Xueba: I study hard in the wasteland world

Chapter 285 Chapter 283, Computer language logic under “np=p”

As various news spread, the Fruit Candy Computer R&D Center has undoubtedly become the focus of the world's attention.

"Although the theory is very complete, I didn't expect that it would be so troublesome to put this thing into practice."

In the laboratory, Su Yu looked at the microelectronic room-temperature superconducting materials prepared in the laboratory through a microscope.

Su Yu has already led the entire team to complete the entire theory of microelectronic room temperature superconducting materials. However, even with the support of the theory, the entire team spent a long time and a lot of money to complete the material laboratory. preparation.

No wonder room temperature superconductivity has been around for many years, and the only areas it has been transformed into applications are power transmission and high-speed rail.

The system involved in this thing is too huge. Until now, China is the only country in the world that can play with room temperature superconductivity.

Of course, the advantage of this thing is also its rolling type, "0 resistance". The simplest point is that the chip made of it does not have any heating problems.

Manufacturers in the field of electronic product radiators can declare bankruptcy.

"Academician Su, although we have overcome the material difficulties now, we have encountered a lot of troubles in the actual manufacturing of new chips."

When Su Yu completed the observation of room temperature superconducting materials in microelectronics, a senior engineer in the team handed Su Yu a project report.

"Really? Let me see." Su Yu took the report.

The engineer pointed to the conclusion at the end of the report and said to Su Yu: "Academician Su, look here, it means that according to the traditional chip design, our project has no problem, and it can completely follow the traditional chip design process. Complete the production of room temperature superconducting chips."

The engineer pointed to the conclusion of the last paragraph of the report, frowned, and said, "But here, I made it according to your new chip design plan, and there were a lot of problems in the operation of the chip."

Su Yu became serious and began to read the report from the beginning.

This team is made up of elites from all over the world. It can be said to be the top computer research and development team. They can handle most of the problems by themselves. Now they have reported the problem to Su Yu, which shows that this problem is What deserves Su Yu's attention is what they can't solve.

Su Yu read the report carefully.

According to the report, using traditional chip design methods and adding SYN inference accelerator, the integration of room temperature superconducting materials on the chip can be successfully completed.

In other words, with the SYN inference accelerator created by Su Yu, it is enough to directly replace the materials in the traditional chip with room temperature superconducting materials.

With this, the room-temperature superconducting chip has actually been completed. The performance of the new chip is dozens or even hundreds of times that of the traditional chip, and there is no need to consider any heat dissipation issues.

But now, in this project, a chip designed by Su Yu is completely different, a completely new computing logic.

If the chip manufacturing is carried out according to the traditional process, then under this operation logic, many technologies will lead to catastrophic accuracy loss based on the SYINNH system.

But this is a completely new field, and the project team has no way of starting, so they can only report it to Su Yu.

Su Yu thought for a while and said seriously: "I'll go back and think about it first and give you an answer in a month. If I don't make any breakthrough, I will follow the traditional path and complete the production of room temperature superconducting chips."

"Is this road really dead?"

In the study, Su Yu turned the pen in her hand and thought with a solemn expression.

Su Yu's original intention for this room-temperature superconducting chip project was to provide hardware support for solving the np-complete problem, thereby establishing a new, faster computer computing logic. Let’s take a simple multiple-choice question in which you choose the correct answer from four options A, B, C, and D as an example:

Traditional computer logic starts verification from A, first to see if A is correct, then to see if B is correct, and then to see if C is correct.

A question with only four options may be easy to answer. Just verify four options, but what if there are thousands, even hundreds of millions, or infinite options?

Even if the computing speed of the computer is soaring, it still has no way to solve the problem when faced with endless options.

However, what Su Yu is looking for is an idea of ​​​​"np complete problem". To simply understand, it is the four options A, B, C, and D. You do not need to verify them one by one. I will tell you directly that "B" is correct. Yes, the computer will directly verify whether what I said is correct and whether B is correct.

This kind of thinking is incredible at first glance.

It is precisely because of this inconceivability that most computer scientists around the world believe that the "NP-complete problem" is not true.

If it is rooted in traditional operational logic, Su Yu also agrees that the problem is not valid. Therefore, Su Yu wants to find a new logic that can make the NP-complete problem valid.

Su Yu has been doing this work for a long time with various partner universities across the country, but the entire project has not achieved much results.

The current room temperature superconducting chip is actually not the result of this project, but more the result of the room temperature superconducting project.

In particular, the team's current verification shows that the chip designed by Su Yu with a new computing logic cannot run at all.

Su Yu tightened her head and carefully browsed all the information on the entire project.

If he hadn't made any breakthroughs, the "International Major Basic Science Project" he was hosting might have been shelved.

Of course, room-temperature supercomputing chips produced according to traditional design ideas will successfully achieve breakthroughs.

But Su Yu was still not satisfied with this.

Because even if this room temperature superconducting chip breaks through, it is only a technical breakthrough. It only improves performance and solves the heat dissipation problem. There is no practical breakthrough. This is just "1+1=2" before. Now This room temperature superconducting chip can be regarded as "2+2=4".

"Huh~"

And when Su Yu sighed and was about to collect the materials and get off work, he suddenly saw an inconspicuous submitted paper in the "Catalogue of Submitted Articles from Partner Universities for Review".

Su Yu opened this paper with great interest. When he had no inspiration, he always read whatever he saw.

This paper introduces an SLH accelerator optimization designed for efficient execution of converters.

The key is that the accelerator proposed in the paper implements per-vector scaling quantization, which adopts an independent scaling factor for each element vector. In this way, the fusion operation of the controversial Su Yu algorithm can be realized.

Su Yu became interested in this paper. Although the paper was not very rigorous and full of loopholes, it was immediately obvious that the author was a "novice".

But in this paper, the author's wild and even unprovable bold ideas attracted Su Yu.

Su Yu sat down again, picked up the pen, and began to read the paper seriously.

(End of this chapter)