Chapter 308 The real 5nm!

"Pa bang bang~"

The entire venue was filled with thunderous applause. China’s double breakthroughs in photolithography machines and processes are worthy of excitement for every Chinese present!

Especially Lin Xiao's last sentence, only the chips born here are the best chips, which gave them a sense of pride.

When in the past have they dared to claim this?

If they dare to make such a claim, let alone the ridicule from abroad, some domestic netizens will have to use the topic to their advantage.

Now, the data comparison provided by Lin Xiao is entirely based on facts. Even if someone questions it, they will at most question it from other aspects, and it is impossible to question it from the performance comparison between the two.

As for why I don’t doubt whether the data given by Lin Xiao is some kind of false comparison method, such as only mentioning my own advantages and not mentioning the shortcomings, this is because Lin Xiao directly gave the running scores of GEEKBENCH5.

GEEKBENCH5 is a professional benchmarking software for mobile phone chips. It is often used to compare the performance gap of different mobile phone processors. The above benchmarks basically represent professional benchmarks. If you want to spend money on this, it is completely It's impossible.

Because of this, when mobile phone manufacturers talked about mobile phone chips at press conferences in the past, they basically did not dare to directly put the running scores of GEEKBENCH5 at the press conference.

And now that Lin Xiao has put it up directly, it undoubtedly shows his confidence.

And his confidence also infected other Chinese people present, causing the applause to last for a long time.

As for the representatives from ASML, Nikon, Canon, TSMC, Samsung and other companies, everyone’s face was very dark.

Especially TSMC and Samsung, at this moment, the pressure has come to them from photolithography machine manufacturers such as Asmail.

At this moment, they all felt unprecedented pressure. Before coming, they had already underestimated this opponent. However, they did not expect that they underestimated it too much.

Their X5 technology can actually achieve this level of performance.

Is it really a 5nm process, not a 2 or 3nm process?

In particular, as Lin Xiao mentioned, Kirin 9000 is more compatible with TSMC's N5 process, but not with this new X5 process. They all know this.

Before chip design, it is necessary to select a cooperative foundry, and then the foundry will synthesize the relevant parameters of the process in a model file and send it to the chip design company, and then the chip design company can import these model files into the EDA software, so Kirin The 9000 chip is designed based on TSMC's N5 process.

But now, Lin Xiao and the others have used the incompatible X5 process to create a stronger Kirin 9000. How good must their process be to achieve this kind of performance improvement?

A chief technology officer from TSMC and a chip process engineer both looked very heavy at this time.

"Is this impossible?" The process engineer couldn't help but say: "Did he make up that data randomly? Is it fake?"

The chief technology officer shook his head and said: "The probability of fraud is very low. How could he fake something like this at a press conference?"

The process engineer frowned. The probability of fraud was indeed very low. With Lin Xiao's reputation, why would he do it? Like the Hanxin incident in China twenty years ago, it is impossible for them to repeat it again.

If you say this, doesn't it mean it's true?

But at this time, the chief technology officer said: "However, there are still two key points that they have not mentioned."

"What two points?"

When the chief technology officer was about to say something, Lin Xiao on the stage looked like he had just remembered something and said with a smile: "Oh, I forgot to mention two important points just now."

The applause that was still continuing stopped, and people were waiting for what two important points Lin Xiao would say.

Lin Xiao smiled slightly: "The first point is the energy efficiency ratio."

The people present were suddenly shocked.

right! And the energy efficiency ratio!

Qualcomm's Snapdragon 8gen1 chip is so powerful, with a single-core score of 1,200, which is nearly 200 higher than the Kirin 9000's 1,020 score, and its multi-core score is also higher than the Kirin 9000. But why is it so ridiculed by the mobile phone industry?

It is because the energy efficiency ratio of this chip is so poor that the power consumption has increased significantly with the same performance, causing it and its brother chip Snapdragon 888 to be called the Fire Dragon. As for why this is so?

This requires mentioning Samsung’s process issues.

It can be said that Samsung’s process is the main reason for its poor energy efficiency.

So when it comes to technology, we must mention energy efficiency ratio.

So when Lin Xiao mentioned this now, could it be that... he was also very confident?

At this time, Lin Xiao also had a smile on his face: "In addition, according to our tests, the energy efficiency ratio of the Kirin 9000 chip based on the X5 process is 15% higher than that of the Kirin 9000 chip based on the N5 process!"

Before the people below were surprised, he went on to say: "Then the second point is the yield rate. As of the start of this conference, we have produced 200,000 Kirin 9000 chips, and our quality products The yield rate is maintained at around 95%! This is an absolutely excellent yield rate!

"These two points are the two important points I want to say!"

"So, I can declare that the best chips can only be born from our hands."

"The chips made by our 5nm process are also true 5nm process!"

In the auditorium, the Chinese audience was once again excited and stood up and applauded again.

Such energy consumption ratio and such yield rate made them excited again.

This is the process that truly deserves to be able to produce the best chips!

The faces of TSMC's chief technology officer and process engineers were even duller at this time.

"CTO, are these the two important points you mentioned?" The process engineer looked at the chief technology officer and couldn't help but ask.

The chief technology officer's eyes were still a little dull, and finally he nodded.

What he wanted to say just now were these two points.

Performance, energy efficiency ratio, and yield rate are the keys to determining whether a chip's technology is excellent or not. The first two determine whether the chip will be popular. In terms of reputation, the yield rate determines the direct profit of the foundry.

Now, all the data about the X5 process given by Lin Xiao has completely defeated their N5 process, and even from a comparison point of view, it may be stronger than their N5p and N4 processes.

N5p is their 5nm enhanced version process, and N4 is the so-called 4nm process.

In fact, they now feel a little unconfident about the N3 process that they are still developing.

The X5 process Lin Xiao mentioned, is this really a 5nm process?

Suddenly, they thought about what Lin Xiao said, "real 5nm process". Could this mean...

"Are they really doing it in full compliance with the 5nm gate length standard?"

The real 5nm process means that the grating length of the transistor is 5nm.

If this is true...

The chief technology officer of TSMC felt deeply unbelievable for a moment. If this is really the case, then they may have to achieve the N1 process, which is the 1nm process with their virtual label nomenclature, to be able to compete with them. The X5's craftsmanship compares favorably.

But for them, it will take at least many years before they start developing the N1 process. The gate length of their planned 3nm process is only 14nm, which is still a full 9nm gap from 5nm!

At first he thought it was impossible, but then he thought that the wavelength of this X-ray lithography machine was 3.56nm. It was obviously easier to do this than they used the EUV lithography machine.

This means that his guess is very likely to be true!

At this moment, the chief technology officer fell into a state of doubt about life.

What was different from his sluggishness was the continued applause around him.

The excitement belongs to the mainlanders, and he has nothing.

…

At the same time, the live broadcast room of the press conference was extremely lively. As Lin Xiao began to introduce their craftsmanship and compared the data, the netizens watching the live broadcast showed that they were better than the LOL Huaguo team that won the championship. Even more exciting.

Especially now, the number of online viewers has increased from 200,000 to more than 1 million.

More than one million real netizens have already become excited after seeing the data given by Lin Xiao that completely beats TSMC's process.

Anyone in the computer industry is interested in chips, and those who are interested in chips often make fun of Samsung's technology and lament that TSMC is awesome.

As a result, now, they actually have a more powerful craft of their own, and for a moment, they only feel that spring is coming.

"Rise up!" 』

"too strong! 』

"Another explosion!" Asmail, Canon, Nikon, TSMC, Samsung, just win them all, five kills! 』

"How dare the little Semiconductor Manufacturing Co., Ltd. try to do whatever it takes?" Take it directly

"We are number one!" 』

"Lin Shen: There is no one who can fight!" 』

"Holy shit, how many Kirin 9000s have been produced?" Is this considered Kirin 9000s? When will Warwick start selling mobile phones? 』

"Awesome 666!" 』

"Lin Shen is too strong! In this life, I will only obey Lin Shen! 』

"Who dares to disobey Lin Shen? Get him! 』

"..."