However, the semiconductor industry is not strictly separated by region, especially in recent years of capital globalization, in many different countries, semiconductor enterprises have actually been closely linked through investment, equity and other forms.

This situation, which has long been developed by the market, is a silent reminder that it is almost impossible to live out of the globalized industrial system to survive in the semiconductor industry. (there are some reasons for this below)

Straight speed race - "Moore's law"

The magic law was put forward by Intel co - founder Gordon Moore (Gordon Moore) in 1965, and then a revision in 1975, the content - the chip will double every 18 months.

How to look at the semiconductor industry in the "Buddha Department"

It is such a seemingly "simple" law that seems to have no arguments, but it has been verified in the next 40 years of the development of the semiconductor industry. The key is that it is not an idea, but a theoretical formula based on the development trend of early semiconductor technology and the specific circumstances of commercialization.

What it reveals is the most important charm of semiconductor technology - more than 30 years of technical space to stabilize the index level, and the overall range has reached 1 million times. (in 2016, CPU could accommodate 10 billion transistors, and in 1976, it could only accommodate 10 thousand transistors).

How to look at the semiconductor industry in the "Buddha Department"

Interestingly, after Moore gave the first law, an American venture capitalist Arthur Rock, which was so close to the semiconductor industry, gave another law, which was also subjected to the "time test", also known as "Rock's law", or "the second law of Moore" - the cost of the wafer plant (Fab), every 4. The year will double.

In 1986, when Intel launched the 80386 processor, the cost of building a wafer fab was roughly $200 million. In just 10 years, when Intel launched the "Pentium" processor, the cost of construction of wafer fabs has risen to $2 billion. And according to the industry's public information, the figure has reached $15 billion in 2015, and this number is likely to be close to $30 billion in 3 years. This is only the cost of a wafer fab, and the technical cost behind the wafer fab is several times that figure.

The backwardness of China is because of missing the "window period"

How to look at the semiconductor industry in the "Buddha Department"

The trend of such a large circle of semiconductors is introduced, and it is easy to find the backwardness of the development in China. Because of various historical reasons, China has missed the final "window period" of the semiconductor industry. Coupled with the above, we have explained the "terrible" Moore's law of semiconductor, making the gap between us and advanced international semiconductor technology constantly widening.

How to look at the semiconductor industry in the "Buddha Department"

Foreign semiconductors are already in full swing. People who have seen computers in China are not many.

In the 90s of the last century, the foreign semiconductor industry was booming and personal computers were widely used, but according to the statistics of the world bank, only 4 per 10000 people in China had personal computers in 1990. Even 24 years later, in 2004, there were only 409 people in every 10000 people in China. The average world level was 1298, and the United States was 7492.

For example, in 2002, when the Chinese finally came up with the "dragon core" 1, which was called the first gun of its own chip, Intel had begun to produce the Pentium 4 series of processors (the seventh generation Intel processor). Let alone the strong alliance of "Wintel" (Microsoft + Intel), which provides the whole ecosystem for X86 architecture.

In the core area of semiconductor manufacturing and general-purpose processor, slow pace is hard to catch up with.

On the contrary, there are many fields in the semiconductor industry that are relatively low in growth, or their own products have been relatively mature, the competitiveness of domestic enterprises is very strong, and many enterprises have already defeated other international rivals in many small fields. In fact, there is nothing to be worried about.