Great Power Technology

Chapter 7 This thing does not belong to this era

Ye Zhou didn't know how long it had passed. During the waiting time, he reviewed all the knowledge about Yinglong EDA in his mind over and over again, and even roughly went through its code composition.

The screen of the phone was soaked with sweat from his palms, but the group of people in front of him remained silent.

Finally, a bald man standing behind Lu Xu spoke first.

"No need to test, I understand."

Lu Xu turned his head, and after recognizing the other party, he nodded slightly.

"Mr. Zhang, this thing is authentic."

"It's fine to call Zhang Yue, and the boss has issued repeated orders not to call him the general manager. It seems that this thing is not just fidelity, the framework design concept in it, I guess even Synopsys can't do it."

Synopsys is the world's top EDA software provider, and their EDA products have monopolized the international market for more than 20 years.

But now Zhang Yue made his point directly.

What Yinglong can do, Synopsys cannot.

This alone is enough to prove Ying Long's originality and advancement.

"I don't think it's as simple as the frame design. How to say this thing, I don't think it's something of this era."

The man next to Zhang Yue interrupted.

Hearing his words, everyone was stunned for a moment, and Ye Zhou's heart skipped a beat.

Of course he himself knows that this thing does not belong to the current level of technology, but he does not intend to reveal this for the time being.

Could it be seen through so quickly?

"How to say?"

Zhang Yue frowned and asked.

"The development direction of EDA is actually nothing more than the optimization of algorithms and architecture. However, have you noticed that this thing provides a completely new route."

"Come on, Lu Xu, let me operate it."

While talking, the man pushed Lu Xu away and sat on the chair.

"Look, I noticed this page when Lu Xu was operating it just now, do you know what it is?"

The people around him gathered together in an instant, and after seeing clearly the words the man pointed with the mouse, there was a neat sound of shit from the crowd.

Even Zhang Yue, who is a big boss, is no exception.

"My goddamn, is this the person who designed this EDA for the Cloud HPC module? From the very beginning, he didn't intend to die on the algorithm and architecture optimization that was close to the bottleneck. He wanted to directly pull the chip design to the cloud. .”

The man nodded, he tried to start the module, but then found that the optional cloud server was empty.

"That's right, when designing this EDA, he took into account the insufficient computing power of existing computers. Therefore, he didn't intend to limit the performance of EDA to a certain computer at all."

"Ordinary EDA, even if you pull over a supercomputer like TaihuLight, its performance has an upper limit."

"However, this Yinglong, his source of computing power is cloud computing."

"The computing power of the cloud server can be as big as it is. It can even be said that Yinglong's computing power is unlimited."

Everyone gasped, but then another question popped into their minds.

What is the use of such a large computing power?

If it is aimed at the current scale of integrated circuits, basically the computing power of commercial computers used by enterprises is sufficient.

This question circled in their minds, and after just a few seconds, everyone had an answer.

No wonder.

No wonder the man said before that Ying Long didn't look like a product of this era.

Because its service object is not the current chip design at all!

It is the chip design after ten or even decades later, when chip design enters the sub-nanometer era and the scale of integrated circuits enters the ultra-large era!
Everyone took a deep breath.

This Yinglong EDA is not just as simple as slapping the existing foreign mainstream EDA.

This is not a product of one dimension at all.

If foreign EDAs still use pens to draw on white paper, Yinglong EDA has already stepped into the threshold of 3D printing.

Once this thing is announced, it will set off a storm in the chip industry.

However, before they turned their surprised gazes on Ye Zhou, someone came out from the crowd and said:
"Cloud computing is really strong in chip design, but it's not incomprehensible. I just saw something that I can't understand."

The eyes of the crowd focused on the speaker, who pushed forward and said while sitting on a chair:
"I'm working on EDA algorithms. You see, we have a concept called ultra-large-scale computing task splitting, which mainly uses automated deployment to split multiple tasks."

"You speak human language. There are not many people who work on algorithms here, and they don't understand the mathematical language you speak."

Zhang Yue interrupted his narration mercilessly.

The man smiled awkwardly and continued:

"Let me give you a simple example. A group of people are cleaning the classroom. What we have to do is assign the taller people to clean the windows, and the stronger people to move the table. This saves time and energy. , which is common practice now.”

"However, I glanced at Yinglong's task distribution interface just now, and it doesn't seem to do that."

"Its task is to distribute in real time. After cleaning the windows, the tall person will drag the ground under his feet by the way, and the person who has moved the table will wipe the windows that he can reach by the way."

"Can you understand? This piece of EDA, it has almost reached the point where the computing power is used to the point of pennies. In other words, it has squeezed out all the computing power it can get."

"Not to mention our own algorithm, even the top algorithm of Synopsys, in fact, can only reach about 40% in computing power utilization efficiency, and this Yinglong is 100%"

"That's what I don't understand."

"I don't understand how this thing is made at all. I guess I can't understand it even if I dig out the source code."

"Now I have a very bold, but I think very credible conjecture that if we do reverse engineering and analyze Yinglong's algorithm, some of us may win the Fields Medal."

Everyone's eyes froze for a moment, then turned to Ye Zhou neatly.

At this time, Ye Zhou was actually a little worried.

He knew that Ying Long was very advanced, but he didn't realize that this thing would bring such a big impact to the opponent.

There is no telling how many precious crystals of different fields are condensed behind every scientific and technological achievement. In the eyes of a knowledgeable person, the mystery can be discovered at a glance, but for Ye Zhou who has been directly instilled with knowledge, on the contrary, I feel a little natural.

It's like this thing should be so advanced.

That's why he came to announce this EDA.

The two sides looked at each other for a few seconds, and finally Zhang Yue broke the silence.

"Little brother, did you really make this thing?"

(End of this chapter)