By 2035, China wants to develop, manufacture and market 35 percent of all semiconductors itself. For this purpose, the state leadership provided various funds with a lot of money. Markt&Technik asked Prof. Leo Lorenz, the President of ECPE (European Center for Power Electronics e.V.) who has lived and taught in China for several years, whether this is realistically possible.
Markt&Technik: Years ago you said that China would soon catch up with Western nations in the semiconductor sector and become self-sufficient. But that seems to be delayed. Why is that?
Leo Lorenz: Unfortunately, I haven't been to China since the beginning of the pandemic, but I am well supplied with information from there by former work colleagues and good friends with whom I am in permanent contact. I think two things happened in China.
Firstly, the technology transfer that is so important for China was abruptly stopped by the USA. The best-known example is Huawei. For many years, the USA was the preferred country for Chinese engineers to study and then take up development or research work in industry or at university. Many of them have developed into world-renowned experts there. As a result, very good connections and cooperation developed, especially in the research area of the universities and institutes in both countries. There was a regular and intensive exchange of scientists. As a result, university institutions were so well connected that research was carried out under the same direction both in the USA and in China.
The US government - beginning with the Trump administration - had rightly said that the results of research funded in the US must remain in the country and not be shared without further consent. Here in Europe, too, we have now become more sensitive to this when it comes to the subject of technology transfer. I think overall we have learned that while globalization is right and important, there are also limits depending on the political systems.
Chinese engineers are extremely good and quick at converting research results into products and manufacturing them in huge numbers for the world market. They are way ahead of us there, because the decision-making processes here take much more time. But when it comes to innovations, i.e. inventing or developing something completely new yourself, people in China still have a hard time with that.
Innovations mostly still come from Western countries, which have a long history of free research – i.e. independent, free thinking and critical questioning starting with university education – have learned. Based on these facts, there is now a pause in "loose technology transfer," and this will significantly affect the speed at which the Chinese government wants the semiconductor industry to develop.
What is the second factor?
The second key point is the pandemic. That's why entire industries and entire economic regions such as Shanghai, Shenzhen, Wuhan and so on had to shut down completely. These massive lockdowns have left their mark. Many people in the country now doubt whether the government's zero-Covid strategy is really good and right because it leads to abrupt disruptions in various supply chains.
For these two reasons, China will no longer achieve the goal it has set itself of developing, manufacturing and marketing around 35 percent of all semiconductors worldwide in the country by 2035. The government is aware of this.
And then there is probably a mentality problem in China, right?
Right. We absorb the topics of reliability and robustness with our mother's milk. In China, engineers are not as sensitive to these issues. In the ECPE we have two working groups that focus on the reliability and robustness of power semiconductor components - one for automotive, the other for trains. And soon there will be another one for renewable energies. The service life plays a decisive role, especially for mobile applications in the transport sector, but also stationary applications in the energy supply with often very different requirements in terms of operating time, mission profiles, operational environment and so on. And even with the well-known silicon components, we find that we have certain reliability data - and their associated measurement, Testing and quality techniques - have to rethink and supplement. And that with silicon, which we have known for 70 years! Not to mention the new materials silicon carbide and gallium nitride.
When it comes to cell structures for power semiconductors, the Chinese manufacturers are quite good, money is not an issue there. Here, on the other hand, you have to fight for every euro of budget or funding. But it will be some time before the issues of reliability and robustness in the base are understood and the engineers there perceive this as an important point.
At the Markt&Technik panel discussion at PCIM Europe 2022, representatives from STMicroelectronics and onsemi said in unison that it was perfectly fine to manufacture standard products in China, but they wanted to keep the latest component technologies here in the West.
That's a good thing. Chinese engineers are very interested in technology, have a very good university education and learn quickly. They are good at copying, as we see again and again at many technology fairs in Europe and around the world. But when it comes to pioneering work, drastic innovations, here in the West we are still superior in terms of mentality and upbringing. In China, people are more relaxed and accept system failures without much discussion when there are technological leaps. If, for example, an electric car breaks down, that's the way it is, and you drive on with another one. But if a vehicle from a German premium brand breaks down here, then all hell breaks loose.
And this difference in mentality is also reflected in semiconductor manufacturing. Here we have an extremely high yield with huge quantities and high quality. This is far from the case in China. It will stay that way for a long time.
Source https://www.elektroniknet.de/halbleiter/leistungshalbleiter/china-wird-das-eigene-ziel-nicht-erreichen.199176.html
