Chapter 153 Project 712 (4k)
Topological semimetals can be used in the chip field. They are hailed as the most likely to become the next generation chip raw material and have been promoted by academic circles.
As an expert in the field of topological semimetals, Lai Jiachen counted roughly ten years from understanding this field to now engaging in research in this field. In these ten years, he has often heard similar things speech.
But the dream is beautiful, but the reality is cruel. Researchers have not been able to develop a topological semi-metallic film until this year. From the laboratory to industrialization, I don’t know how many years it will take.
So in his opinion, the application of this thing in the chip field is more or less the same as controllable nuclear fusion. Twenty years ago, it was said that it would take 50 years, and 20 years later, it will still be said that it will take 50 years.
I don’t know how many pitfalls there are in the middle waiting for people to step on.
And now he sees the fact that the biggest problem in the application of topological semimetal in the chip field has been solved, the contradiction between light sensitivity and stability. Solving this problem means that its original progress is impossible. Predictive, you can't estimate how long it will take to make it happen.
But now you can make a plan, and you can sort it out according to the plan.
After Lai Jiachen got the answer he wanted from the other party, he sighed in his heart: People are really different. Problems that are difficult to solve for the outside world are just easy for geniuses. have to.
After the discovery of room-temperature superconductivity, Chen Yuanguang disappeared after accepting an exclusive interview with Yang Ma. Although he did not make many public appearances before, he still participated in some activities.
This time, including the opening of the Human Hibernation Center of Guangjia Technology, the opening of room-temperature superconductors, the launch of Guangjia Aerospace’s new recyclable rocket, the completion and acceptance of the Huaguo Compound Eye, etc., Chen Yuanguang did not attend even once.
The external network said that Chen Yuanguang was under house arrest, but domestically, in conjunction with Joseph's assassination, they said that he was under protection.
As a professor at Shenhai Jiaotong University, Lai Jiachen had his own source of information. He knew that Chen Yuanguang had been staying in western China during this period, but he didn't know whether it was the capital of Shu or where.
After seeing the topological semi-metal this time, his first reaction was that no one except Chen Yuanguang could make this thing. After he asked, the other party kept silent.
Now that the suspicion has been confirmed, Lai Jiachen doesn't even have a sour mentality. We are not from the same world, so we can't be sour at all.
He recalled a BBC report he saw not long ago, which praised Chen Yuanguang as the biggest X factor between China and Amerikan.
He felt that this statement was an exaggeration at the time, but now he feels that this statement is still a bit conservative.
“So, using topological semi-metal as a raw material can make the chip have stronger computing power?” After meeting with Zhao Pu, Lin Jia asked, she still didn’t quite understand.
Chen Yuanguang explained: "Originally we fixed the area of the chip, say one square centimeter, and then we had to put as many transistors as possible on this square centimeter. 7nm refers to the size of the transistor. It is 7nm, not the size of the chip.
The smaller the transistor, the more transistors the chip has in the same area. The computing power of the chip is linked to the number of transistors.
Now we use new materials as raw materials for chips. We don't have to stick to one square centimeter. We can put transistors in an area of 4 square centimeters or even 16 square centimeters. Although we say that the volume of a single transistor is larger, we can put transistors. The area is also larger.
That is to say, the number of transistors of the chip we make using the 28-nm process is the same as that of the chip made using the 5nm process, because our substrate is larger.
Originally, China’s idea was to catch up with foreign technology. Others were working on 3nm. We haven’t even finished 28nm yet. The gap is very big, but because 3nm can be said to be the limit of technology.
Because the lattice structure and physical properties of silicon-based materials limit the further reduction of chip manufacturing processes, silicon-based materials have almost reached their limits at 3nm, and it is now difficult to guarantee yield at 3nm.
Samsung has fallen behind in this manufacturing process, and the only one that can still maintain a yield rate is TSMC.
If we continue to catch up with lithography machines, we will need the joint progress of all upstream and downstream suppliers, which is too difficult.
But by directly applying new materials, we can achieve the same computing power based on existing technology.
Even if the domestic ASML 7nm lithography machine is modified and the silicon wafer is made of topological semi-metal, the chip's computing power can perform better than 3nm, infinitely approaching 1nm.
This means that their strategy has completely failed. Our chip computing power is more advanced than theirs, so who is stuck?
And if you give us the 3nm process, we can break through the current shackles together and fill up the chip, or we will all be stuck at 1nm.
This is a matter of mutual benefit. "
Chen Yuanguang continued: "This new topological semimetal was undoubtedly discovered during the preparation process of room temperature superconductors, but I have never figured out how to design polarized light to engrave chips. That step.
During my stay in Mianyang, I finally thought about this issue clearly. ”
From the beginning to the end, Chen Yuanguang has never thought of using a photolithography machine to solve chip problems, because even if the photolithography machine is solved, as long as you still rely on this technical system, others can always find the key technical points. It makes you uncomfortable.
A ship is full of holes. You can't fix them all.
Even if it is possible to make up for everything, it will take far more energy than imagined.
It’s better to just replace it with an engine. It used to rely on people to row, but now add an engine to it. As long as the speed is fast enough, water will not leak in.
Topological semi-metal is also a technology carefully selected by Chen Yuanguang. It is suitable for the current environment and does not require too much effort to transform the existing technology system.
After listening to this, Lin Jia asked: "Can't a quantum computer also solve this problem?
Why do it multiple times with one fell swoop?"
Chen Yuanguang continued: "Quantum computers are still in the research process, and it will take at least three to five years before they can really work. Moreover, quantum computers are difficult to export, and foreign countries will not adopt our technical architecture. At best, they will buy it back and reverse engineer it." Crack.
They will not help you with the software adaptation of quantum computers, it is more of a tool for enterprise use
or as a key infrastructure for cloud computing.
Topological semi-metal can produce chips that are as good as foreign ones in up to one year or even less, which will have a greater impact on them.
It can better achieve our goals and force them to do so. Accept cooperation.”
Lin Jia asked: "What if they still insist on not cooperating?"
"Then by the time their process is close to 1nm, quantum computers have been researched, and then everyone will think that silicon-based computing power is more powerful. , or the computing power of quantum computers is more powerful.
It is impossible for them to catch up in the field of quantum computers, unless the process of ultra-temperature superconductors is leaked. "Lin Jia nodded: "Now you have to pay more attention to safety. It's better to stay in Mianyang during this period. The safety factor here is much higher than that of Shenhai.
If we surpass Amerikan in terms of chips, they may launch it. What's crazy?
Originally, Mr. Lei invited you to attend Xiaomi Automobile's press conference. He hoped that you could come on stage and say a few words and let me invite you. Under such circumstances, I declined for you.
Or I can attend in your place. ”
With Chen Yuanguang’s communication, Lin Jia has already served as the vice president in charge of marketing at Xiaomi Motors. For girls, cars are obviously more interesting than aerospace and human hibernation.
“ Okay, please help me refuse. There are too many people at the conference and I will definitely not attend. "Chen Yuanguang said.
Lin Jia then asked: "Since we want to develop chips, we must cooperate with consumer companies, right?
Who are you planning to collaborate with this time? Huawei? Millet? Or a blue and green factory? "
Chen Yuanguang said: "Huawei, only Huawei.
They dare not accept cooperation with the other three companies, which means they only want the domestic market and not the foreign market.
And the other three companies do not have chip design capabilities, so the Kirin chips designed by Huawei cannot be used in Xiaomi mobile phones.
Huawei is an unavoidable partner in chip technology. They have already mastered the 7nm process. If we can continue the discussion, we might go directly to the 7nm process and use topological semi-metal to bring the computing power close to 1nm. of chips. ”
Lin Jia murmured: "If this is really done, Apple will be driven crazy by you.
Apple's market share has been declining. If Huawei surpasses Apple in computing power, Apple will simply stop playing. Now, Huawei dominates the high-end market.
Don’t play with other brands. What is the concept of domestic chips with 1nm process power and cooperate with Huawei’s marketing? "
Chen Yuanguang said: "Not sure yet. It depends on whether they dare to gamble and transform the equipment of the 7nm process. If they mess up, their current 7nm production capacity cannot be guaranteed."
"You have to bet. This is not an economic calculation. If we can surpass Amerikan in chip technology, it will cause a huge shock to the world. 3nm is stuck on the neck of 1nm, which would be a big joke
Americen can’t even think about continuing to raise interest rates. Even if they block China’s new energy vehicles, they rely on falsified data to make the world continue to pay for the dollar. The impact of room temperature superconductivity, U.S. stocks rely on OpenAI and The GPU is struggling to support itself.
Every time Yellen comes, it is ostensibly to persuade us to give up support for the green energy industry, but in fact, she wants us to cooperate with the Fed in cutting interest rates.
Every time, they pay lip service but don’t do it. They want us to buy U.S. Treasury bonds, buy U.S. dollars, and let the Federal Reserve cut interest rates. The appreciation of the RMB against the U.S. dollar is within control, but when it comes to letting China When America enters China's new energy vehicle industry and opens up its chip supply chain, other key issues are considered.
In this case, let’s use 1nm chips to open the door to the global market and let Apple’s plunge sound the death knell for U.S. stocks.
See at that time, whether the Fed will raise interest rates or cut interest rates. ”
The Federal Reserve has been raising interest rates for a year, which has had a major impact on the global economy. It has cut global leeks by raising interest rates, but it is impossible to keep raising interest rates. Amerikan's banks and commercial real estate are struggling to support themselves in a high interest rate environment.
Silicon Valley Bank is the victim of crazy interest rate hikes.
Once the interest rate is cut, it will cause the U.S. dollar to depreciate significantly against the RMB, which means that the price of goods imported from China will increase significantly, and domestic inflation will easily get out of control.
However, the inertia of US stocks and the chain of interests maintained on it are still too many, resulting in a limited duration of the crash.
Normal temperature superconductivity is the future after all, and once the 1nm process chip is sold on Huawei mobile phones, it will directly attack Amerikan’s present.
"If I can have a mobile phone chip made with a 1nm process, then Huawei's Ascend chip can also be made with a 1nm process. The versatility is not enough, and the adaptable architecture is not enough. We can rely on technology.
Once a breakthrough is made, the former Attack Apple, and the latter attacks Nvidia. A100 and H100 are not sold to China, and some castrated versions of GPUs are sold to Chinese companies. Then we will directly use Ascend with the latest process.”
Ascend is launched by Huawei. Chip for artificial intelligence training.
“This is a rare opportunity. We have to plan carefully and cooperate with the financial layout to give the United States a hard blow.
We can appropriately disclose the news to Goldman Sachs, Big and Small Morgan, Institutions like Sequoia will definitely cooperate with us in shorting U.S. stocks.”
"The premise is that this thing can be built. There is no doubt about Chen Yuanguang's credibility, and the answers given by the experts are also very affirmative. The application of topological semi-metals on chips is completely feasible.
But in terms of process Regarding whether, as Chen Yuanguang said, 7nm can compete with 1nm, the stability of the final chip needs to be confirmed.
Let Huawei intervene, and don’t use their own 14nm lithography machine. The Xinana machine was allocated for this project.
The 7nm lithography machine happened to be idle, so I decided to use it.
A new company was established to take charge of this project. Jiangcheng State-owned Assets will take some shares to cover this bad debt.”
Hongxin Semiconductor can be regarded as the tears of the times. When semiconductor projects were being carried out all over the country, Hongxin came out of the blue, attracted some TSMC executives to join, and then bought a 7nm lithography machine from ASML. This lithography machine The engraving machine is still the only 7nm lithography machine in China to this day.
As a result, Hongxin collapsed before the lithography machine could even start construction. The lithography machine was mortgaged to the investor, Jiangcheng State-owned Assets, and tens of billions were wasted.
“Chen Yuanguang’s calculation of technology investment depends on whether he does it in his own name or in the name of Guangjia Aerospace.
The foreign investors in Guangjia Aerospace must withdraw their shares as soon as possible. They will not withdraw. They have to withdraw, this is not a negotiation, don’t give them a chance to bargain.”
Under the coordination of Hua Guo, the 7nm lithography machine mortgaged by Hongxin was directly allocated to this project, and Huawei also entered the market. , the project was internally named Project 712.
Update one chapter first
(End of this chapter)