If you want to explain Wang Zhengguo's nonsense proton computer for a while, this difficulty is really not ordinary, compared with the quantum computing mechanism proposed by Richard Feynman in 1982, the proton computer should be regarded as a more high-end level.

Unless it is supported by the three-body star technology, from the perspective of the current human society technology, let alone the development, even if you want to understand it theoretically, it is difficult.

For example, how to explain the expansion of protons in the three-dimensional state to the two-dimensional, and the infinite expansion, carving, and return to the three-dimensional state, Ni Guangnan himself can't understand it anyway, let alone explain this thing in depth.

In the final analysis, in view of the outdated knowledge structure of these scientific research scholars at the ISSCC conference, Ni Guangnan finally gave up using proton computers to burn his brain and said a simpler plan.

The proton computer is to expand the three-dimensional proton to the two-dimensional, and then carry out the engraving and processing of the circuit, and now Ni Guangnan's idea is just the opposite of the proton computer routine, what he wants to do is to make the two-dimensional state of the semiconductor integrated circuit into a three-dimensional structure.

Well, isn't it still very brain-burning to say this, then let's have something that normal people can understand.

"Our current semiconductor integrated circuits are all processed on a flat surface of silicon, first of all, the area size is limited by other computer hardware, in order to enhance the performance, we can only reduce the size of a single transistor with the same area. ”

Then, he motioned to the staff to pull over the small blackboard in the corner, which was indispensable at all times.

First he drew a small single-story bungalow, then he drew a high-rise skyscraper on the side of it based on the same footprint, and after the two stick figures were completed, he pushed it to the front of the podium and pointed to the single-story bungalow next to it.

"In order to integrate more transistors within the original microprocessor area with the same precision in the semiconductor manufacturing process, we first have to abandon the current 2D processing model of building a single-story bungalow in a single plane and instead build a skyscraper, and then add a second, third, and more..."

Of course, this design needs to solve a lot of problems, such as how to connect between layers, which is a problem, and the corresponding heat dissipation, which actually requires us to find a way, but Kaiyang Semiconductor has made a breakthrough in this field, for this new technology, I call it: 3D stacking, which generally belongs to the classification of semiconductor packaging technology. ”

When it is difficult to break through the manufacturing process of semiconductor integrated circuits, turning to the development of more advanced packaging technology such as 3D stacking can actually achieve a huge leap in technology.

Is this kind of thing feasible?

Needless to say, in the 2005 planning and development document, the US Defense Advanced Research and Planning Agency directly positioned 3D stacking technology as one of the three basic technologies for the future defense technology transformation of the United States, which will become the core technical support platform of the eight major national defense strategies of the United States, and will have a revolutionary impact.

Therefore, things like 3D stacking are destined to be extraordinary in the future.

Now Chief Engineer Ni Guangnan has revealed the 3D stacking packaging technology, in fact, this thing has a certain amount of technology accumulation on Kaiyang Semiconductor, but it has not yet reached the time of full outbreak.

According to the plan, if Kaiyang Semiconductor cannot break through the 0.5 micron process technology in a short period of time, the company will directly use 3D stacking technology to improve the performance of the processor.

Before 2000, the number of transistors of microprocessors has not reached a certain level, and the so-called heating problem is not too significant, plus the processor of RISC instruction set is still relatively well controlled in terms of power consumption and heating.

However, with Lin Benjian joining Kaiyang Semiconductor and leading the team to quickly overcome the process problems of the 0.5 micron process, in the Kaiyang 915 processor project, he finally did not plan to adopt 3D stacking so early, and chose traditional solutions to control risks.

The corresponding technology developed before has been transferred to backup, and it has also continued to invest in perfection in case it is needed, and if the process technology is stuck one day, it may be possible to use it to fight fires.

The use of 3D stacking technology to develop microprocessors will be plagued by the problem of heat dissipation caused by the spacing between different stacking layers is too small in the future, but if it is used in storage devices, the most difficult heat dissipation problem will not exist, and based on the working nature of memory devices, it will not have the possibility of high heat generation at all.

In 2011, the world's major memory equipment companies at that time turned to 3D stacking technology, and quickly launched a variety of memory chips with more than 30 or 40 layers of stacking layers, which could be ten times the memory capacity and read and write speed of traditional DDR3 while occupying the same area size.

Before Wang Zhengguo's rebirth, the 3D NAND memory chips made by using 3D stacking technology had been launched and rapidly updated and iterated, which was completely expected to directly eliminate the life of NAND/DRAM.

As for why he announced this kind of good thing today, there is nothing to say, Kaiyang Semiconductor has already got the patent applied for, and the announcement of this technical route at this time is clearly to turbulent the entire industry and disrupt the current stable technological development route, so as to fish in troubled waters.

Anyway, once the 3D stacking technology is patented, there will be a publicity period, and those top technology companies, which are not always paying attention to the trends of the patent offices of various countries, and what new technologies will soon know as soon as they come out.

"In my opinion, 3D stacking technology is promising in the future, it is the best solution to solve the failure of Moore's Law at the dawn of the 21st century, and in the process of scientific and technological development, it is essential to move from 2D to a higher level of 3D world. ”

The keynote speech lasted one hour and 15 minutes, the whole process was high-energy five urine points, and all kinds of revelations continued, which can be described as the highest quality keynote speech in recent years.

As the stage announced the end of the speech, Ni Guangnan turned around and left the audience, and the scene did not know who took the lead in applauding first, and in less than two seconds, thunderous applause resounded throughout the audience, even the Japanese scientific researchers who were originally full of disdain had to be convinced, this speech is indeed very standard.

Several IBM bigwigs in the front row looked at each other, and their feelings were really indescribable, originally preparing to come up with a serial plan of "borrowing a knife to kill people, crossing the river and demolishing the bridge", but now it seems that the quality of this bridge is too good for them to dismantle.

"It's very interesting, Kaiyang Semiconductor's ability is very good, no wonder Lin Benjian will go there to work, or this is the next Intel company, and this Ni Guangnan, as Gordon Moore of Kaiyang Semiconductor, he is a wise man who has to be admired. ”