip design company to go public on NASDAQ with independent intellectual property rights, and it was gaining momentum.

The people at Vimicro Corporation said, "Mr Zhao, if you can make the product according to this request, and do this price, and meet our request, you can make ODM for Vimicro Corporation."

This product is a sensor mounted on a computer camera. To capture Vimicro Corporation as a customer, Zhao Lixin took the order.

In 2006 and 2007, many peers entered the PC space. With Taiwanese, Korean, and American companies competing with them in the sensor industry and looking to lose money in 2007, Zhao Lixin made a key decision: to enter the mobile phone market.

In 2006, China sold 109.19 million mobile phones, and at the end of 2007, Galaxycore made the first camera sensor designed to the phone's specifications, and in 2008 the company rapidly expanded from a few million dollars to more than $20 million in revenue.

Riding the easterly winds of mobile development, Galaxycore's revenue exceeded $100 million by 2010 and peaked in 2014.

Also changing the fortunes of the mobile phone market is Feng Chenhui, co-founder of Maxscend Technologies Electronics, which is now the leading RF front-end chip company in the A-share market.

In 2009, Maxscend Technologies' first product, the digital audio chip, was introduced. At the time, FM radio was in common use, and Feng Chenhui and the team were looking at the DAB standard for digital audio broadcasting.

In 2010, the money from the angel round was almost spent, the existing products had no market and no customers, and the new products were not yet available.

Feng Chenhui described the state of work at that time in terms of suppressed, confused but persistent. Still inspired by the exchange with old classmates.

Yu Renrong said, "Definitely look at what the major customers are doing, look at what the major chipmakers are planning, and how they're going to work with it, rather than coming up with something behind closed doors that's different from others."

In 2014, Maxscend Technologies turned the ship around and made RF front-end chips the company's main focus.

"For quite some time, Samsung has been our biggest customer and now includes Huawei, OPPO, Vivo, Xiaomi as well."

In Feng Chenhui's view, the domestic mobile phone makers that have been growing over the years are the breeding ground for chip companies like them to continue to grow. The Science and Technology Creation Board and the National Fund are more like icing on the cake.

The next decade

When China's chip industry blossomed everywhere, time went back 35 years ago, Gao Feng they face a completely different scene, with the words of investor Lv Huang, thirty years ago, China's semiconductor industry and the gap between foreign countries is "the gap between kindergarten and university", now it is "the gap between junior high school and university".

Liu Weidong still remembers the days when teachers were given "odd jobs" during school when people used coordinate paper to engrave red film to design chips, while abroad were already using computers to design chips.

To the Chinese Academy of Sciences Microelectronics Institute of Gao Feng, in memory of that year, "the most advanced domestic production line can only do 4 inches and 5 inches, cleaning equipment to clean wafers are manually pulled out, according to the instructions of manual cleaning, while the international has a lot of 8-inch automated production line."

Gao Feng's work experience with Singapore Licensed Semiconductor and TSMC gave him a better understanding of the huge gap in the domestic semiconductor industry.

In the early days, this group was a bit like a lone sailor paddling a small sampan toward a spinning storm.

In Tsinghua Institute of Microelectronics, this at that time has represented the leading level of the domestic semiconductor industry in the laboratory, can often see such a scene, a group of researchers lying on a large table, close to the drawing to check for errors, verify the design layout with the naked eye.

Even by the 1990s, after the reform and opening up of a large amount of foreign investment into China, China semiconductor began to have momentum, but still slow progress, "market for technology" thinking, it is difficult to bring substantial help to the development of semiconductor.

Gao Feng remembers his graduate studies at South Asia Semiconductor Co., Ltd. in South Korea, where he learned to design chips using the American Compass, an internationally advanced LSI design software.

In 1995, this country, which is one sea away from China, achieved DRAM catch-up with the United States and Japan and developed 256M DRAM ahead of Japan and the United States.

As one of the few female semiconductor majors, Yu Qunhui stayed in his teacher's company after graduation - Hongsi Electronic Technology, which was acquired by GEM listed company Feitian Integrity three years ago, focuses on the independent development of information security application IC chip design company.

In order to survive, they start from reverse design, take on the task of design commissioned by customers, disassemble and analyze the existing products on the market, figure out the functions, and then make their own products, and then take them to customers for production, which is equivalent to design services.

"This (reverse design) is also accumulation, and without this process, it's hard to have the ability to design-forward behind us." Yu Qunhui said. The path of reverse design is the path taken by many Chinese chip companies in the 1990s, much like the U.S. learning from the fruits of the British Industrial Revolution back then, "learning, absorbing, reinventing," says Yu Qunhui.

The market is fickle. While Maxscend Technologies was gradually getting out of the woods, Liu Weidong's Juho Electronics was just launched in 2014, focusing on sensor signal conditioning chips. But this type of chip has a long payback period and few opportunities for investment. The first investment still came from roommate Lv Huang.

Liu Weidong originally wore a beard, saying it stood out in the crowd. Later, the beard was shaved and the students laughed and said they were much younger.

The content of his WeChat friends circle is almost always the introduction of his company's products, and even during the holidays, he can always see Liu Weidong posting about Jiu Hao electronics in his friends circle. He is also the secretary-general of the Tsinghua Alumni Association of Semiconductor Industry Association, and his classmates call him "Mr. Liu" and enthusiastically organize various alumni activities related to it.

Before the interview, he even typed out his answers on the computer one by one according to the outline.

Like Liu Weidong, LV Huang, who has a strong dialect accent, says he is not good at storytelling.

When he visited Anxi in Quanzhou, he used to chat while making tea. Those who have invested, he has taken with him but is happy to share his judgment on the industry.

After graduating, he returned to his hometown of Quanzhou to work in the industry, and after getting involved in investment, LV Huang found that his old classmates were "more involved in chips".

Combined with the old expertise, looking back at LV Huang's investment layout, the chip project accounts for about 70%. Of this, Lv Huang invested $4.1 million when Will was founded in 2007, and he was one of the early investors in GigaDevice.

The first time Liu Weidong asked LV Huang what he would do if his business failed, and LV Huang said he couldn't, and if he failed, he would pay back the money he had bought him for drinks.

It turns out that when he was an undergraduate at Tsinghua, one year LV Huang didn't come home for New Year's, Liu Weidong returned early from his home in Shanxi, carrying bamboo leaf green and rabbit legs.

Lv Huang picked him up at the south gate of the school and said he would definitely invite Liu Weidong again after he graduated and earned money.

"So be nice to your classmates, especially those in the same dorm." Speaking about the experience years later, Liu Weidong joked.

So far, many have crossed the eye of the storm, but all are still on their way. For this group, the next two years are a crucial test.

Now, the new paradigm shifts brought about by digital transformation and artificial intelligence, the Internet of Things, drones, Industry 4.0, etc., all these new things contain advances in semiconductor technology.

The students of Tsinghua University EE85 have undoubtedly become a force to be reckoned with in this trend. Many have attributed this to "the times". In their eyes, a "once-in-a-lifetime" opportunity lies ahead in the next decade or more.

The people interviewed are happy to analyze the situation, different experiences, each person in the interview has their own focus on the analysis of the situation, but one thing is the same, just like Gao Feng remembers what the university class teacher Zhou Yucheng often said to them: semiconductor is difficult, but there must be hope for the future, great development prospects.

The new coronavirus epidemic looms large at the moment, with a registered capital of 204.15 billion yuan of the National Integrated Circuit Fund to begin the second phase of investment. Gao Feng has no plans to slow down its investment, despite the impact on business confidence.

Driven by the domestic "new infrastructure", his view of the semiconductor, the need for domestic replacement is receiving unprecedented attention.

5G infrastructure, new energy vehicle charging piles, large data centers, industrial interconnection ... semiconductors will continue to play an important role in it.

"Including Wei Shaojun, Chen Da Equivalent this group, they count our older generation, the elder brother, they have seen, experienced, and even personally participated, they always want to see in their hearts the company they support, to be able to fulfill the original promise, to be able to really make China's big chip."

As the interview was coming to an end, it was still early in the afternoon, after 4 pm in Beijing, and looking at some office buildings already lit up in the distance at the Wudaokou entrance, Zhao Lidong suddenly took a turn: "All these people have this dream in their hearts, and I feel it very deeply, you know?"