At 8: 00 p.m. on July 29, Yang Jiankai, a senior analyst at Yuanchuan Research Institute, made a transoceanic communication with Yu Zhihong, product line manager of Renesas Electronics USA. He has been involved in the Silicon Valley chip industry for more than a decade, has been involved in designing Tesla, Inc. Model S's IGBT chip, and now manages a product line worth US $100 million.

In the live broadcast, Yu Zhihong thought:

95% of the chips are not just R & D as the first priority, but more of a collective product. The role of the product manager is to integrate the people of each functional department, so there is no fixed nine-to-five workflow, everything depends on the product and market demand.

Under the general trend of chip localization and substitution, there is a good opportunity for China to have more tracks, several orders of magnitude more than that of the United States.Some applications, such as lighting chips, including some simple analog chips, are no longer possible in the United States, because the market prices in these areas are very low and there is no gross profit to speak of, but in China, such a business can still be done.

In overseas markets, domestic chip design companies will encounter fewer price challenges. But building a brand takes a long time, it is impossible to dominate the world by low price, and for more responsible customers, the price of the chip is not even the main concern.

Many chip companies are telling the story of electric cars. To determine whether there is a real opportunity, ask two questions: how many of the company's chips will be used in a car, and how much can a chip cost?

In the field of IGBT, domestic manufacturers still have a long way to go to establish internationally competitive brands.The gap between the specifications of many home chips and Infineon is very limited, but if more customers are sold, the problem begins to emerge.

Silicon carbide and gallium nitride have the revolutionary potential to completely replace silicon devices in some market segments, but there is no way to replace existing solutions in most markets.

The following is the streamlined live Q & A record.

The way to attack the Product Manager

YukawaMr. Yu used to work in Los Angeles, the birthplace of American electric cars, but how did he switch to the chip industry and be responsible for the chip design of Tesla, Inc. 's Model S model? Can you tell me about your specific work experience?

Yu Zhihong：I studied in the School of Machinery and Power of Shanghai Jiaotong University. After graduation, I went to the United States to study for a master's degree in electronic engineering, majoring in cybernetics, focusing on mathematics and partial to theory. But my first job was as a hardware design engineer, including power motor control, embedded design, and so on. In 2010, I began to understand a little bit about car chips when I worked in an electric car company in Los Angeles, working on buses, electric truck chargers and motor control.

In 2012, I joined International Rectifier (hereinafter referred to as IR), the International Rectifier Company (acquired by Infineon in 2015), which is responsible for transferring some industrial chips to cars for development and application. At that time, Tesla, Inc. was a very small company. IR happened to be working with Tesla, Inc.. I was involved in the development of the electric system of Model S, which became popular after mass production in 2013.

Since then, I have never left the chip industry and have been engaged in technology-based marketing and product management. Currently, I am the Chief Product Line Manager of Renesas Electronics in the United States, where I am responsible for an industrial power product line that manages approximately $100 million in sales. Based on about 10 years of product management experience in the American chip industry, I organized what I learned into a book.《I am in Silicon Valley Tube Chip-Chip Product Line Manager Survival GuideI would like to share with you today.

YukawaBoss Yu recently wrote a book called "I manage chips in Silicon Valley-- Survival Guide for Chip Product Line managers." how do you understand the role of product managers in chip companies? What kind of process does a good chip product manager follow to work or start a business?

Yu Zhihong：In the impression of many people outside the industry, the chip is a choking industry, which needs to concentrate research and development efforts to tackle key problems.However, 95% of the chips are not just R & D as the first priority, but more of a collective product, which, through good management, brings a group of people together, performs their duties, and then becomes a good company. not one or two bulls can decide everything.

When most chip companies are set up, there must be a main founder with the responsibility of product manager. The role of the product manager in the chip company is to integrate the people of each functional department. For example, how to understand downstream applications, define products, conduct research and development, support sales, serve customers, participate in competition, and so on, the product manager is a major force in kneading many links. Therefore, the day of the product manager is generally complicated, perhaps the first hour is to introduce the product at a key customer, an hour later to solve an internal production problem, and then an hour later, report to the superior leader why you want to make a chip.

However, no matter what you do, the goal of the product manager is that the company can make a profit. Specifically, it is through the implementation of a product R & D budget, grasp the appropriate time node, the product to the market, reach customers, large-scale sales, and finally produce an ideal sales, to achieve the established profit growth target.

Therefore, generally speaking, the product managers of chip companies do not have a fixed nine-to-five workflow, and everything depends on product and market demand.When I do a half-year work report with my superiors, the most important thing is to explain the changes in profits in the past six months and their reasons, and then whether I have made some new big clients, and finally, are there more new opportunities than before? These are some of the most fundamental criteria for measuring the work of a product manager.

YukawaThe domestic substitution of Chinese chips is a major trend, start-ups are born every day, and there is fierce competition on some tracks. In this context, how should chip startups choose the right track? As far as your experience is concerned, are there any good cases that can be used as a reference for current chip entrepreneurship?

Yu Zhihong：The chip industry, from the upstream refined sand, to silicon wafer production and equipment production, to chip design, wafer foundry, packaging and testing, and finally to downstream sales agents, is rich and diverse, each has its own way out. It's hard to say what the right track is.

In the United States, for example, some Cyrus executives have recently come out to be power companies adapting to FPGA chips, because they are familiar with it and can easily cut into the familiar surrounding tracks. There are some AI processor chips, the market is booming recently, they are also willing to take risks.

Under the general trend of chip localization and substitution, there is a good chance that there are more tracks in China, several orders of magnitude more than in the United States. Some applications, such as lighting chips, including some simple analog chips, are no longer possible in the United States, because the market prices in these areas are very low and there is no gross profit to speak of, but in China, such a business can still be done.

But just because there are plenty of opportunities doesn't mean you can do anything. Entrepreneurs should take their own background as the benchmark and do what suits them.

For example, employees of chip sales agents, after mastering some customer relationships, can jump out and become their own agents. Relatively excellent chip design engineers, who find that the sales prospect of a kind of chip they are familiar with is very good, they can bring their own experience out and set up a new chip design company. Product line managers are also a background for chip companies to start a business, because they are very familiar with the whole process from R & D to sales, and have a lot of resources to know how to do a thing well from beginning to end.

Finally, I would like to remind you that it is very difficult for chip startups to cross the border.Chips are very different from consumption, the Internet and other fields. If the founder does not have more than 10 years of experience in the relevant technology field, I personally think that it is very difficult to have the vision to make the company more long-term.

YukawaIn the domestic chip design industry, there are many generations of champions. After many companies have developed a small number of successful chip products, they become silent and no longer grow. The fundamental problem is that there is something wrong with the product route planning. So, how should chip companies do a good job in product planning?

Yu Zhihong：It is because many companies do not think about the long term at the beginning, or the so-called strategy only exists on PPT, and they have not done a deep thinking.

For example, when some domestic companies were established, in fact, it was mainly because a salesman or engineer found that TI had made a chip and had a sale in Huawei. He thought that the TI chip was not difficult, and I could also do it. After adopting the follow-up strategy, perhaps by chance, it was really sold to Huawei.

The question is, he knows what it is, that is, he knows that the chip is used by Huawei, but he does not know why, that is, why TI opened the chip at that time, how the chip got into Huawei at that time, and what other customers besides Huawei can be hit. What are the plans for TI in the future, and so on.He may not have thought about these questions. Gradually, other domestic companies began to copy his chips and compete on costs and products.

What is a long-term plan?

Vertically, for example, based on the success of chip product A, you can make a version 1.1 or 1.2 to further improve performance and strengthen customer relationships. Horizontally, other suitable products can be developed based on a good process and a good control core derivative change. Generally speaking, a company with relatively strong comprehensive strength, such as Si Lijie, used to make electricity meter chips and had a lot of meter customers, so, can it do other chips on the electricity meter? If you don't have a certain technology, can you buy it? In this way, he can dig deep along one application and sell more products to the same application.

Chip companies go out to sea

YukawaWhat are the implications for the mainland from the development experience of the chip industry in Taiwan?

Yu Zhihong：In fact, as early as the 1960s and 1970s, the island realized that the island was so small and lacked resources, and whether there were some industries that could bring good profits and solve the employment problem of many people. The final conclusion is that we can only do high-tech. Apart from Taiwan, the same is true of Singapore.

Therefore, they all chose the chip industry as a breakthrough. There are many people on the island studying in the United States. Invite them back, and then negotiate the introduction of some production lines. On this basis, the electronic industry has gradually developed. Through electronic OEM, the island has mastered some experience in design and application. In this way, the island gradually integrates the entire electronic industry chain, deepening understanding, integrating upstream and downstream, each link can complement each other and develop together.

As far as the enlightenment of the mainland is concerned, it is our industrial planning that should be based on the long term and have a long-term perspective for decades, not just to support some companies to list. If we really want to establish a tight supply chain, we need to make a decade or two of investment in the core links of foreign monopoly such as upstream raw materials and electronic automation software, or even customers will have to try and make mistakes for a long time in order to do a good job in these industries. Therefore, we should not be short-sighted and should establish some more long-term plans.

YukawaWhat do you think of the position of Chinese chip companies in the American chip industry and its enlightenment to us?

Yu ZhihongThere are many Chinese people working in the chip industry, and there is a Huamei Semiconductor Association in Silicon Valley, many of which are leaders. However, since they have lived in the United States for a long time, they do not necessarily have a lot of sense of belonging at home, so we should not expect them to manage and position themselves with some Chinese mode of thinking.

Our problem is that sometimes there is more internal friction and there are many non-market factors, so there may be no way for the market to freely choose the most competitive companies.

There is also a more important question: why many Chinese have developed the top innovation ability in Europe and the United States, and can start very cutting-edge companies, while many powerful Chinese in China can only do some business model innovation. This may be a problem that we need to think about in education.

YukawaAt present, some domestic chip companies have begun to go to sea, such as Si Lijie, or it is quite surprising. How should domestic chip companies go to sea?

Yu Zhihong：At present, few domestic chip design companies go to sea, and the market prices of many domestic chips are much lower than those abroad. Because European and American chip companies are generally large listed companies, the gross profit threshold is high, then the price will be relatively high. Therefore, in overseas markets, domestic chip design companies should encounter fewer price challenges. But building a brand takes a long time, it is impossible to dominate the world by low price, and for more responsible customers, the price of the chip is not even the main concern.

In addition, domestic chip companies, in some market segments such as data center chips, can consider setting up some R & D centers in Europe and the United States to understand the cutting-edge trend.

YukawaMany domestic chip companies began to make frequent mergers and acquisitions at sea. Just take the power semiconductor industry as an example. CRRC once acquired a British company, and Wen Tai also planned to buy one. what do you think of this phenomenon?

Yu Zhihong：In Europe and the United States, semiconductors are not a very popular industry, but have been very mature, which provides a lot of investment targets for the domestic capital market. For the more chicken-rib business in the European and American markets, put into the domestic restructuring, it is possible to coruscate youth. This is a good shortcut for companies with strategic vision.

For example, Zhuzhou CRRC Times Electric's acquisition of British Dennis Semiconductor, I personally think, is a very great strategic decision. Dennis did not have a large business at that time and did not have any local market, but it did have the right technology. It can be used to drive electric trains. This kind of core power electronic devices are very scarce in China. After completing the acquisition, Zhuzhou CRRC Times Electric fully mastered, digested and absorbed the technology, and then set up a factory in China, magnifying a small factory that originally had only some design capacity and small-scale production capacity, and its influence was magnified many times.

There is also Si Lijie's acquisition of Mei Xin meter chip business, which is also an excellent acquisition case. For Mei Xin, its meter chip can not play a synergistic effect, but for Si Lijie, it can provide a solution for the entire meter chip.

Of course, I also see some other acquisitions, although the target value is very large, but the gross margin of the assets purchased is very poor, it is difficult to make a profit, there is no new R & D department, but the sales are relatively high, and such acquisitions will be relatively losses. of course, the money they spend is not necessarily their own.

Automotive chips and the third generation semiconductors

YukawaWhat do you think of the opportunities that the accelerated growth of the electric vehicle industry brings to the semiconductor industry?

Yu Zhihong：Many chip companies are telling the story of electric cars. To determine whether there is a real opportunity, ask two questions: how many of the company's chips will be used in a car, and how much can a chip cost?

For example, the new power of car-building electric vehicles, there must be a lot of large processors, CPU, GPU, FPGA and so on, this is the opportunity for digital chips. A larger processor can be worth tens of dollars or more.

The other is power devices, such as MOSFET and IGBT, which are new applications and are of course a major boon for related companies, because a car can sell at least tens to hundreds of dollars in revenue.

But for simpler chips, such as analog switches and sensors, a car may sell for a few dollars or less. Even if they can tell stories about electric cars, it doesn't really have much impact.

YukawaWhat is the current level of domestic IGBT chips?

Yu Zhihong：This problem is rather complicated. There are several points to consider for IGBT:

One is the performance of a single chip, which is an one-on-one competition.

The other is the ratio of performance to price. Many domestic IGBT production lines are basically 6 or 8 inches, and the cost is naturally more expensive, because Infineon and other large international factories have 12-inch factories that specialize in IGBT. Normally, except for factory investment, 12 inches is basically about half the cost of 6 or 8 inches. However, because European and American manufacturers have high requirements for gross profit, in fact, everyone is basically even.

So, the rest is a question of reliability, that is, whether customers dare to use domestic chips or not. Because the application of IGBT is generally industrial motors, railway locomotives, electric vehicles and other occasions with strict reliability requirements, it is impossible for the whole machine factory to support domestic chip companies like Lei Feng, but it depends on quality and reliability.

In terms of quality indicators such as consistency and reliability, domestic manufacturers still have a long way to go to establish internationally competitive brands. The gap between the specifications of many home chips and Infineon is very limited, but if more customers are sold, the problem begins to emerge.

Of course, if power semiconductors such as IGBT are developed for a period of time, the gap will get closer and closer. Because the industry has reached some physical limits, there is no way to improve it any better. So I am very optimistic about the domestic IGBT industry, if there are 10 years, I think the gap will become very small, the rest is to do the market to do the brand problem.

YukawaDo domestic automotive chip companies have advantages over foreign countries? For example, in the power semiconductor industry, in terms of cost, synchronous research and development, product definition and other dimensions?

Yu Zhihong：Very simply, if domestic car chips can be sold abroad, their advantages will really come out, but this has not been seen yet, and even consumer chips have not yet been sold abroad.

Take Tesla, Inc. as an example, I have not heard of any domestic power semiconductor manufacturers that can get close, that is, let Tesla, Inc. 's people take a look, no one has reached this state, and the share of European, American and Japanese manufacturers is basically 100%. In fact, Tesla, Inc. has a lot of Chinese engineers, but the problem is that there is no way to see China's automotive chip companies in the supply chain for the time being. In the localization projects of foreign cars in China, there are some cases of cutting into domestic programs, but few.

YukawaWhat is the practical application and promotion of the third generation semiconductors such as silicon carbide and gallium nitride in the market? How big is the gap between China and foreign countries? Is it the hope of overtaking in the corner of China's semiconductor industry?

Yu Zhihong：Silicon carbide and gallium nitride have the revolutionary potential to completely replace silicon devices in some market segments, but there is no way to replace existing solutions in most markets.

Relatively speaking, silicon carbide has greater potential and has proved its performance, reliability and quality in many applications, especially in electric vehicles, using demanding inverters. Even so, silicon carbide can only be used in high-voltage situations, but it is completely useless for most consumer products.

As far as gallium nitride is concerned, I have opened several gallium nitride controller products myself, which have a narrower range of applications. On the one hand, in 5G communication, in the power amplifier, it can replace the existing scheme. On the other hand, there is a lot of fast charging on mobile phones.

However, the entire market potential of gallium nitride is far overvalued. I suspect that many gallium nitride companies that are now good storytellers may have a thunderstorm in the next few years.

Because Europe and the United States have been doing gallium nitride for decades, no company has gone public, and so far no one has been able to balance profits, even if there are a lot of good people. The reason is that the gallium nitride market is too narrow to support so many listed companies. This is where we need to be vigilant.

The end.

Recently, the chip sector has once again become the darling of A shares. Driven by the rising price tide of chip shortage, people have turned their attention to those domestic chip companies that can solve the problem of "sticking neck". The capital market is talking about technology overcoming difficulties, and even optimistic about the industry.

However, through a live broadcast with Silicon Valley chip expert Yu Zhihong, we also observed another perspective of the chip industry: product management.

The growth of chip companies has never been a single technical problem, but complex, requiring multiple dimensions of the deficiency.

Price and performance, time and technology, team management and coordination, a good chip, not only needs the results of chip technology, but also the results of project management.

Just like Internet companies, chip companies also need product managers to make a deep insight into the market, play a key role of co-ordination and coordination, and continue to launch marketable chip products through wise business strategies.

So when China's big Internet companies enter the chip industry, can they replicate past successful experiences and apply complex and rapid project management experience to the chip industry? This may be a question that we have to think about.

But our revelation is that chips are not only a technical problem, but also a commercial problem.