LATEST UPDATES

Published at 1st of March 2022 07:05:59 PM


Chapter 1457

If audio player doesn't work, press Stop then Play button again




No technology can be invented for no reason. Although talented scientists may speed up the research and development of technology, talented scientists also need to spend money.

It can be said that all the technological progress of mankind is made by money. Jiang Yanhai seems to have a fast speed of scientific research, which doesn't seem to cost much money.

But is that really the case? Not at all.

A very simple thing. Last time, Jiang Yanhai was going to the Research Institute near Yangcheng to have a video contact with Professor Liu and Professor Hu. At that time, the whole research institute was closed, and all scientific research activities were suspended. What should be taken off, what should be taken off, and what should be left behind were left behind.

These need money! Some studies stop for a day. The power of computers alone is not a small number.

Jiang Yanhai calls any supercomputing if he wants to call any supercomputing, and even doesn't hesitate to directly use the strategic reserve. Doesn't the strategic reserve need money? These are all money.

Therefore, scientific research does not save money, but Jiang Yanhai is more efficient here. Others may have an effect of 10000 yuan, while Jiang Yanhai may have an effect of 100 yuan.

But this 100 yuan also needs to be spent.

Warwick has today not because of how strong their researchers are, but because they are willing to invest.

What is the concept of throwing away a 100 billion company every year? How many companies have a national market value of more than 100 billion?

"Well, it's easy to say. Let's speed up the progress. In fact, there is probably a law in the design of the chip and the matching with the system..." Jiang Yanhai said with a smile.

Of course, they are talking about silicon-based chips. At present, the ultimate performance of silicon-based chips has not been squeezed out.

The next step of 3nm is 1nm. It is possible to even break through the limit value of nano and reach a few tenths of a nanometer.

What is the concept of a few tenths of a nanometer? It's the size of an atom! In other words, in the future, a transistor will be as big as an atom.

But is this technology the limit? No, even nano robots can exist, which is nothing.

Therefore, technology has no limit. Think about how a star several times larger than the sun will eventually shrink into a black hole a few centimeters in diameter, and you will understand how magical matter is.

After entering Warwick's Research Institute, Jiang Yanhai stopped talking to these bosses because talking to them had no effect on the upcoming research. He directly asked these researchers for the next generation design drawings of Kirin series chips.

This is a 3-nanometer chip design drawing. In fact, the completion degree of the whole drawing is very high. If it is not Jiang Yanhai, they estimate that they will enter the streaming stage. If the streaming is successful, it also means that the chip is successful.

Of course, the success of streaming does not mean that the chip is perfect. The chip with design defects is not without heat, energy consumption, power and other aspects.

This is why many chips are a series and an architecture. In this way, although these chips are constantly improving, the performance of this chip can also be continuously optimized by them.

However, in China, all companies can only choose to overtake at corners, including Warwick, because they have too little reserves in this regard.

Think about when the earliest IBM was founded in the United States? Companies including Intel have been around for decades.

But what about China? When was the first real chip company established?

The difference between the two is too great. There is a huge lack of experience alone, which is why Li ruobai and his colleagues choose to work directly on new materials and new structures.

Because the limit of silicon crystal materials is about to be reached, the current silicon crystal materials may still be used for 20 or 30 years, but if they have achieved results in new materials and new architecture at the same time, these 20 or 30 years will be enough for them to go to the forefront of the world.

In this respect, they have advantages, at least many new materials, but the companies next door don't.

Jiang Yanhai didn't choose to delay. After checking the new generation of Kirin chips, he directly said, "your Kirin's design on 3 nm is still based on the architecture design of the previous generation of Kirin chips. In this regard, I probably have some improved ideas, which you can record together."

Warwick adopts a series of architectures, which is inevitable. They can't directly start designing new architectures and new chips like Li ruobai, because the purposes of both sides are different.

Warwick's idea is very simple. They want to put it into use quickly, but Li ruobai is responsible for the national strategic plan, and the support between the two sides is naturally different.

However, although the chip design of Warwick is not as simple and efficient as that of Jiang Yanhai, and has made progress in energy consumption and other aspects, the performance gap is not large.

At least users can't feel it.

Jiang Yanhai constantly puts forward some improvement suggestions he has found before, while Warwick engineers are constantly verifying. Some verifications can be simulated directly on supercomputing, while some must be carried out after successful streaming and compared with the previous chips to carry out various limit tests to obtain the latest data.

In the past, the speed of the second test method would be very slow, because we didn't have a lithography machine before, but now it's different.

3nm lithography machines are being produced at a crazy speed in China. It can be said that OEM enterprises such as TSMC and even Samsung have sought to buy 3nm lithography machines.

Orders in this regard, the domestic natural is not to refuse, not only to refuse, but also hit the price of cabbage!

ASML's 7 nm lithography machine costs US $100 million. Sorry, we only sell it for us $80 million, 3 nm.

And the data of yield is absolutely not bad! They are all international top data.

80 million dollars is the price of cabbage!

This price is all over the world. There is no one who can beat it.

Although ASML has been trying to speed up its progress, according to the data analysis of Erdan, it will take at least a year and a half before they can come up with a mature 3-nm lithography machine.

But in a year and a half, the cauliflower is cold, okay?

In fact, not long after Xiaomi's mobile phone was on sale, the Ministry of Commerce next door has quietly revoked all regulations on chip embargo and lithography embargo.

However, if Jiang Yanhai remembers correctly, not long ago, the Ministry of Commerce issued the latest relevant bill, which prohibits the export of 3-nm lithography machines, but I can sell them.




Please report us if you find any errors so we can fix it asap!


COMMENTS