Lishan Semiconductor is huge, with chip factories stretching one after another.

From the lowest 8 micron process factories to the most advanced 0.5 micron factories, everything is available. Some chips with a relatively low level of technology, although they are used to produce products such as CPU, memory, and flash memory, are no longer available.

However, the chips in the semiconductor industry are not only the commonly used types of chips, in fact, almost all industries that use electricity may need semiconductor parts, and some semiconductor parts do not need to be so advanced, but more need to be durable. In addition, human beings have tens of thousands of types of semiconductor chips, and only a few are general-purpose consumer chips like DRAM particles and CPUs.

A lot of niche chips, not widely used, the market is small, but it is also very needed. For example, some military chips may not be more advanced than civilian CPUs, but they are more special-purpose and need to be specially designed, and the process used is not advanced, but the cost may not be cheap.

After all, the development of CPU chips may be bought by tens of millions of users, so the research and development cost of each chip is already very low.

Some niche chips, made with very backward technology, are priced at tens of thousands of dollars, mainly because the research and development of these chips is costly, and the demand may only be hundreds of thousands of orders, so the cost is also very expensive.

For a long time, China's semiconductor industry, not only cutting-edge technology, was not developed. Even if it is a type of semiconductor chip, the process level is not first-class, but the key is that its product range covers more than 10,000 different semiconductor chips. Many general-purpose chip manufacturers are unwilling to develop and produce chips, but TI is still willing to do it, and gradually monopolize the niche business.

TI is different from companies such as Intel, mainly because of different positioning. Intel has done the field of general-purpose chips, and TI has no advantage if it does it again, so it has become a company that does everything, no matter how niche the chip, as long as the customer can afford to pay for it, they are willing to develop it.

Lishan Semiconductor Co., Ltd., even a lot of production lines with backward processes, has not really been eliminated. Instead, they began to gradually turn some production lines with different processes to make some niche products, anyway, these chip factory equipment is already zero cost. The rest is more about labor and material costs.

It is precisely because of this that these dilapidated production lines are not eliminated by Lishan Semiconductor, so it employs more people than Xinfeng Semiconductor.

The types of industrial chains of Xinchuang Semiconductor are certainly not as rich as those of Lishan Semiconductor.

To a large extent, Lishan Semiconductor is not only for civilian use, but also, because it is a state-owned enterprise, also has a military background. Moreover, the process level is more advanced than those semiconductor factories of the old aerospace and military industry, so many domestic aerospace and military industries need to use some chips that are not in high demand, and they also go to Lishan Semiconductor to ask for foundry.

Many aerospace-grade chips that were originally not exported to China by foreign countries, Lishan Semiconductor was also able to tackle key problems with the aerospace team, and successively broke many technical blockades. It is precisely because of the sweetness that many domestic aerospace researchers have not been able to introduce some technologies abroad, so they will not talk about it directly, and let Lishan Semiconductor cooperate with independent R&D and manufacturing.

Undertaking these foundries, the profit may not be much, but there is no doubt that the background of Lishan Semiconductor is undoubtedly stronger.

However, because it was set up in a special zone, and the initial mechanism was less bureaucratic and more market-oriented. In addition, Lishan Semiconductor is a joint-stock enterprise, so although it is not as efficient as Xinqi Semiconductor, it is definitely better than most companies.

Shen Xubang held a Luban 2nd generation quad-core desktop chip and said proudly: "In the performance test of this chip, the single-core performance is more than 3 times that of the Pentium chip, and the quad-core is more than 7 times that of the Pentium chip." Even if the Pentium chip uses a multi-core architecture, in this regard, Intel is not a match for the Luban 2 generation. As for, compared with our own Luban 1st generation, the performance is also more than 3 times improved. At present, Luban's second-generation chips can be said to have directly widened the gap with its opponents by more than one generation! If it weren't for Intel's X86 platform, which has about $50 billion worth of stock software, the X86 platform would not have survived at all!

The compatibility of the moat of the chip company, the more software in stock, the more for the chip company. The more secure it is, because many users are reluctant to use the commonly used application software, so they have to continue to use the hardware of the platform.

However, due to the lack of software ecological support, many of these high-performance CPU chips have gradually failed to develop after a few years or ten years. This failure is more due to commercial failure than technical failure.

In addition to high performance, XRM architecture chips are also paying more and more attention to the compatibility of the software ecosystem and building a moat.

In the past, it was game software, and now of course it is also important to games, but any kind of application software is a chip of XRM architecture, which is constantly encouraged and enriched.

Lin Qi said with a smile: "There is no eternal company, and there is no technology that will always be ahead of the ages. If you want to take the lead, you need to keep innovating and constantly developing into the future. In the stage, Intel may lag behind us in performance due to architecture problems, but the market does not completely look at performance. ”

"That's right, the main thing is to look at software applications, and over the years, the application of Pangu has also developed rapidly. Although the market value of the stock of application software is lower than that of X86, the growth rate is not slow. Shen Xubang said with a smile.

After that, Shen Xubang showed Lin Qi another good thing, the 32-core Luban 2nd generation processor!

It seems like it's just a bit more core, but in reality, it's much more technical. If the multi-core wants to play a better performance, the technical content is also very high. Platforms such as various high-performance supercomputers and servers have become more and more inclined to use chips with more cores to pile up performance, and more cores have been piled up when they leave the factory, which is obviously more considerate than purchasing a bunch of chips and finding a way to solve the problem of chip collaboration.

"Just such a multi-core chip, floating point operations per second, at most, more than 500 million times!" Shen Xubang said with a smile, "Ten years ago, it was already the level of floating point operations of supercomputers." Now with such a chip, such a complex budget has been achieved! It can be said that this 32-core Luban second-generation chip should be the strongest chip in the world at present! I can't find a second chip, which is even stronger than it! And the cost of such a chip is only more than 100,000 yuan!

"It is estimated that individual users will not use such a strong chip, but this chip should have a bright future. Lin Qi looked at this chip, perhaps, a Celeron chip for individual users at the beginning of the 21st century may be more powerful than this chip. However, a chip of 100,000 yuan, plus other supporting chips, should be able to achieve less than 1 million yuan, and configure a platform comparable to a mainframe.

The current market price of mainframes is at least more than one million dollars. Although, there will be some software and technical services included in it. However, the mainframes on the market start at a million-dollar rate. And the performance may not be towards the 32-core Luban 2nd generation chip in front of you.

What's more important is power saving, a CPU chip with an integrated 32 core, no matter what, will save more power than those mainframes stacked with hundreds of CPUs.

Therefore, this chip is not only an order in the commercial market, but also in scientific research, military industry and other fields, and I am afraid that this high-performance chip is also in short supply.

Therefore, Lin Qi gradually understood why Lishan Semiconductor attaches more importance to the field of XRM chips, after all, high-performance chips can now immediately save domestic scientific research costs.

As for the ARM chip that saves power on the mobile terminal, it is the new venture semiconductor that pays more attention to it, although the chip is energy-saving, but the efficiency is also relatively low, and in the high-performance application market, even in the future, it does not occupy the mainstream of the market.

After visiting Lishan Semiconductor, Lin Qi basically understood that the Luban 2nd generation CPU could be listed with the Pangu 8th generation computer. However, the inventory of Pangu 8 must be very low at the beginning, because, by Double 11, Lishan Semiconductor is estimated to be able to produce 100,000 chips.

If all Pangu 8 generations are supplied, there will only be 100,000 units in stock for Double 11.

It can be said that this is the lowest stock in the Pangu computer of the past dynasties, and it is not intended to play monkey marketing, but when the time comes, it will have to play monkey, and a small amount of spot and more pre-sold futures will be sold.

https:

Genius one second to remember the address of this site:. Mobile version reading URL: m.