While HiSilicon and Huaxin were busy designing and making breakthroughs in chips, Xu Chuan was also studying these mathematical problems related to chip design in the villa.

Compared with the CPU architecture and GPU architecture, the NPU neural network architecture is an emerging computer field.

It was first proposed by psychologists and neurobiologists to be applied to computers by simulating the structure and operation of nerve cells in the human body or other living brains.

Neural network architectures are more akin to 3D nature than originally flat computing architectures, and often provide a relatively simpler way to solve complex problems.

As a result, neural network architectures have attracted more and more attention from the computer industry in recent years.

HiSilicon and Huaxin chose this piece to make a breakthrough, on the one hand, they wanted to overtake in corners, and on the other hand, they were helpless.

No way, compared with those Western countries and enterprises, Huaguo started too late in the basic research of CPU architecture and GPU architecture, and it is far less than the strong experience and strength of others.

And, of course, the most critical patents.

Relying on the moat of patents, companies in Western countries such as Qualcomm, Apple, and Google have almost completely monopolized this field.

If you want to get around, you can only find another way.

That's why HiSilicon and Huaxin chose to start with the NPU neural network architecture.

Although this architecture is still based on the ARM architecture researched by ARM, it can pave the way for future independent research and development.

Flipping through the information on the computer, Xu Chuan carefully studied and analyzed.

Generally speaking, neural network architecture mainly uses a variety of mathematical models to describe and simulate the biological nervous system at different levels from different perspectives.

Representative network models include BP network, RBF network, Hopfield network, self-organizing feature mapping network, etc.

Compared with the traditional CPU architecture and GPU architecture, he started this architecture faster.

After all, mathematical models and modeling, as well as the mathematical formulas behind them, are one of his research areas.

For example, the RBF neural network architecture (radial basis function network architecture) is a kind of forward network constructed based on the function approximation theory.

The learning of this kind of network is equivalent to finding the best fitting plane of the training data in a multidimensional space, and each hidden neuron activation function of the radial basis function network constitutes a basis function of the fitting plane.

To explain it mathematically, the principle is actually very simple.

That is, find a function g in a selected class of functions so that it is an approximate representation of a known function in a certain sense, and find the error caused by the approximate representation of g. ”

This is one of the functional foundations of high school math.

For Xu Chuan, the research on these things was initially just to solve the problems of HiSilicon and Huaxin, but as time went on, he became more and more interested in this kind of neural network.

This kind of structure, which starts from the mathematical mechanism and uses the model to complete the description and simulation at different levels, has many applications in other fields in addition to chips.

For example, artificial intelligence, automatic control, robotics, statistics, and so on.

These are all areas of considerable interest to Xu Chuan.

Needless to say, artificial intelligence is the trend of the future.

Automatic control and robots are one of the core key points for the progress of social productivity.

As for statistics, it seems that it is an applied mathematics that mainly works for other fields.

But in fact, it can also be applied to the field of mathematics itself.

Of course, he is more interested in the statistics and analysis of data in the field of high-energy physics or materials science through neural networks.

The latter two are quite large amounts of data and require more robust and simple mathematical statistics and analysis methods.

And after studying these days, Xu Chuan keenly feels that this kind of neural network seems to be more suitable for quantum chips and quantum computers than ordinary silicon-based chip computers.

If the traditional silicon-based chip composition computer is considered to be 2D or 2.5D, then the quantum computer is 3D, which is inherently beyond a level in dimension.

In fact, from the perspective of biology, the neural network computing method can be seen from the perspective of biology, which is inherently a three-dimensional architecture in 3D.

Of course, this statement may be somewhat inaccurate.

But he does feel that it could be more suitable for quantum computers.

It's just that there is no complete quantum computer available to test this idea for him.

Solving the problems of HiSilicon and Huaxin is not something that can be done in just a few days, and while studying and researching, Xu Chuan also devotes part of his energy to other work.

The days passed like this, and in the blink of an eye, the Lantern Festival arrived.

In the villa, Tang Sijia knocked on the door of the study.

After a while, there was movement inside, Xu Chuan walked out, glanced curiously at the life assistant in the white coat, and asked, "Is there something wrong?" ”

It was the first time he had seen the life assistant in such a doctor's uniform, which looked somewhat similar to the white coat in the laboratory.

Tang Sijia carried two bags and said concisely: "Today is the date of your quarterly physical examination, I will take samples and do a physical examination." ”

"Oh, okay, wait a minute, I'll save the data on my computer." Xu Chuan suddenly understood and nodded.

Since he was assigned a life assistant last year, physical examinations have become the norm for a while.

At first, he would check it every month, but then he felt that it was too much trouble to come once a month, and wanted to change it to once a year.

After all, ordinary people generally only have a physical examination once a year, but Tang Sijia argued with him, and finally decided to do a simple physical examination once a quarter and a detailed physical examination once a year.

Counting the time, today is almost the first quarter.

After drawing two samples of blood for physical examination, Tang Sijia first saved the sample, and then handed over a tool.

"This is for urine collection, do you need me to help you?"

Xu Chuan coughed lightly and took the tool: "I'll do it myself." ”

How can others help with this kind of thing?

After handling the physical examination, Tang Sijia checked and recorded the physical information such as blood pressure, weight, height, eyesight, hearing, etc., and after packing up the tools, he handed Xu Chuan the other bag he brought.

Xu Chuan took it casually, opened it and took a look, it looked like a lunch box?

"What is this?"

Tang Sijia smiled and said, "Tangyuan, isn't it the Lantern Festival today, I guess you won't remember it, so I asked Aunt Lin to make a reunion and bring it over." ”

Xu Chuan took out the lunch box and opened it to see that it was really white and moist rice balls, steaming.

Closing the lid, he smiled and said, "Thank you." ”

Eating glutinous rice balls during the Lantern Festival is also a habit in the south, but he hasn't eaten it for a long time.

After all, at the point of the Lantern Festival, he has basically left his hometown.

When he lives alone, he will not deliberately remember this day and order a rice ball.

With the rice balls, Xu Chuan turned around and entered the study.

On the desk, the phone was ringing, and it was the president of NTU calling.

"Academician Xu, the Lantern Festival is healthy." The phone was connected, and the voice of talking about Shao Yuanle came from the other end.

Xu Chuan responded with a smile: "Have fun and have fun, and I also wish the principal good health." ”

After a few words of greeting, Tan Shaoyuan asked with a smile: "I don't know when Academician Xu is going to return to Jinling." ”

Xu Chuan: "I've gone back to Jinling, is there something wrong?" ”

Tan Shaoyuan: "That's the case, this time I called you, mainly because at the end of last year, you proved the perfect unified framework theory of strongly correlated electronic systems. The school has received many inquiries from physicists at home and abroad, all of whom are asking when you will explain the opening report. ”

"So I want to ask you when it's convenient, and the school should prepare in advance."

Hearing this, Xu Chuan thought for a moment and said, "Then in ten days, I have some other things to do here for the time being." ”

These days, he has become familiar with a lot of things related to neural network architecture, and he plans to spend the next ten days to solve the problems of HiSilicon and Huaxin.

Although the time is short, it may not be possible to make the optimal solution, but judging from the previous exchanges, it is quite urgent to do so.

In that case, let's solve the problem first. As for optimization, you can do it again when you have time in the future.

This kind of thing can't make a big fat man in one bite.

Tan Shaoyuan: "No problem, just take your time as the standard, the school can do it at any time." That's it, the school will send out the notice and announcement first, and get ready. ”

Xu Chuan nodded and said with a smile: "Okay, it's just troublesome to the school." ”

Tan Shaoyuan: "Hey, how can this be said to be troublesome?" You must know that Mizuki Peking University can't ask for it. Then I won't bother you much today, Academician Xu, and we'll talk again when we have time. ”

"Okay."

Hanging up the phone, Xu Chuan looked at the phone with a black screen and shook his head with a smile.

Speaking of the unified framework of the strongly correlated electronic system, from the year before the year to the present, almost a month has passed, and it is estimated that many people in the physics community can't wait, and if they can't find him, they can only find NTU.

But after so much time, it's time to give a response to the outside world.

PS: The second is to ask for a monthly pass

(End of chapter)