This article gives the most detailed information about chips!

Source: The content comes from "The Elementary School Monk at the Foot of Laohe Mountain", thank you.

In the contest between China and "foreign countries", which country has the upper hand? Some say that China beats foreign countries, while others say that foreign countries easily beat China to the ground. Both sides cited a variety of examples, which made us, the spectators, confused. Of course, the centrists will definitely say that both countries have their own advantages and disadvantages, but this conclusion is correct but not nutritious. If you want to be knowledgeable on the topic of China and foreign countries, you must first understand what technology is?

If we classify technologies, the first category can be called "copycat technology" or "pure money-burning technology". Some people like to burn money on the left side, while others like to burn money on the right side, thus burning out different application technologies. This is essentially the integration of old technologies into new things, such as the Saturn V that the United States used to land on the moon, the cross-sea bridge built by Geotechnical Engineering, the Maus tank built by Moustache, and even the Great Wall of China and the Egyptian pyramids. For example, this is a bit like the Guinness record: the longest hair, the longest nails, etc. As long as the money is in place, anyone can burn out these things. The key is whether there is demand, so these can also be called application technologies.

For example, this kind of bridge-building machine can be made by several industrial powers, but they can only be used as toys. Only earthworkers can make money by making it.

After the development of geotechnical engineering, massive demands emerged, which promoted the explosion of various money-burning application technologies. The money earned can be used to tirelessly improve various details. Therefore, it is no exaggeration to say that China's application technology has become on par with the entire foreign world.

The second type of technology is temporarily called "non-copycat technology" or "time- and money-burning technology". If you try to disassemble any awesome equipment in detail, you will eventually find that it is all material technology.

Making materials is similar to cooking. I can tell you the ingredients of fried tomato and eggs, but your dish is not as delicious as mine. This is the core technology. In addition to biomedicine, the core technology is material technology. Let's look at a series of examples:

Engines, the jewel in the crown of industry, are the most criticized shortcoming of geotechnical engineering. To put it bluntly, the core technology is that the turbine blades are not strong enough. If you step on the accelerator too hard, they will fall apart. Whether it is aerospace engines, aviation engines, or gas turbines, as long as they have the word "machine" in them, geotechnical engineering is a bit weak (see the previous article "The Tragedy of Materials: The Difficult Birth of Chinese Hearts" ).

In addition to burning money and time, material technology sometimes requires a bit of luck. Let's take the engine as an example: metal rhenium, this stuff mixed with nickel, can make amazing turbine blades. The global proven reserves of rhenium are about 2,500 tons, mainly distributed in Europe and the United States, and 70% are used to make engine turbine blades. This strategic material is definitely banned by the United States. A few years ago, a rhenium mine with a reserve of 176 tons was discovered in Shaanxi. Geotech was so happy that he immediately spent a lot of money. His hard life has only improved in the past few years.

Rare earth permanent magnets are magnets made of rare earths. They can maintain their magnetism all the time and are very useful. Most of the high-grade rare earth ores are distributed in China, so China is more capable than the United States in technologies related to "magnetism", such as nuclear fusion and space dark matter detection. It is said that China also imposed an embargo on the United States a few years ago, forcing the United States to exchange rhenium, and with the little rhenium dug out by Shaanxi and Anhui, the J20 engine has some prospects.

As the high-end machine tool of the "mother of industry", geotechnical engineering is basically at the same level as the men's national football team, and can only look up to Japan, Germany and Switzerland. Materials are one of the biggest limitations. For example, during high-speed processing, the friction between the spindle and the bearing produces thermal deformation, which causes the spindle to lift and tilt, and the tool wear, etc. Therefore, geotechnical engineering can only look at the "foreign" work with extremely high processing precision.

Optical crystals and some of Geotech's products can also be embargoed against the United States, so the technologies related to light are not weak, such as laser weapons and quantum communications. Aerodynamic shape, thanks to the accumulation of Qian Xuesen's generation, the technology related to it is also very good.

If we continue to list, we will find that in terms of basic materials with broad applications, China still lags behind foreign countries, but in terms of relatively narrow application fields, China is gradually taking the lead.

Sit up straight, little friends, here comes the point! There are about 130 kinds of this key core material in the world, which means that as long as you have these 130 materials, you can assemble any existing equipment in the world, and then produce anything that already exists.

To some extent, the core technology of mankind refers to these 130 materials, of which 32% are completely blank in China, and 52% rely on imports. The proportion is even more disparate in cutting-edge fields such as high-end machine tools, rockets, large aircraft, and engines. Although parts are made in China, 95% of the equipment for producing parts relies on imports. These are not old news, but data released by the Ministry of Industry and Information Technology in July 2018, which is still fresh.

In terms of core material technology, it is not an exaggeration to say that "foreign countries are still holding China down". This is actually easy to understand, after all, China has only been in business for a short time, and material technology requires not only money but also time.

I have to emphasize here that applied technology is not less important than core technology. It requires a combination of funds, demand and social conditions. Although foreign countries have the ability to spend money, they may never have the opportunity to spend money in their lifetime. Someone must be arguing here: they just don’t want to spend money, otherwise they will kill you in a minute! Haha, if you force yourself to spend money, the consequences will be similar to those of the Russians.

After all this fuss, it's time to get back to the point. The reason why semiconductor chips are difficult is that they involve not only a large amount of application technology that costs a lot of money, but also a lot of material technology that costs a lot of money and time. In order to make it easier for children to understand, we have to start from the principle.

Many illiterate people think that quantum mechanics is just a mathematical game and has no application value. Haha, let's find an ancestor for computer chips. Please see the demonstration:

We can understand conductors and insulators. But the little ones who are confused by physics for the first time may be semiconductors. So I will first pay back the debt to your physics teachers.

When atoms form solids, many identical electrons are mixed together, but quantum mechanics believes that two identical electrons cannot stay in the same orbit. Therefore, in order to prevent these electrons from fighting in the same orbit, many orbits are split into several orbits. So many orbits are squeezed together and accidentally get close to each other, forming a wide orbit. This wide orbit formed by many thin orbits squeezed together is called an energy band.

Some wide orbits are packed with electrons, so they cannot move, while some wide orbits are empty, so electrons can move freely. The ability of electrons to move is manifested as conductivity on a macroscopic scale, while conversely, if electrons cannot move, they cannot conduct electricity.

Okay, let’s keep things simple, not mention the concepts of “valence band, full band, forbidden band, conduction band”, and get ready to focus on the key points!

Some full orbitals and empty orbitals are so close that electrons can move from full orbitals to empty orbitals without any effort, and thus can move freely. This is a conductor. The principle of conduction of monovalent metals is slightly different.

But often there is a gap between two wide tracks, and electrons cannot cross it by themselves, so it does not conduct electricity. But if the width of the gap is within 5ev, adding extra energy to the electrons can also cross to the empty track, and they can move freely after crossing, which is conductive. This solid with a gap width of no more than 5ev can sometimes conduct electricity and sometimes not, so it is called a semiconductor.

If the gap is more than 5ev, it is basically over. Under normal circumstances, electrons cannot cross it. This is an insulator. Of course, if the energy is large enough, let alone a 5ev gap, even a 50ev gap can still pass through, such as high voltage electricity breaking through the air.

At this point, the band theory developed from quantum mechanics has almost taken shape. The band theory systematically explains the essential differences between conductors, insulators and semiconductors, that is, it depends on the gap between full orbits and empty orbits, or in academic terms, it depends on the band gap width between the valence band and the conduction band.

Semiconductors are still far from the principles of chips, so don't be in a hurry.

Obviously, there is nothing to do with a straight man like a conductor, so the conductive wire is still copper wire to this day, with no technological progress, and the fate of insulators is similar.

The ambiguous nature of semiconductors is the easiest to cause trouble, so industries related to electronic equipment basically belong to the semiconductor industry, such as chips and radars.

The following is a bit brain-burning.

Scientists use silicon as the base material for semiconductors for some simple reasons. Silicon has 4 electrons in its outer shell. If a solid is made up of 100 silicon atoms, then its full orbit is filled with 400 electrons. At this time, 10 silicon atoms are replaced with 10 boron atoms. A trivalent element like boron has only 3 electrons in its outer shell, so the full orbit of this solid has 10 vacancies. This is equivalent to freeing up a few empty seats on a crowded bus, providing conditions for the movement of electrons. This is called a P-type semiconductor.

Similarly, if 10 phosphorus atoms replace 10 silicon atoms, a pentavalent element like phosphorus has 5 electrons in its outer layer, so there are 10 more electrons in the full orbit. This is equivalent to 10 more people hanging outside a crowded bus, and these people can easily get off the bus. This is called an N-type semiconductor.

What will happen if we put these two semiconductors face to face? It is obvious that the extra electrons of the N-type will move to the vacancies of the P-type until the electric field is balanced. This is the famous "PN junction". (The animated picture comes from Zhang Yun's blog post on ScienceNet)

If a forward voltage is applied at this time, the extra electrons in the N-type semiconductor will continuously run to the vacancies in the P-type semiconductor. The movement of electrons is electric current, and the PN junction is conductive at this time.

What if a reverse voltage is applied? Electrons are drawn from the P-type semiconductor to the N-type semiconductor, but the N-type is already full of extra electrons. The extra electrons continue to enhance the electric field until the applied voltage is offset, and the electrons stop moving. At this time, the PN junction is non-conductive.

Of course, there is still a slight movement of electrons in reality, but it is negligible compared to the forward current.

If you are confused, it’s okay. To sum it up in plain words: PN junction has unidirectional conductivity.

OK, now we have a unidirectional PN junction, what next? Connect the two ends of the PN junction with wires, and it becomes a diode:

With a diode, build a circuit:

The triangle represents a diode, and the direction of the arrow indicates the direction in which the current can flow. AB is the input terminal, and F is the output terminal. If no voltage is applied to A, the current will flow out along the line of A, and there will be no voltage at the F terminal; if voltage is applied to AB at the same time, the current will be blocked at the other end of the diode, and there will be voltage at the F terminal. Assuming that voltage is regarded as 1 and no voltage is regarded as 0, then only when 1 is input from AB terminals at the same time, the F terminal will output 1. This is the "AND gate circuit".

Similarly, if we change the circuit to this, as long as one of AB inputs 1, the F terminal will output 1. This is called an "OR gate circuit":

Now that we have these basic logic gate circuits, we are not far from the chip. You can design a circuit whose function is to transform a string of 1s and 0s into another string of 1s and 0s.

To give a simple example, applying voltage to the second and fourth input terminals is equivalent to outputting 0101. After passing through a specific circuit, the output terminal can become 1010, that is, there is voltage at the first and third output terminals.

Let's play a slightly more complicated game:

There are 8 input terminals on the left and 7 output terminals on the right, each of which corresponds to a light-emitting diode. A string of signals is input from the left: 00000101, and after passing through a bunch of circuits in the middle, another string of signals is output on the right: 1011011. 1 represents voltage, and 0 represents no voltage. Voltage can light up the corresponding light-emitting diode, that is, 5 of the 7 light-emitting diodes are lit, so a number "5" is obtained, as shown in the figure above.

Finally, we have figured out how numbers are displayed! If you want to add 1+1, the complexity of the circuit is beyond the IQ of 99% of people. Even if I design the circuit myself, the computing power of the circuit is not as good as an abacus.

Until one day, someone used 18,000 electron tubes, 6,000 switches, 7,000 resistors, 10,000 capacitors, and 500,000 wires to form a super complex circuit, and the first computer was born. It weighed 30 tons and had a computing power of 5,000 times per second, which was less than one-tenth of the current handheld calculators. I don’t know how many times the engineers at that time had brain cramps in order to install this pile of circuits.

The next idea is simple: how to integrate these 30 tons of things into a place as small as a fingernail? This is the chip.

In order to shrink the 30-ton computing circuit, engineers threw away all the extra stuff and made PN junctions and circuits directly on silicon wafers. Let's start with silicon wafers and talk about the chip manufacturing process and China's level.

First: Silicon

If you chlorinate this thing and then distill it, you can get very pure silicon, which can be cut into slices to get the silicon wafers we want. The evaluation index of silicon is purity. If there are a lot of impurities in silicon, then the electrons can't run smoothly between full orbits and empty orbits.

Solar-grade high-purity silicon requires 99.9999%, and more than half of the world's silicon is made in China, so it has long been sold at bargain prices. Electronic-grade high-purity silicon used in chips requires 99.9999999999% (don't count, 11 nines), and is almost entirely imported. It was not until 2018 that Xinhua Company in Jiangsu achieved mass production, and currently produces 5,000 tons per year, while China imports 150,000 tons a year.

What is rare is that Xinhua's high-purity silicon is exported to South Korea, a semiconductor powerhouse, so the quality should be pretty good. However, 30% of the manufacturing equipment still has to be imported...

The traditional leaders in high-purity silicon are still Germany's Wacker and the United States' Hemlock (a joint venture between the United States and Japan), and China still has a long way to go.

Second: Wafer

When silicon is purified, it needs to be rotated, and the finished product looks like this:

So the sliced ​​silicon wafer is also round, so it is called a "wafer". Does this word sound familiar to you?

After the wafer is cut, thousands of circuits will be assembled on the wafer. This is done in a factory called a wafer fab. Think about it, with current human technology, how can this be accomplished?

Using atomic manipulation? You're overthinking it, my friend! By the time you master the art of flying a sword, humans may not be able to manipulate atoms one by one to form various devices. The wafer processing process is a bit cumbersome.

First, a layer of photosensitive material is applied to the wafer. This material melts when exposed to light. Where does the light come from? The photolithography machine can use very precise light to carve a pattern on the photosensitive material, exposing the wafer underneath. Then, it is washed with something like plasma, and many grooves are carved into the exposed wafer. This set of equipment is called an etcher. When phosphorus is added to the grooves, a bunch of N-type semiconductors are obtained.

After completion, clean it, re-apply photosensitive material, carve the pattern with a photolithography machine, carve the grooves with an etcher, and then sprinkle boron on it to get a P-type semiconductor.

The actual process is more complicated, but that's the general principle. It's a bit like 3D printing, where you put the wires and other components in layer by layer.

The small squares on this wafer are chips. When you magnify the chips, you will see piles of circuits. These circuits are no more sophisticated than those in the 30-ton computer. The most basic layers are simple gate circuits. It’s just that more devices are used to form a larger circuit, so the computing performance is naturally improved.

It is said that this is a NAND gate circuit:

Here's a question: Why not make the chip bigger? Wouldn't that allow for more circuits to be installed? Wouldn't the performance be comparable to foreign products?

This question is very interesting, and the answer is surprisingly simple: money! A 300mm diameter wafer can produce 100 chips with a 16nm process, and 210 chips with a 10nm process. Therefore, the price is half as cheap as before, which can suppress competitors in the market. The money earned can be used for more research and development, and the gap is thus widened.

As an aside, China's military chips have basically achieved self-sufficiency, because we don't care about money! We can make the chips bigger. In addition, the larger the silicon wafer, the greater the probability of encountering impurities, so the larger the chip, the lower the yield rate. In general, the cost of large chips is much higher than that of small chips, but for the military, this is not a big deal.

Don't confuse "Loongson" with "Hanxin"

Third: Design and Manufacturing

The thought of using hundreds of millions of devices to form such a huge circuit is mind-boggling, so chip design is extremely important, so important that it is on par with material technology.

If the traffic lights at an intersection are not set properly, it may cause a huge traffic jam. If there is a problem with the PN junction, the electrons will also cause a traffic jam. There is only one way to test this kind of sophisticated circuit design, that is: use it! Use it in large quantities! Now you know the importance of chip cost, because you will not spend more money to buy a computer with the same performance, and chip companies will easily fall into a vicious circle if they lose their market share.

Because of this, chip design not only costs money, but also takes time to develop. It is a core technology that costs money and time. Since it is a core technology, it will naturally develop into an independent company. Therefore, there are three types of chip companies: those that do both design and manufacturing, those that only do design, and those that only do manufacturing.

Semiconductors are one of the few technologies in which Taiwan is still ahead of mainland China. Due to the de facto division between the two sides of the Taiwan Strait, mainland China and Taiwan will be discussed separately for the time being.

In the early days, design and manufacturing were all done together. The most famous ones are: Intel in the United States, Samsung in South Korea, Toshiba in Japan, STMicroelectronics in Italy and France; in mainland China: China Resources Microelectronics, Silan Microelectronics; in Taiwan, China: Macronix Electronics, etc.

Among foreign countries, Taiwan and the mainland, the most backward is the mainland. Its products are mostly concentrated in low-end fields such as home appliance remote controls, and high-end chips such as mobile phones and computers are almost blank!

Later, as chips became more and more complex, design and manufacturing were separated, and some companies only designed and became pure chip design companies, such as Qualcomm, Broadcom, and AMD in the United States, MediaTek in Taiwan, and Huawei HiSilicon and Spreadtrum in mainland China.

A few comments on each one.

I won’t say much about the famous Qualcomm, which is used in half of the world’s mobile phones. Broadcom is the chip supplier for Apple phones, and it is no surprise that it ranks second in mobile phone chips. AMD and Intel have basically monopolized the computer chip market. These are all American companies, and their dominance in the world is not exaggerated.

Taiwan's MediaTek follows the low-end route, and its mobile phone chip market share ranks third. Many domestic mobile phones use it, such as Xiaomi, OPPO, and Meizu. However, it has been hit hard by Qualcomm recently, and its sales have been falling.

Huawei HiSilicon is the most competitive. I'm sure you've read a lot of stories about it, so I won't go into details. In addition to communication chips, HiSilicon also makes Kirin chips for mobile phones. Its market share has ranked in the top five as Huawei's mobile phones grow. From my personal experience, HiSilicon's chip progress is really quite good (this wave of advertising does not charge Huawei a penny).

Spreadtrum is a school-run enterprise of Tsinghua University, and one of the earliest chip companies in mainland China. After all, it cannot be shaved bald, so it has to bite the bullet and take the low-end route. There were many crises some time ago, and then it was said to be the beginning of a change. It has not been easy, and it is far behind the world's giants.

There are also a number of chip design companies in the mainland, such as MStar Semiconductor, Novatek Technology, Realtek Semiconductor, etc., which are all subsidiaries of Taiwan's big brother. Their products are used in televisions, portable electronic products and other fields, and they are quite profitable.

Among the chip design companies in mainland China, Taiwan holds up most of the market!

There is also a type of foundry that only manufactures but does not design. We must first talk about Taiwan's TSMC. It was the emergence of TSMC that separated chip design and manufacturing. In 2017, TSMC accounted for 56% of the world's foundry business, ranking first in the world in terms of scale and technology. Its market value even exceeded Intel, making it the world's number one semiconductor company.

The wafer foundry industry is dominated by big brother Taiwan. In addition to the giant TSMC, Taiwan also has UMC, Powerchip Semiconductor, etc. Even the United States and South Korea have to stand aside.

The largest foundry in the mainland is SMIC, and Shanghai Huali Microelectronics is also not bad, but their technology and scale are far behind Taiwan. However, constrained by the treacherous social situation in Taiwan, TSMC began to expand into the mainland and settled in Nanjing. In recent years, Taiwanese and foreign companies have been frantically building wafer foundries in the mainland, which is comparable to the joint venture automobile industry in the past.

China's SMIC has 28nm technology and 14nm production lines are on the way, but unfortunately they are not profitable yet. People are still willing to hand over this job to TSMC, which has almost won 70% of the world's foundry business below 28nm.

The United States, South Korea, and Taiwan already have 10nm processing capabilities. In recent months, TSMC has just launched a 7nm process, which has surpassed Samsung. The first batch of customers is Huawei's Kirin 980 chip. These two buddies have been partners for a long time. Huawei designs chips and TSMC processes chips.

To be honest, if the mainland can integrate Taiwan's semiconductor industry and use flexible policies and a huge market to promote its further upgrading, it will be at least half as easy to catch up with the United States. Now, the mainland has a long way to go!

Fourth: Core equipment

The chip yield rate depends on the overall level of the wafer fab, but the processing accuracy depends entirely on the core equipment, which is the "photolithography machine" mentioned earlier.

As for lithography machines, ASML of the Netherlands is the leader! Sorry, the production is not high yet, just wait and see! Whether it is TSMC, Samsung, or Intel, whoever buys ASML's lithography machine first will be the first to have 7nm technology. There is no other way, it is just so powerful!

Japan's Nikon and Canon also make lithography machines, but their technology is far inferior to ASML. In recent years, they have been beaten so badly by ASML that they can only grab market share in the low-end market.

ASML is the only manufacturer of high-end lithography machines, each of which costs at least US$100 million. Only 12 machines were produced in 2017, and 24 machines were expected to be produced in 2018, all of which have been snapped up by TSMC, Samsung, and Intel. 40 machines are predicted for 2019, one of which will be for our SMIC.

Since it is so important, can't we pay a little more? First, Intel has 15% of ASML, TSMC has 5%, and Samsung has 3%. Sometimes, money is not everything. Second, the United States has made the Wassenaar Arrangement, which prohibits the sale of sensitive technologies. China, North Korea, Iran, and Libya are all restricted countries.

Interestingly, in 2009, Shanghai Micro Electronics successfully developed a 90nm lithography machine (with core components imported). In 2010, the United States allowed the sale of equipment above 90nm to China. Later, China began to work on 65nm lithography machines. In 2015, the United States allowed the sale of equipment above 65nm to China. Later, the United States began to lose control of its younger brother, and SMIC had the opportunity to grab a high-end machine.

But we don’t need to be discouraged. Any of our real estate companies can easily surpass ASML in sales. Yeah!

China's etching machines are second only to photolithography machines in importance. The situation is much better in China. The 16nm etching machines are already in mass production, and the 7-10nm etching machines are also on the way. Therefore, the United States was considerate enough to lift the blockade on China's etching machines.

To implant elements such as boron and phosphorus on the wafer, an "ion implanter" is needed. In August 2017, the first domestically produced commercial machine was finally produced, and the quality is not mentioned for now. 70% of the market share of ion implanters is held by Applied Materials of the United States. To coat photosensitive materials, a "coating and developing machine" is needed, and Tokyo Electron of Japan has taken 90% of the market share. Even auxiliary materials such as photoresists are almost monopolized by Shin-Etsu of Japan and Dow of the United States.

From 2015 to 2020, the domestic semiconductor industry plans to invest US$65 billion, of which US$50 billion will be invested in equipment, and US$48 billion will be used to purchase imported equipment.

All in all, China has invested an average of 13 billion yuan per year in recent years, and Intel alone has invested more than 13 billion US dollars in research and development.

When it comes to semiconductor equipment, China has an extremely heavy responsibility and a long way to go!

Fifth: Closed Beta

After the chip is made, it has to be cut from the wafer, connected to wires, put into a casing, and tested. This is called packaging and testing.

The packaging and testing industry is once again dominated by Taiwan's big brother, ASE, which ranks first in the world, is followed by a group of powerful younger brothers: Siliconware Precision Industries, Powertech Semiconductor, Chip Motion, Shinbon, and King Yuan Electronics.

The three major packaging and testing giants in mainland China, Changdian Technology, Huatian Technology, and Tongfu Microelectronics, are all doing well. After all, they are only at the end of the chip industry and the technical content is not high.

When talking about Chinese chips, we have to mention the "Hanxin Incident". In 2003, Professor Chen Jin, Dean of the School of Microelectronics at Shanghai Jiao Tong University, bought chips from the United States, erased the original markings, and presented them as independent research and development results, defrauding countless funds and honors, consuming a large amount of social resources, and the impact was unprecedented! For a long time, the scientific research circle was terrified of talking about chips, which seriously interfered with the normal development of the chip industry.

In most areas of silicon raw materials, chip design, wafer processing, packaging and testing, and related semiconductor equipment, China is still in a state of "heavy tasks and a long way to go". How long will this state of confusion last? According to the theory of "burning money and burning time", it is about 2030! The "Outline for the Development of the Integrated Circuit Industry" issued by the State Council clearly states that by 2030, the main links of the integrated circuit industry chain will reach the international advanced level, a group of enterprises will enter the international first echelon, and the industry will achieve leapfrog development.

At present, the overall level of China's chips is almost at the stage of just achieving zero breakthrough. Although the market share is negligible, it has a presence in every field and the prospects are still promising.

At the end of the article, I habitually complain about the immaturity of human technology. Chips, as the highest level of technology that everyone can reach with their brains, the band theory that is the basis of them is actually only an approximate theory, and the behavior of electrons still cannot be accurately calculated. To put it in a broader sense, although the current technology is complicated, it is actually just playing with electrons. As for the other hundreds of particles, we still don’t know how to play with them!

The chip processing accuracy has reached 7nm. Although Samsung boasts that it will reach 3nm, so what? Can you continue to do it? 1nm is just a few atoms. The quantum effect is very significant, and the approximate theory does not work. The behavior of electrons is even more difficult to predict. The semiconductor industry has to stop here.

Whether it is burning money or burning time, the end result is theoretical physics. In addition to burning money and time, basic science also burns people, and the burning is extremely tragic. 99 out of 100 people with high IQs are just stepping stones! Engineers can become monks later in life, but physicists must be trained in a professional field. Basic science has been neglected in China for more than 5,000 years, and now the enthusiasm for filling out the application every year is not as high as that of playing a trick.

We can't just play around with electrons anymore, in order to make use of neutrinos as well, we have to quickly fool people, oh, no, we should call on more children to learn basic science!

Today is the 1695th issue of content shared by "Semiconductor Industry Observer" for you, welcome to follow.

Moore Elite is a leading chip design accelerator that reconstructs semiconductor infrastructure to make it easier for China to make chips. Its main businesses include "chip design services, wafer packaging and testing services, talent services, and incubation services". It covers more than 1,500 chip design companies and 500,000 engineers in the semiconductor industry chain, and has precise big data on integrated circuits. It currently has 200 employees and is growing rapidly. It has branches and employees in Shanghai, Silicon Valley, Nanjing, Beijing, Shenzhen, Xi'an, Chengdu, Hefei, Guangzhou and other places.

Click to read the original article to learn more about Moore's elite