Recently, Japan** announced some secrets of the Kirin 9000S chipHuawei's Kirin 9000S chip is called a 14nm process, not 7nm as rumored by the outside world. Huawei uses a special architecture to increase the density of transistors, and finally achieves performance close to that of the 7nm process.
Seeing this, I couldn't get angry, your master didn't dare to say that it was 14nm, you said, is this a slap in the face and a fight?
In that case, let's analyze it and see how I slap Japan in the face.
Let's first understand what a process is.
We talk about 14nm, 7nm, 5nm every day, but what is it?
We know that the basic unit of a chip is the transistor, which is composed of a source, drain and gate, and the process is the length of the channel from source to drain.
However, when the process reaches 22nm, the problem of short channel effect becomes more and more obvious, and it is no longer possible to control the electronic transitionResearch and development of new processes.
So FinFET technology appeared, which effectively controlled the electronic transition, so that Moore's Law continued, but it had a problem, that is, the channel length was difficult to measure, in order to reduce the difficulty and easy to observeThe gate length is simply used as a reference standard for the process.
In other words, the current process is no longer standard, just an approximation.
It is an exaggeration that this approximation will deviate more and more as the process decreases.
Before 2000, iterations from 60nm to 180nm were made with every 0At 7x, the number of transistors on the chip increases by a factor of 2.
After 2000, the 130nm iteration to the 45nm process was completed, and the process node was reduced by 0At 72 times, the number of transistors on the chip can be twice as many as before.
In the process below 14nm, this moisture is even greater.
TSMC's 7nm process is equivalent to Intel's 10nm, because the transistor density and performance of the two are about the same, and they belong to a generation of products.
However, the chicken thief's TSMC has changed this competitive landscape by naming it, making everyone mistakenly believe that TSMC's process is several generations ahead of its opponents.
No wonder Intel executives angrily criticized TSMC for not speaking of martial arts.
It's like selling wine, all marked with 5 years or 10 years, and consumers mistakenly think that it is wine that has been cellared for 5 or 10 years, and happily buy it back to drink, and drink it from friends, you see that I have been delicious for 10 years!
In fact, 5 years is 3 years of cellaring, and 10 years is 6 years of cellaring. If it sells well, it will not be stored for 5 years, and it will be sold directly, and if it is 10 years, it will be sold after two or three years of cellaring. Rich don't make bastards!
You can imagine TSMC's most advanced 3nm and 4nm processes, how much water there is, it is estimated that 52 degrees of wine can be mixed with 38 degrees, if it reaches 1nm, it is really liquor can be mixed into beer.
Therefore, the chips used in mobile phones now cannot reach the 4nm and 5nm standards at all, which is why chip manufacturers like to squeeze toothpaste, because the more difficult it is to upgrade the technology later, the more money is consumed.
Overseas chip companies can mix water and squeeze toothpaste, but Chinese chip companies cannotWho are you bullying?We didn't set off fireworks in Hiroshima and Nagasaki.
After the United States set off two fireworks that year, Japan directly announced its surrender, and Japanese girls lined up to serve American GIs. It's really envious, jealous and hateful.
Of course, our country has always advocated peaceful development, and it is easy not to set off fireworks on Mt. Fuji, but you have to keep teasing, come and hit me, hit me!Then we can also meet your needs.
After the release of Huawei's Kirin 9000s, it can be said that it shocked the chip field, so many bloggers played the game of dismantling the chip, I saw it for the first time, thousands of yuan, tens of thousands of yuan to buy a mobile phone, just to disassemble and play, if you don't use it, donate it to the poor mountainous areas!
While disassembling Huawei's mobile phones and chips, while earning the traffic of the Chinese people, to be honest, I don't like it.
It's understandable that everyone wants to be famous, want to be an Internet celebrity, and want to make money, but don't waste it for the sake of traffic. China has a population of 1.4 billion, but only 1Of the 400 million households, only 33 million households can save 100,000.
You easily dismantle a Huawei Mate 60Pro worth 6999 yuan, this kind of chic and domineering, compared with the altitude of 4000 meters, minus 20 degrees, eating ice ballast for breakfast, and climbing more than ten kilometers to go to school compared with poor children, it is somewhat abrupt!
The Kirin 9000s was disassembled, and chip experts finally found a problem: the Kirin 9000s did not use any American technology, it used a completely new technology that had never been done before.
Huawei's chips have changed tracks, Qualcomm, Apple, TSMC, Samsung, and Intel have occupied all the original tracks, and Huawei can only find another way, and as a result, it has found a broad road.
The most terrible thing is that after the chip was disassembled, the foundry was not found, and it was said that it was TSMC at first, and then it was SMIC, and even Intel was involved.
The funniest thing is that Hongxing Erke and Golden Monkey issued a statement that the company is not a foundry of Huawei Kirin 9000s.
This is a big deal for the United States, Japan, and the Netherlands. You can't let Chinese chips develop, otherwise we will have a good life.
began to do everything possible to restrict the rise of Chinese chips, restrict the export of semiconductor equipment, restrict the export of semiconductor materials, and restrict Nvidia's supply of advanced AI chips.
And the U.S. Bureau of Industry and Security plans to raise the budget to prevent sensitive technologies from entering China and Russia.
Although Japan's strength is insufficient, it can play the role of a-stirring stick, and through ** voice, it is said that Huawei Kirin 9000s is not a 7nm process, it is a 14nm process, but it has improved performance through a certain technology.
To be honest, which company develops chips not for use, as long as it can meet the performance requirements and meet the demand, it is a good chip.
You claim 5nm and 4nm, but as soon as you use it, the temperature is dry to 76, can this work?Therefore, the product should not always look at the advertisement, but also look at the efficacy.
What's more, Huawei's Kirin 9000s has a very high degree of domestic production, using pure domestic technology + domestic processor architecture + domestic "self-developed 4.."99999G baseband" (5G ultra-fast network speed) + domestic manufacturing.
If this is placed in the ** market, it will be a "heaven and earth board", can other players not be jealous and hateful when they see it?
This chip works with the Hongmeng system, which really complements each other, and the strengths can be more visible, and the shortcomings are almost gone. The overall usage effect surpasses that of the iPhone 15.
If this rate continues, the next step will be to replace Japan as a semiconductor material.
According to the statistics of research institutions,At present, 14 of the nine semiconductor materials necessary in the world have reached a market share of 50%, and they have a monopoly on EUV photoresists.
This is Japan's biggest and only advantage in the field of semiconductors.
Don't say that Nikon's lithography machine is all in the past, and soon the domestic lithography machine will come out, and it will be a matter of minutes to surpass Nikon. Besides, our target is not Nikon, but Dutch ASML.
For Japan, what we have to do is to defeat it in semiconductor materials, especially to achieve a counterattack of EUV photoresist and completely kick Japan out of the chip field.
This requires the joint efforts of our whole country, not only the efforts of scientific research institutes, the Chinese Academy of Sciences, Tsinghua University, industrial companies such as Shanghai Silicon Industry and Nanda Optoelectronics, but also consumers like you and me.
Only by supporting domestic chips can we overcome obstacles and achieve counterattack and transcendence.