Han Yan and Chen Yan only stayed in the capital for two days. Zhou Buqi accompanied them to watch a Peking opera, listen to a cross talk, and went to the China World Trade Center to buy a lot of things, spending a total of more than 400,000.

Especially two LV bags with a retail price of 160,000 yuan, which made Chen Yan dumbfounded. Teacher Han Yan was so frightened that she repeatedly postponed it and refused to take it at all.

After Zhou Buqi's persistence, she accepted it gracefully, and the smile on her face stayed with her all day long.

That morning, Zhou Buqi arranged a flight to take them back to Northeast China.

Go back to the company and deal with some important things.

Shen Xiangyang's report on AI chips has been typed up. It is a 13-page A4 paper. The various data and technical directions listed above also clearly explain the cutting-edge ideas and future expectations of AI chips.

There is a lot of content that I can’t quite understand.

But the general content is understood.

The so-called AI chip is actually an AI accelerator, a module specially used to handle a large number of computing tasks in artificial intelligence applications. It can exist as a separate chip. In the future, when the technology becomes more powerful, it can also be integrated into a general-purpose chip.

Based on a foresighted case, Zhou Buji had a deeper understanding of AI chips - face recognition.

Facial recognition is not a cutting-edge technology. It appeared 20 years ago. It has been widely used since the early 1990s. At the beginning of the 20th century, my country's public security departments were able to arrest criminals based on facial recognition technology.

But why didn’t smartphones really start to use facial recognition technology until iPhoneX?

Because it was not until this time that Apple's A-series chips finally had bionic functions, that is, the AI ​​chip technology was integrated into the A-series chips.

The so-called face recognition is a type of artificial intelligence. It is a kind of computer-inspired judgment that requires a huge amount of calculation. This problem is difficult for general-purpose chips on the market.

Even if some Android phones with face recognition appear, they are fake. They are 2D recognition, which does not require a real face and can be unlocked by taking a photo of the face.

3D structured light for face recognition is not too difficult, and there are many companies around the world that can do it.

The difficulty is the bionic chip.

There is only one Apple in the world.

With the strong support of AI processing technology, the computing power and processing power of Apple's A-series chips have been greatly improved. The 4-core chip can easily beat the 8-core chip of Android.

Now in 2008, AI chips are still the most cutting-edge field and a new track, and technology companies around the world are groping forward.

According to Shen Xiangyang's report, the industry has never been able to reach a consensus on the future technical direction of AI chips and is divided into several major schools. First, the research and development of algorithm models such as deep learning are not yet mature, and secondly, the basic theory of AI is still There is a big gap.

In other words, the AI ​​chip is truly a "software and hardware integration".

The soft level is the AI ​​algorithm at the code level; the hard level is through the chip design of the integrated circuit. The chip design method and the application of embedded language are based on the AI ​​algorithm at the software level.

The two must be deeply integrated and highly unified, and the software technology must be demonstrated through hardware, in order to develop the most powerful AI chip.

Going deeper... AI chips will involve wider and more complex fields, not only the integrated circuit semiconductor industry, but also have to be deeply integrated with Internet technology and software technology!

Top chip companies such as Qualcomm, Intel, Samsung, and IBM have their strongest strengths in the "hard" aspects, but they are lacking in the "soft" aspects.

They are not an Internet company or a software company, and they have technical shortcomings.

The core of an AI chip is AI. You must first have AI technology, and then you can carry out related chip design based on AI technology. The most powerful AI chip must have the best AI technology and the best chip design.

The same example is cloud computing.

IBM, Cisco, HP and other server giants cannot develop cloud computing platforms, and the market has been divided up by Internet giants. Because the core of cloud computing platforms is the virtual capabilities and scheduling capabilities of the cloud operating system, which is a "soft" technology. Millions of units

The distributed connection of the server is just a supporting facility.

AI chips, like cloud computing, are a field where the Internet dominates hardware. The leader should not be chip companies, but Internet companies.

For Zhou Buqi and Ziweixing, now is an opportunity!

The artificial intelligence laboratory of Ziweixing Research Institute is headed by Shen Xiangyang, and it studies AI technology. It can completely expand the technology and combine AI technology with chip design.

Although it requires combining two technologies, this project is not necessarily more difficult than the 4G universal chip that Li Yinan is working on.

Because general-purpose chips have universal attributes, they have more functions, are more precise, and are more complex in design. AI chips have a single function, so the complexity of chip design is not that high.

The biggest shortage in the country is talents in the chip field. In comparison, there are still many AI-related talents. From this perspective, making AI chips is more likely to be successful than making general-purpose chips.

Moreover, this is a new track. Once it is built, it will be the best in the world.

Zhou Buqi's mind immediately jumped out of Shen Xiangyang's report.

The first step is the general-purpose chip; the second step is the AI ​​chip; the third step is the combination of the general-purpose chip and the AI ​​chip to launch a bionic chip similar to Apple's A-series chips.

Obviously, for most of the current chip giants, what they lack is the ability of AI chips.

Ziweixing has a good layout in the field of artificial intelligence. It not only has Shen Xiangyang, a top scientist, but also hired Land Machine from Yahoo. It can be said that it is full of opportunities.

It was almost 6 o'clock in the afternoon, and Zhou Buqi called Shen Xiangyang, "Old Shen? Come and have a chat if you have time."

This was a big deal. Shen Xiangyang put down the meeting in his hands and ran over immediately, "Is it about the AI ​​chip?"

"Yes, I have decided that this project must be done!"

"Really?"

"Well, it must be done, this is our chance."

Shen Xiangyang looked at him deeply, "This direction is more risky than cloud computing, and the country lacks relevant supporting facilities and enough talents. If you want to make AI chips, you need people who are proficient in both soft and hard technologies.

Top talents must understand both artificial intelligence and chip design."

"And you?"

"I'm not good at it," Shen Xiangyang shook his head, "I lack research on the direction of chip design. But Li Yinan is okay. His Internet knowledge is not top-notch, but it's pretty good."

Zhou Buqi waved his hand and said, "Anyway, this matter is settled and will start immediately. On the Ziweixing side, we will make AI chips; on the Asda side, we will make 4G universal chips. Gemini, both are indispensable."

Shen Xiangyang had already guessed this result.

This is in line with his understanding of Boss Zhou.

After a moment of silence, he said slowly: "This will be a big project. The country not only lacks related industries, but also lacks related talents. I think it needs to be divided into three steps."

"you say!"

"First, let Li Yinan join Ziweixing's technical committee and serve as the vice president of Ziweixing Research Institute to coordinate and improve the technical cooperation between the two parties."

"good!"

Of course this is no problem.

Although they have changed their names, Ziweixing and Aisida are truly a family. Although they are not a group company, they are a holding relationship.

Just like Alibaba and Ant after the split, they are two companies from the perspective of equity structure, but from the perspective of management system, Ant is still within Alibaba's organizational system.

Shen Xiangyang said: "Second, cooperate with Tsinghua University. I talked with Li Yinan, and he also went to the Institute of Microelectronics for inspection. Making chips is a long process, and it would be good if the results can be produced in five or ten years. Domestic If there is a lack of talents, then we will train them from scratch. We can design professional courses according to our needs."

Zhou Buqi twitched the corner of his mouth.

Depend on!

Let’s assign tasks again.

This kind of thing must be coordinated by Boss Zhou. Not only must we communicate well with the school, but we must also negotiate with the central and education authorities.

To establish a new major in Tsinghua University, and with this level of high technology... it will require an investment of tens or even tens of billions at least. To invite the top professors at home and abroad, it must be promoted by the power of the state.

Shen Xiangyang also knew that this was very difficult, and said: "If we cannot set up undergraduate majors, we can set up master's and doctoral majors to absorb undergraduate students majoring in computer science and integrated circuits."

This is much simpler.

There is no need for the state to promote it. Ziweixing can provide it with some money. Zhou Buqi looked much better and nodded lightly, "Well, we will organize a technical delegation later to contact Tsinghua University."

Shen Xiangyang said: "The third point, which is the most important point, is to develop in Silicon Valley and establish a biocomputing laboratory and an AI chip laboratory in Silicon Valley. Bring back the research results there and compare them with those in China. Research docking can also be applied in teaching.”

To put it bluntly, it's money.

Ordinary engineers can only do some ordinary technical positions, and those who work in scientific research laboratories are all top engineers, which means that the per capita salary in this type of laboratory is very high.

Preliminary assessment shows that the annual salary and welfare expenses of a scientific research laboratory with a team of 50-60 people in Silicon Valley will reach 40 million to 50 million US dollars.

Previously, it had been planned to establish four laboratories at Ziweixing American Research Institute: cognitive computing laboratory, artificial intelligence laboratory, deep learning laboratory, and big data laboratory. Now there are also biological computing laboratory and AI laboratory. Lab on a chip.

Annual expenditures for this institute may reach US$300 million.

Zhou Buqi has become numb to money figures and so on, and does not realize there is much difference between US$200 million and US$300 million.

As long as the trend of mobile Internet is captured, making money will be faster than robbing a bank.

This amount of R&D funding is nothing at all.

Artificial intelligence is the next wave of huge opportunities after the mobile Internet.

Get in as early as possible.
Chapter completed!