The current chip process is simple, it is very simple, that is, the silicon wafer engraving pattern, injecting conductive impurities, and forming a switch.

It is said that it is complicated, and it is also very complicated, if it is at the centimeter level, the pattern engraving and conductive impurity injection are fine, but at the nanometer level, these processes are very troublesome to complete.

However, Huang Haojie saw another processing method from the handling capacity of C31 fullerene.

That's right, it is to directly use the handling capacity of C31 to engrave lines, inject conductive impurities, and plating copper on the silicon wafer, and these processes can all be completed by C31.

However, the premise is that Huang Haojie completes the precise control of C31 fullerene.

Semiconductor Research Institute, Micro-Cryptography Laboratory.

More than a dozen researchers and engineers are debugging a huge instrument.

Atomic engraving machine made by the Semiconductor Research Institute.

Zheng Xinwang is a master's degree student in semiconductors who just graduated this year, and he never thought that he would come to work at Galaxy Technology.

Originally, he wanted to study for a doctorate, but there were unforeseen circumstances, and the doctoral supervisor he reported happened to have an accident, and Zheng Xinwang happened to meet Galaxy Technology recruiting, so he came to work at Galaxy Technology.

After coming to Galaxy Technology, Zheng Xinwang discovered that the semiconductor process that Galaxy Technology was researching had subverted all the current chip manufacturing processes.

He is using an electron microscope and a computer, carefully manipulating carbon nanofabrication needles and carving them on silicon wafers.

In order to make this carbon nanoneedle, it took them half a month to complete this five-nanometer carbon nanoneedle.

The circuit diagram of the chip is input into the computer, and then the computer calculates it with the electron microscope.

The C31 of the carbon nanoneedle is stimulated by specific conditions to combine with the silicon atoms on the silicon wafer, and then the silicon atoms are taken out of the silicon wafer, so that a groove is formed.

Countless grooves are connected together to form a groove circuit.

"Xinwang, how's it going?" A gray-haired middle-aged man walked over and asked.

"Zhao Gong, I'm adjusting some data, the five-nanometer level is too small, and about 43% of the circuit carving is inaccurate." Zheng Xinwang said helplessly.

Zhao Gong didn't show any disappointed expression.

"Everything is difficult at the beginning, keep working hard, I'll go and take a look."

Of course, Zheng Xinwang knew that this kind of thing could not be rushed, because once the first circuit chip was carved out.

Then the rest is very good.

On the other hand, Xingtu Company also held a lot of banknotes and directly merged three domestic semiconductor design companies in Dongtang.

The domestic chip manufacturing industry is not good, but the chip design industry is still so-so, such as Huawei's HiSilicon Semiconductor, which is the top semiconductor design company.

Huang Haojie, who didn't have time to grind, directly asked Wang Bosi to launch a silver offensive and use the merger method to give birth to Xingtu's chip design capabilities.

Du Wei and the others added the chip designers obtained from three semiconductor design companies, as well as a large number of 30~90 nanometer chip patents purchased by Zhang Rujing, and started the independent design plan.

Although these chip patents have been eliminated by others, it does not mean that they are all rubbish, after all, many things in chip design are common.

For example, the 5nm chip design can be used in a 50nm chip, and the same 50nm chip design can be used in a 5nm chip.

This may lead to a regression in chip performance and the original design, but the key is not in the chip design, but in the chip manufacturing.

After all, what if you can design a 1-nanometer chip, without the ability to make a 1-nanometer chip, everything is in a flash.

Just like Huawei's chips, your HiSilicon design capabilities are very powerful, but the problem is that you don't have chip manufacturing capabilities yourself.

Once Taiji Electric has any winter melon and tofu, Huawei Company will have a chip shortage or supply interruption, which is not to finish the calf.

Similarly, Zhongxin Communications Company was disgraced by Mi Lijia in 2016, and the billions of fines were equivalent to two years of profits of Zhongxin Communications Company, so it was in vain.

Huang Haojie would rather study it slowly by himself than easily hand over the production of core components to crooked nuts.

Like Sanxin, the company likes to do this kind of thing the most, whether it is the mobile phone screen and memory.

DRAM, or Dynamic Random Access Memory, is our most common system memory.

Whether it's a mobile phone, a computer, or any smart device, memory (e.g., memory sticks) is made up of single or multiple DRAM particles.

According to 2016 data, the global DRAM market size is about 41 billion US dollars, of which Sanxin, SK hynix, and Micron have shown a three-point posture, with the three accounting for more than 95% of the market share, and the industry has oligopoly characteristics.

Sanxin has absolute advantages in technology, production capacity and market share.

After the first mass production of 20nm DRAM process in 2014, it once again led and mass-produced 18nm DRAM. SK hynix mass-produced 18nm DRAM in early 2017. Micron is still on the road of 20nm, and is expected to complete the migration to 1Xnm products in the second half of 2017, with a technology gap of about 1 year with Sanxin, and the monopoly of the industry has a tendency to increase.

If a fire burns down Sanxin's factory, it means that memory production will plummet by 57%, and supply and demand will be greatly imbalanced, so the world's memory will inevitably rise in price.

So these guys' factories will always catch fire or some kind of chemical pollution for some reason, and then the price of memory modules will skyrocket.

Lao Ma once said: once there is a proper profit capital, be bold, if there is a 10% profit, it is guaranteed to be used everywhere; With a 20% profit, it becomes active; With a 50% profit, it takes risks; For the sake of 100% profit, it dares to trample on all human laws; With a 300% profit, it dares to commit any crime and even risk going to the gallows.

This is the essence of capital.

So cooperate with these guys, it's better to be prepared for the back road, otherwise you may be killed by them.

Xingtu Design Company has preliminarily designed a 5nm-level chip, of course, because it is a trial design, so they directly copied some of Qualcomm, Intel, and Huawei's chip designs.

Anyway, this design is just to experiment with the performance of the atomic imprinting machine, and Galaxy Semiconductor does not plan to produce it for sale, so there is no problem of patent infringement.

This 5nm-level chip consists of a total of 32 layers of circuitry.

Therefore, the micro-engraving laboratory needs to carve out 32 layers of circuits, but considering that the extraction of a little silicon atom will be very slow, which is not conducive to large-scale production.

Therefore, Huang Haojie found another way, his solution is very simple, separate the 32 layers of circuits, carve a silicon wafer in one layer, and then make nanoneedles on the grooves of these circuits to complete the production of these nanoneedles.

It is possible to directly make circuit printing machines, and these circuit silicon wafers with nano needles can be carved out of a layer of circuits at one time.

A silicon wafer only needs to go through 32 circuit printings to complete 32 layers of circuits.

Of course, in addition to circuit printing machines, the semiconductor research institute also needs equipment such as circuit injection machines and circuit copper plating machines.

These devices are based on C31 fullerene, so they are also called carbon nanocircuit production lines.

The first step in this method is to accurately complete the original film, which is the key point in the production of the chip.

For example, the 5-nanometer chip designed by Xingtu Company needs 32 original chips, and these 32 original chips are used as the basis to make 32 circuit printing heads, 64 circuit injection heads, and several circuit copper plating machines.

There must be no errors in these original films, otherwise there will be problems in the production of chips, or even scrapped.

And once the original footage is made, the rest of the work is pretty straightforward.

In this case, it is equivalent to using the laboratory method to make a high-precision chip, and then holding this high-precision chip and directly copying it, and the two are exactly the same.

All atomic stamping machines are the key equipment of Galaxy Technology, and C31 fullerene is the top priority.