Chapter 263 Let competitors be indisputable (2/4)

Fireworks and firecrackers are not allowed in the capital now, but the habit of watching the year is still maintained. After returning home, Ye Zishu and his brothers sat in the living room to watch the Spring Festival Gala after taking a shower.

The two quietly waited for the arrival of the New Year's bell, but after 12 o'clock in the morning, they didn't seem to feel anything. After thinking about it, it was because they didn't hear the sound of firecrackers all over the city.

People who came out of the countryside all resigned from the old year amidst the continuous sound of firecrackers, but suddenly it disappeared, feeling that the taste of this year was extremely dull.

"Go to bed early, you don't need to turn off the lights at night!" Ye Zishu stood up and said to Ye Zihua.

It was just after 12 o'clock in the morning, so he didn't intend to stay any longer. He might as well go to bed early and study in the virtual library while he had time.

In fact, Ye Zishu has basically experienced the usual silicon-based semiconductor chip technology in his previous life, and 5nm or even 3nm process manufacturing technology has appeared.

Chip manufacturing technology has reached this level, and the need to continue shrinking is already very low, because compared with the cost, the performance improvement is very limited.

This is also why the price of chips in the previous life is getting more and more expensive, and it is even so expensive that it takes a long time to think about buying a graphics card. It is because the manufacturing process has reached a certain level, and the cost of chip manufacturing will increase more and more.

Moreover, with the improvement of the manufacturing process, the production efficiency of the chip will gradually decrease, because the more precise the manufacturing technology, the slower the production efficiency, which is a general rule.

Leaf Book has now shortened the process technology of silicon-based chips by nearly 10 years, and will continue to accelerate the update and iteration of chip process technology in the future.

Since we want to kill other competitors in the field of silicon-based chips, we cannot slowly come up with technology like squeezing toothpaste, because competitors will not give up this promising industry because of a temporary failure.

Even when he really monopolizes the global chip market, the other party may use national power to concentrate on research and development of better technologies, hoping to break out of the encirclement.

He knows how important chip technology is, and other companies and countries will also know it. After the duel between companies is defeated, it is very possible for the country to organize forces to tackle key problems.

Coupled with the talent advantages they have accumulated in chips, if they really concentrate on doing big things, it is really possible to break through the monopoly he established in a short time.

Therefore, for a long time to come, Xuanwu Technology's chip manufacturing technology still needs to be iterated at a relatively fast speed.

The time interval for process update must not exceed 5 years, and the shortest interval is about 2 years.

Because the iteration cycle is less than two years, the return that Xuanwu Technology can obtain is relatively low, so it is necessary to strike a balance between preventing competitors and earning benefits.

In addition to the advantages in chip manufacturing technology, the biggest advantage is the lead in chip architecture, which is still something that the opponent cannot break through in the short term.

Even if the chip manufacturing technology level of the two parties is the same, the chip architecture he proposed will still make the computing power of the chip many times ahead of the other party, and the chip architecture is the core competitiveness.

However, according to this iteration speed, it is estimated that within a decade or so, the chip process technology may be advanced to 3 nanometers, or even within the range of 1 nanometer.

Therefore, he must prepare for the next generation of chips in advance. What he is looking at now is carbon-based chips, and the technology in this area is still very complicated.

Because carbon-based chips involve material technologies such as carbon nanotubes and graphene, the manufacturing method is very different from the existing silicon-based chip manufacturing technology.

It can be said that carbon-based chips are basically rebuilding a chip manufacturing system, so there are still a lot of technologies involved, and he needs to spend a lot of time on this information.

However, he basically covers the theoretical content. As for the technical details, he doesn't know, because for him, it is a waste of time.

If you really want to take it out at that time, you can just copy it. There is no need for such content now, and leading technical principles are enough.

In fact, carbon-based chips can't be said to be very powerful. There are all kinds of weird chip manufacturing technologies in the virtual library, and their performance is even stronger than carbon-based chips.

Among them, there are many chips with composite materials as the substrate, which are much better than carbon-based chips. However, as long as the chip manufacturing does not break through the traditional method, in fact, even if the substrate is replaced, there will be no substantial improvement.

Just like replacing a silicon base with a carbon base, data transmission and computing are essentially achieved through the migration of electrons. No matter how the substrate is replaced, the migration rate of electrons will still be subject to various restrictions and cannot be increased indefinitely.

Because electrons are relatively large, they are easy to collide during migration, which is the origin of resistance, which not only consumes energy, but also affects the speed of electron migration.

Therefore, the best chip material is to achieve superconductivity, so that under the same process technology, the calculation speed will be increased by thousands of times. Superconducting materials are the limit of electronic chips.

In fact, instead of using superconductors to manufacture superconducting chips, it is better to develop directly in the direction of photonic chips, and achieve dimensionality reduction strikes through the unique physical properties of photons.

There are many things that can be done in scientific research, but in many cases it is not necessary to do so, because the comprehensive cost needs to be calculated.

In fact, the advancement of semiconductor manufacturing technology does not increase the performance of the chip very much. The biggest gain is the design capability of the chip.

The chip design architecture technology he is going to present, under the same process technology, is more than five times the performance of other chip design architectures.

This is not the most advanced chip design architecture he has come up with, and there are also many chip design architectures in his hand that have improved performance by tens or hundreds of times.

Therefore, after the chip process technology reaches a certain bottleneck, do not rush to find new manufacturing technologies, you can slowly update the chip design architecture.

In order to obtain more powerful performance, other companies have to invest huge sums of money in research and development of more advanced chip manufacturing technology, and he only needs to come up with a new design architecture technology to be able to beat it.

At that time, the chip performance gap will be huge, and the cost gap will also be huge. The two will not be at the same level of competition at all.

So in the face of competition on chips, he doesn't have to worry at all, because he has too many technologies that can guarantee an absolute competitive advantage in this field.

Technical issues are not his biggest concern, but how to unlock the next level of the virtual library is his top priority.

Before, he guessed that he might need to finish reading the books and materials on this level by himself, but now he feels that the possibility of such a way of opening is very low, because he is a person after all, and his energy is limited, and he may not be able to finish reading after exhausting his life.

So he speculated that it might be related to the level of civilization, that is to say, if the system judges that the level of civilization has increased, the level of open technical data will also increase.

Although he is not sure, he feels that this guess is more likely to be close to the truth, so it is very necessary to strengthen attempts in this area.

As for how to improve the level of civilization, he is not very sure about this aspect. It means improving the knowledge level of human beings and making human beings more advanced, or it is only manifested in technology.

There is still a big difference between the two. The former needs to improve the scientific ability of the group as a whole, rather than polarization. The powerful people are very powerful, and the stupid people are very stupid.

If it's technical, it's relatively simple, and elite education can be adopted to continuously improve the top technology, but this will cause a disconnection within civilization, which should not be a good thing.

Therefore, he personally prefers to improve the civilization level of the entire ethnic group, rather than polarize it, because this is not conducive to the overall development of the entire ethnic group.

In the future, there will be more and more problems, and eventually there may even be ethnic differentiation in the biological sense.

: