Tianlin Wang, Managing Director of Intel Capital and General Manager of China
[Editor’s note] The author of this article, Wang Tianlin, managing director of Intel Capital and general manager of China, is from Aijiwei, which is authorized by Entrepreneurship State.
At the end of last year, I wrote an article sharing the analytical framework of investment in the chip design industry, which caused some repercussions. The semiconductor industry has continued to thrive in the past year, and industry friends suggest sharing their observations on the industry again, as well as their experiences on investing in other areas of semiconductors. The following are some of my recent thoughts and analyses on the industry. I hope to share with you.
The nature of capacity shortages
The key word in 2021 is production capacity. Almost all friends in the electronics industry are concerned about how long the production capacity shortage will last. This chip shortage, whether it is the span of time or the depth and breadth of the impact on the upstream and downstream industrial chains, is unprecedented. In fact, the essence of the production capacity shortage is the change of supply and demand. The key lies in figuring out which are temporary factors and which are long-term fundamental changes in order to make better judgments.
From the demand side, although the short-term impact of hoarding and epidemic disruption to the supply chain caused by trade disputes is relatively large, it is not a fundamental change. The fundamental reason is the increase in demand for semiconductors in many industries. For example, the chips used in cars before were limited, mostly simple 8-bit and 16-bit MCUs. However, the development of the Internet, artificial intelligence, and new energy vehicles has amplified the demand for automotive chips: vehicle body intelligence has fundamentally changed the architecture of automotive electronics, and new demands for 32-bit MCUs and ECUs have emerged; in-car entertainment and smart cockpits have been continuously upgraded, so that the market for automotive SoCs has grown year by year; it is only a matter of time before autonomous driving chips become mature and popular, and the amount of related power management, battery management, and charging management chips and devices will also increase exponentially. These new trends also expand the demand for traditional chips in automobiles. For example, many sensors have greatly increased the amount of data communicated in the car. Ethernet chips that were not used in the past may have to be equipped with several or even dozens in each car in the future.
In addition to cars, the spread of 5G technology has also increased the demand for chips for mobile phones and base stations. Both cars and mobile phones belong to the category of end-users, and the data generated on the end needs to be processed by the cloud computing and data centers behind them, which will in turn drive the demand for server chips. These are long-term fundamental changes that will not disappear with the improvement of trade disputes or epidemics.
From the supply side, although advanced processes tend to attract more attention, most of this wave of capacity shortage is actually concentrated in mature processes. Take the iPhone 13 Pro as an example, only four of the nearly 50 chips in it use advanced processes: SoC, Modem, DRAM, and NAND. The remaining 40 or so are manufactured in mature processes above 20 nanometers, typically represented by RF and power management chips. However, most of the wafer fabs have not planned to expand the mature production process in advance, and the construction and verification cycle required for new or expanded production is not overnight. Based on this, the chip shortage may continue for some time before the new production capacity on the supply side is released in large quantities.
Now is the best time for technology entrepreneurship
There’s no question about it.Now is the golden age of China’s semiconductor development, and unprecedented challenges have also brought unprecedented opportunities.As mentioned in the previous article, there are three elements in the development of enterprises: market, capital, and talent. After two decades of development, China has become the largest market for semiconductors in the world. According to statistics, the global share of semiconductors purchased by Chinese customers has increased from a few percent at the beginning of this century to one-third or more at present. Taking the mobile phone industry as an example, the main customers of chip companies have gradually evolved from Motorola, Ericsson, and Nokia to Huami OV and Apple Samsung. More importantly,The trade dispute has spawned demands from local customers for supply chain security, giving them more incentive to bring in new local suppliers. These multiple positive factors provide chip entrepreneurs with an unprecedented opportunity.
Secondly, a series of policies in the capital markets have also greatly promoted the development of the semiconductor industry. It can be said that the semiconductor industry is one of the biggest beneficiaries of the Science and Technology Innovation Board: 47 of the top 347 companies listed on the Science and Technology Innovation Board are related to semiconductors, accounting for one-third of the total market value of the Science and Technology Innovation Board; among the top ten companies by market value, five are semiconductor companies. The newly established Beijiao Exchange will also be a good platform for many smaller companies. Due to the many industry segments, many semiconductor companies are located in sub-segments with limited size but indispensable position in the industrial chain. Such "small but beautiful" companies are very suitable for the positioning of Beijiao as "specialized and innovative". In other words,No matter "big and strong" or "small and beautiful", as long as it is a semiconductor company that seriously serves customers with products, it has the opportunity to occupy a place in the capital markets.
In terms of talent, many peers believe that high-quality semiconductor startup teams have been tapped, but we believe the opposite. From the perspective of product technology, even listed companies are still significantly behind world-class semiconductor companies. The market opportunity is by no means just "domestic substitution", and the upside potential is infinite; and the maturity of customers and the favorable capital markets will also attract more industry veterans to join the ranks of entrepreneurs.The current Chinese market is huge, with abundant capital and gradually mature talent training. It can be said that it is the most suitable place in the world for technology entrepreneurship.We believe that we have also seen more technology companies and talents invest in true technological breakthroughs and upgrades with confidence.
From ABC to AAA investment philosophy
I often tell entrepreneurs: "Going public is not an exit, let alone a destination, but a new starting point." When I first started investing in the industry, after careful analysis, I divided each project I saw into three categories: A, B, and C: Class A is definitely something that should be invested in and supported. The company will develop benignly and go public successfully in the future. If you miss this project, you will regret it so much that you can’t sleep. Class C generally has various shortcomings and should never be invested. Class B is somewhere in between, which belongs to whether you can invest or not. But we may also missee and sometimes mistake Class A projects as Class B, so most of the projects our team has launched are Class A, with a small number of potential Class B projects. As more and more investment companies went public, we found that even among Class A companies, there were significant differences. Therefore, we drew on the ratings of the bond market and further subdivided Class A into three grades: A, AA, and AAA. To put it simply, the market value of Class A companies is several billion, AA companies are the leaders in their respective market segments, with a market value of around 10 billion to 56 billion, and AAA is stable at more than 70 billion (10 billion U.S. dollars).Now the focus of our team’s work is also how to findHow have Class A companies evolved?Discover the potential to becomeAA and AAAAt the same time, we also believe that excellent investors can help entrepreneurs achieve success through systematic empowermentClass A companies grow into AA companies.FromAA grows to AAAThe leap.
In the long run, we believe that the domestic secondary market will continue to differentiate into a pyramid structure, and the semiconductor industry is no exception: a very small number of AAA companies located at the spire are companies that have a profound impact on the industrial chain, and are popular stocks allocated by fund managers; a few AA companies, as the leaders of their respective market segments, are within the scope of fund managers’ attention and may be upgraded to AAA companies; and with the increase in the number of listed companies, most companies will belong to Class A with limited liquidity, and few secondary market fund managers will pay attention. Some Class A companies with small performance and limited growth prospects have even broken up after listing, and their liquidity is limited. They are likely to be acquired by AA or AAA companies in a few years. The domestic M & A market is likely to gradually form and mature at that stage. The enthusiasm of the secondary market has also promoted the popularity of the primary market. In the past two years, various investors have rushed into semiconductor projects. For a while, rare goods can be found, and valuations have risen. Due to the pursuit of new investors, the valuations of many A-class semiconductor companies have risen to billions or even billions. If companies cannot transform and upgrade to AA, it is easy to have a situation where the primary and secondary markets are upside down. We believe that entrepreneurs and investors should stay calm in this heat and avoid blindly pursuing high valuations.
The founder’s ability and structure are often the most important factors in determining where a company can goIn order to achieve leapfrogging and become an AA or even a AAA company, in addition to its own efforts, it is also important to expand product categories and new markets by introducing shareholders and partners who "can help and not cause trouble".The vast majorityAAA was gradually developed from AA.Huaqin Technology is a good example. In 2017, we and Yitang Huachuang jointly led the first round of investment in Huaqin. Last month, Mr. Chen Datong and I briefly reviewed: at that time, we thought Huaqin had just entered the ranks of AA. Through the unremitting efforts of the founders and the company team, the strong support of partners, and the empowerment of shareholders in the past four years, we look forward to Huaqin’s early growth into a AAA company.
How to identify leaders in segmented fields
Each field of semiconductors is complex and limited in space. Here are some brief insights into each segment.
Chip manufacturing (wafer fabrication) investment is a big deal. You can use 20nm as the demarcation point and divide it into two fields: advanced process and mature process. The technology and investment threshold of advanced process are getting higher and higher. After years of marathon running, only three or four companies in the world are still holding on. Unlike some industry insiders, I am a big fan of "Moore’s Law", but this does not mean ignoring the importance of mature process, both of which are important parts of the semiconductor industry. Most of the production capacity shortage this time is in mature process. Among them, 22/28nm is tighter than 12/14/16, and 40/55 is tighter than 22/28. Everyone is working overtime to expand production capacity. However, history has proved that,After each industrial expansion, there is overcapacity. To maintain long-term competitiveness, the differentiation of fabs is of great importance, especially in the differentiation and uniqueness of special processes.Taking the automotive industry as an example, a chip that realizes the same function has many times stricter product definition and manufacturing difficulty than consumer and industrial grades: car grade requires a temperature range between minus 40 degrees and 155 degrees above zero, far exceeding the 0 to 40 degrees of ordinary chips; the working life must be more than 10 years, far exceeding the 1-3 years of consumer grade. The domestic passenger car market has exceeded 20 million units per year, ranking first in the world, and is expected to exceed 30 million units in the near future. There is no mature car grade wafer factory in China, and the market potential is huge.
The upstream of manufacturing mainly includes three categories: equipment, materials, and partsThe threshold for equipment companies is relatively high. There are many domestic leading companies that have been or will soon be listed, such as Huachuang, Zhongwei, Shengmei, Yitang, and Tuojing, which have occupied a certain market share in their respective fields and established a clear leading position.If a startup team wants to create a similar product, it has no advantage in funding or customer certification, so it must rely on technological and product innovation to find a breakthrough.Since only a few domestic equipment manufacturers have been validated by advanced processes in some application fields, there is actually a lot of room for improvement. In addition, in the fields of polishing and grinding (CMP) and metrology (Metrology), we believe that more professional teams are needed to tackle innovation.
Semiconductor production processes are complex and require a wide variety of materials and gases, so each market segment is limited in size. We see many "small but beautiful" materials companies making a difference in their respective fields, but the room for growth is limited. From another perspective, managing many suppliers at the same time, the work of the fab supply chain team is also very trivial.In the next round of competition, we are optimistic about companies with the ability to expand and diversify their product portfolios – in addition to their own organic growth, it is also a very effective way to become bigger and stronger through mergers. We believe that the domestic materials industry will eventually emerge several leading companies. In fact, not many domestic materials teams start businesses from scratch, and more use the accumulation of other industries to cut into semiconductors.For example, Bochun Materials has successfully switched to the semiconductor field due to its years of accumulation in the photovoltaic industry. It is currently one of the few domestic material suppliers that have been certified by international wafer factories such as Intel, TSMC, and Micron. It will have the opportunity to expand its product portfolio to related fields in the future.
Parts and components were often overlooked by investors in the past, mainly due to the limited market size and the low technical threshold, especially the long time and high uncertainty to pass the certification of equipment companies. In a sense, this industry is a typical long cycle and limited return. However, due to the demand for supply chain security caused by trade disputes, and the determination of Beijiao to name and support "specialized and new" enterprises, more and more local parts and components companies have attracted the attention of equipment companies and chip manufacturing plants; we also hope that more parts manufacturers can directly provide users with professional, timely and customized services. Shenzhou Semiconductor is a good example. They have passed a rigorous evaluation with professional service capabilities and become an Intel-certified supply chain field service provider.
Due to the trade dispute, EDA has become the hottest track for semiconductor investment in the past two years. EDA tools are similar to materials and components, and there are many and complex sub-fields, but each volume is relatively limited. Although there is a large influx of capital, it may not be entirely positive for the long-term development of the industry.Reference globalEDAWe believe that subsequent mergers and acquisitions and integrations are only a matter of time, and eventually there will be two to three companies involvedAA/AAA leaders dominate the market.Therefore, investors entering the market at a multi-billion valuation at this moment must rely on projects that have selected AA/AAA to generate positive returns.
The packaging capacity of the semiconductor back-end can be said to be completely accompanied by a serious shortage of chip manufacturing. Therefore, most OSATs in China are still focused on expanding production capacity to meet the immediate needs of customers. But we believe that advanced packaging will be an important trend in the semiconductor industry in the next few years, and it must be more closely integrated with the previous chip manufacturing – the era of relying on one or two general-purpose chips to eat all application scenarios is over.Future computing is based on heterogeneity; and as one of the key technologies to support heterogeneous computing, advanced packaging will bring many new entrepreneurial and investment opportunitiesFor example, in the substrate, patch, bonding, cutting and other aspects, new equipment and material requirements will be stimulated. Huafeng Technology is a representative manufacturer in this field. They have entered the advanced packaging placement machine market through innovative design and technology, and successfully entered the market including Sun Moonlight and many domestic first-class OSAT customers.
One last thing, yeah.chip designThe company’s opinion. There are many domestic design companies, and they have made considerable achievements in low-power and low-cost chips, which are not inferior to international manufacturers. A good ecosystem has been formed on the end (consumer electronics, mobile phones, and the Internet of Things). The next breakthroughs are likely to appear in higher-performance computing, such as servers, base stations, and other industries used in cloud computing and edge computing. The design concepts and know-how of high-performance chips and low-power consumption are quite different. Take power management chips as an example.We are pleased to see that many domestic startups are actively expanding into high voltage on the basis of low current.We have invested in JEWAT Microelectronics, which is a good example. Although several founders are from leading international companies, the early stage of entrepreneurship began with consumer electronics and lighting to meet the power chip needs of the domestic industry in the early stage of development. After the low-end products became a red sea, JEWAT introduced partners to become shareholders with financing, while fully developing high-voltage and high-current products with high barriers and difficulty, and developed applications such as base stations, servers, automobiles, and personal computer CPU core power supply. Through several years of unremitting efforts and the support of shareholders, customers, and fabs, the product portfolio has been transformed into a high-end, and its performance has improved by leaps and bounds. It has grown into one of the leaders in the sub-industry.
Conclusion: Gathering and selecting, deeply empowering, long-term companionship
The general environment has given entrepreneurs rare opportunities, but in fact the number of high-quality targets is limited, and there are more and more institutions engaged in semiconductor investment, which poses a challenge to investors. Looking forward to 2022, Intel Capital will continue to empower the ecosystem, serve entrepreneurs, and be valuable capital. In addition to the above views on various segments, here are some of our ideas and plans:
First, the industry valuation is at an all-time high, so it is necessary toAt the earliest possible moment, find the best teamNow that the market is large, capital is sufficient, and talent training is gradually mature, more industry veterans will join the ranks of entrepreneurs. If the valuation has risen to billions, be sure that the company has the potential to grow into AA.
Second, "small but beautiful" cannot be ignored, especially those companies that are limited in size in their market segments but have an indispensable position in the industrial chain. The positioning of the newly established Beijing Exchange "specialized and innovative" provides new platforms and opportunities for these companies.
Third, focus on selection before the investment, and deeply empower after the investment. If the street is full of goods, deep empower is shouting slogans. An entrepreneur said to me: "If you are invested by Intel Capital, you will stand on the shoulders of giants." I half-jokingly replied: "There are not too many people standing on the shoulders of giants." We have been investing in China for 23 years, and we have always controlled the scale of 5-8 new investments per year, and the existing portfolio is maintained at around 40. Although I have spent two-thirds of my time on post-investment empowerment, time is always not enough.
Fourth, there are very few institutions that can truly achieve "long-termism".China Semiconductor needs entrepreneurs who adhere to long-term principles, as well as investors who truly adhere to long-term principles.We invested in Chiporiginal Semiconductor in 2005 and stayed with us until the company went public in 2020. On the one hand, we are one of the few industrial direct investors without investment period and duration restrictions; more importantly, we are strategic investors and recognize Chiporiginal as a long-term strategic and ecological partner of Intel.
Thank you for the support and trust of friends inside and outside the industry. On the premise of doing my job well, I will try to write a dry article every year to share my views and experiences. I wish all entrepreneurs all the best and make more and stronger AA and AAA technology companies.
About the author:
Wang Tianlin is currently the managing director and general manager of Intel Capital in China. He joined Intel in 2005 and led Intel’s Centrino Wireless Product Research & Development and Marketing teams in Oregon and Israel. After graduating from Wharton School in 2014, he returned to China and was responsible for business development and risk management of Intel’s end point computing department in China. He joined Intel Capital in 2016 and started to lead Intel Capital’s business in China in 2017, and re-established an investment team with an industrial background. He focuses on investing in technology innovation companies, including semiconductors, cloud computing, Internet of Things, artificial intelligence, communication technology, etc. Mainly invested in Huaqin Communication, Espressif Information, Zhuoyi Information, Kangxi Communication, Extreme Krypton, Airborne Intelligence, and Yizhi Technology, etc. At the same time, he led the China team to lead the investment of many start-up companies including Nanxin Semiconductor, Jiehuate Microelectronics, Gelun Electronics, Bochun Materials, etc. In addition, he has also served as a director or board observer in more than 20 companies, including Unigroup Zhanrui, Horizon Robotics, Chongqing Blue Bank and No. 9 Company.
This article is authorized by the columnist to publish in Entrepreneurship State, and the copyright belongs to the original author. The article is the author’s personal opinion and does not represent the position of Entrepreneurship State. Please contact the original author for reprinting. If you have any questions, please contact editor@cyzone.cn
关于作者