novelslam Chapter 806 Kneel Down! Completely On My Knees! You Can’T Do It Without Kneeling!
Chapter 806 Kneel Down! Completely On My Knees! You Can’T Do It Without Kneeling!
"There is no doubt that this will be another victory for Moore's Law!"
"Not bad! Not only 22nm, 14nm, and even 7nm, there is a lot of room for imagination."
Everyone was very excited.
actually.
A small amount of shrinking transistors is not a huge gain.
From a physical perspective.
Although transistors have shrunk, the power consumption per unit area has not decreased.
This has become a serious problem for transistor shrinkage.
That is, the theoretical calculation is an ideal situation. In fact, the power consumption not only does not decrease, but increases with the increase of integration level.
Around 2000.
According to the development of Moore's Law, people have predicted.
If there is no technological progress, transistors will continue to shrink at the same rate, and by around 2010, their power consumption density will reach the level of rocket engines.
It is impossible for such a chip to work properly.
Even if this level is not reached, the temperature is too high and can affect the performance of the transistor.
As of now.
The industry has still not found a solution that can completely solve the problem of transistor power consumption.
The actual approach is.
On the one hand, the voltage is reduced (power consumption is proportional to the square of the voltage), and on the other hand, the clock frequency is no longer pursued.
therefore.
After 2005, CPU frequency has almost stopped increasing, and performance improvement mainly relies on multi-core architecture.
This is also known as the "power wall".
At this moment.
All the experts and scholars present at the meeting were very excited.
After a period of excitement, everyone began to return to rationality.
One person adjusted his glasses and said: "Even without supercomputer simulation verification, I dare to make a judgment now. This is definitely feasible."
"But the hard part is realizing this idea."
"Why?"
"Because the part of silicon that stands up, that is, the silicon used as the channel, is really too thin, less than 10nm."
"This thickness is far smaller than the minimum size of a transistor, and it is also far smaller than the minimum size that can be lithographed by the most sophisticated lithography machines currently."
"therefore."
"Another difficult problem before us now is how to come up with this plan and turn Pakistan's vision into reality!"
As soon as these words came out.
Many people at the scene nodded.
The problem of how to break through the 22nm process node can be considered solved for now.
But they had to face new problems.
That's how to make it.
Not only did it have to be done, but it had to be done well, which became a real problem.
There is no way to avoid this problem.
But the essence of the problem lies in the photolithography machine.
Because even the most sophisticated photolithography machines currently can’t handle it.
not to mention.
Longguo is so outclassed by Europe and the United States in terms of lithography machines that even if you have money, you can't even buy the most advanced lithography equipment.
all of a sudden.
Everyone who was originally excited seemed to have been poured cold water on them.
Said without politeness.
How to make it is even more difficult than the process problems that troubled them before.
Because this design is actually not difficult to think of.
The hard part is being able to do it.
Seeing everyone frowning and thinking hard, Lu Yi smiled and spoke again.
"Actually, it's not as difficult as you think."
As soon as the words left his mouth.
Swish, swish, swish!
Everyone looked over.
Lu Yi picked up the mineral water at hand and took a sip, and said slowly: "A detailed explanation requires many process flow charts, so I won't go into details here."
"The basic principle is that this part of the silicon is not photoetched, but grown.
As soon as these words came out.
Everyone was shocked.
Everyone looked puzzled.
Growing out?
How long?
No one asked.
Because everyone knows that Lu Yi will definitely talk about the core points next.
"``It first uses ordinary precision photolithography to carve out a bunch of 'frames', and then deposits a layer of silicon, so that a very thin layer of silicon will grow on the edge of the frame, and then uses selective Etching removes the excess material, and what remains are these standing, ultra-thin pieces of silicon.
Deadly quiet!
No one expected that Lu Yi would solve the problem that they thought was difficult to reach the sky in just a few words.
Oh, does this make them look too stupid?
After a while.
"Kneel down! I am completely kneeling down. This IQ is so overwhelming!"
Yu Chengdong let out a long sigh.
His words also caused everyone present to nod.
It's really a bit of bullying for their lack of intelligence.
In their eyes, it is an unsolvable problem.
But with Lu Yi (Zhao Hao), it was solved easily.
This is really an IQ crushing!
They knelt!
Completely kneeling!
It’s impossible not to kneel down!
If.
Lu Yi previously proposed how to solve the difficult problems blocking the 22nm node process, and the ideas he put forward made them marvelous one after another.
So this moment.
Lu Yi's ideas on how to turn designs into reality are truly astonishing.
Said without politeness.
What Lu Yi said here today will definitely lead chip semiconductor process technology and take a big step forward.
What really lies ahead is a smooth road.
If chip companies like Intel hear what Lu Yi said today, they will definitely kneel down and call them daddy.