The Millennium Semiconductor Survival Guide.

Chapter 180 In-depth cooperation between Xinxin Semiconductor and ARM

Back at the apartment, the smug Pony Ma didn't know that the Matrix prototype he saw in Zhou Xin's office was one of the few prototypes.

Moreover, the built-in chip of Matrix is ​​not the final product, and there is not even a sample of the final chip. The chip of the Matrix prototype is a handheld computer chip.

The chip is still on the blueprint. In order to design the Matrix chip as soon as possible, Guan Jianying has not stopped from the end of New Year's Day to the beginning of Shenhai.

Since Zhou Xin is in San Francisco, Guan Jianying sometimes has to answer emails at night due to jet lag.

Compared with Silicon Valley, Huaguo's lack of talents in the chip field is what makes Guan Jianying feel the most helpless.

If he is short of money, he can apply to Zhou Xin. Electricity, Internet, land, water resources, Shen Hai can meet his requirements.Only in terms of talents, whether it is chip design or lithography machines, it is too difficult to find suitable people.

Unlike SMIC, which is a chip foundry, SMIC can poach a large number of people from Wanwan.

At this point in time, Wanwan is not good at chip design and lithography machines. The lithography machine is a little better, and he can recruit people from Neon. In addition, Shenhai Microelectronics has a few engineers who have worked on lithography machines. Let’s absorb the cutting-edge Technology can be put to work.

Chip design is really bad. IC design engineers are concentrated in American and Korea. In Korea, Samsung has a certain right to speak in the field of chip design.

These people can't dig at all.

Relying on his more than 20 years of chip design experience in Silicon Valley, Guan Jianying reluctantly recruited a dozen chip design engineers from Silicon Valley, with work experience ranging from three to 15 years.

But this manpower is far from enough, after he knew Zhou Xin's development goals for Xinxin Semiconductor.

Zhou Xin and Guan Jianying talked about it. Xinxin Semiconductor needs to design chips similar to handheld computers in the short term. After this business is completed, it will slowly expand to other fields.

Compared with ordinary computers, the market for handheld computers is relatively small, which also leads to few players in this field in chip design.

There are only Intel, Texas Instruments and Samsung, and IBM, Apple, and AMD have tried to design chips for handheld computers, but they have not been recognized by the market in the end.

Because the manufacturers who are still insisting on making handheld computer chips are all big factories, it is difficult to recruit people from big factories, and they are still recruited to work in China.

As for why not set up a research and development center in Silicon Valley, the office is located in Silicon Valley, it is much more convenient to recruit people.

If you want to recruit people from major companies such as Intel, Texas Instruments, and Samsung, or from Silicon Valley to Zhangjiang, the difficulty factor is about 100, and if the office is located in Silicon Valley, the difficulty factor will drop to 80.

Guan Jianying has mentioned that it will be divided into two steps. First, establish a R&D center in Silicon Valley, adopt the model of two centers in two places, and gradually move the R&D focus to Zhangjiang in the future.

In the first three to five years, the research and development center in Silicon Valley must still be the main focus.

This plan was rejected by Zhou Xin. In Zhou Xin's view, Samsung's chip design plan in his hand is enough to support until around 2010.

Ten years is enough to cultivate a capable chip design team. It took less than ten years for Huawei's HiSilicon Kirin to evolve K3V2 into 980.
Not to mention that matrix can still take the first-mover advantage of mobile phone chips.

If the R&D center is located in Silicon Valley, what is the difference between Xinxin Semiconductor and other multinational giants?
Microsoft and Intel also have R&D centers in China. The R&D centers of these multinational giants in China are still marginalized after 20 years.

Zhou Xin knew that as long as he opened the door to Guan Jianying and let him set up a research and development center in Silicon Valley, he would definitely not care about Zhang Jiang's team. This is the necessity of human nature.

It's like when later generations of Tencent firmly grasped most of the traffic entrances in Huaguo and counted money on the credit book, even the top product manager teams in the past no one cared about user experience anymore.

KPIs replace user experience.

It is human nature that no one will do the hard things when the short-term gains from doing the easy things are greater than the short-term gains from doing the hard things.

Except for this matter, Zhou Xin satisfied Guan Jianying almost unconditionally.

Guan Ying can be called a living fossil in the field of chip design, and he knows all kinds of technical materials like the back of his hand.

After knowing that Zhou Xin's goal is a handheld computer, he wants to place the R&D team in Zhangjiang, and claims that there will be no requirement for revenue and profit within three years.

Guan Jianying made another request:
"Newman, the requirements of handheld computers and desktop computers are completely different. Handheld computers are limited by battery capacity, and the positioning of handheld computers has always been an office tool.

Pocket PCs don’t need to consider entertainment, so unlike the core appeal of chips, performance, mobile phone chips need to consider battery life and power consumption more.

I suggest considering Apple's Newton handheld, which Apple has discontinued but was an absolute success.

The failure of Apple's other business lines has dragged them down from continuing to invest in handheld computers, and they had to cut this product line as a last resort.

Purely from the perspective of chip design and product design, I think Newton is very good.

We can buy the patent of the XScale chip from Intel. This set of chips based on the ARM architecture is the chip used by Apple.

A full set of patents is not necessary and cannot be afforded. Intel will definitely open its mouth. I am sure about buying a patent license.

XScale was not taken seriously within Intel, their best partner Newton was discontinued, and other brands of PDAs gradually turned to PDA chips from Texas Instruments and Samsung.

Intel's top executives have always had disagreements on mobile chips, and there are often voices internally that they want to package and sell the mobile chip business, so if we just buy licenses, the price will not be high, and the probability of their agreement is very high.

The XScale series of chip patent authorization alone can play a role in training the team, and I think the money is worth the money. "

Zhou Xin agreed without hesitation.

Guan Jianying went on to say: "Because of our entire engineering team, there are only 15 engineers who are willing to go back to Zhangjiang to start a business with me, plus my initial engineer team of 16 people, only 5 people have experience in mobile chip design.

Of these five people, only one and a half focus on mobile chip design, and I am that half.

Why buy Intel's XScale chip patent license, because almost all handheld computers now adopt the ARM architecture, which is quite different from the architecture used by our engineering team.

Compared with the x86 architecture, the Arm architecture has lower power consumption and cost. Even if it is difficult to compare with the X86 architecture in terms of performance, the handheld computer still chooses the ARM architecture. What follows is that the ARM architecture chips perform very well in the personal computer market. generally.

The architecture of the XScale chip is StrongARM, which is a RISC processor that supports WinCE3.0-PocketPC system, that is to say, it can become a bridge for switching from other chips to ARM architecture chips, and can train the team. "

In the end, Xinxin Semiconductor obtained the patent license of XScale at a price of 2000 million US dollars.

Intel once took the lead in handheld computer chips, and later developed mobile phone chips along the XScale technology route, which is the second largest smart phone chip in the industry after Texas Instruments mobile phone chips.

In the end, Intel packaged and sold the entire technology to Marvell at a low price of $6 million.

It's also human nature. It's so cool to make money on personal computers. I don't even care about the benefits brought by mobile chips. It's better to sell them to make the company's financial report look better.

Guan Jianying, who returned to China with XScale technical materials, patent authorization and 15 engineers, was apprehensive and excited.

In the days of Shen Hai, Lin Benjian's progress was faster than Guan Jianying's, not only because of the talent, but also because he bought a complete set of lithography machine technology from Nikon.

The lithography machine based on the most advanced 150nm process is two generations behind, that is, the 220nm process lithography machine.

The first prototype of Xinxin Semiconductor has been built, and the prototype is jointly tested by Hua Hong Semiconductor.

For the current Huaguo, the Ministry of Science and Technology of Huaguo has organized the research and development of 863nm lithography machines to be added to the five-year 193 plan.

That is, one generation behind the most advanced 150nm process.

Huaguo did not put all the treasures on the Xinxin lithography machine, even though Zhou Xin showed great kindness.

Experts also have projects to do, and not all experts can be poached by Xinxin lithography machines.

It's just that Huaguo's official 193nm lithography machine is still in the planning, and the production of the prototype will not be completed until 2005, while the 220nm process lithography machine sample of Xinxin lithography machine has come out.

Mass production is planned to begin in the first half of next year.

While mass producing, technology research and development catch up with advanced manufacturing processes.

The Ministry of Science and Technology and Shenhai City are two lines. The Ministry of Science and Technology has its own considerations, and Shenhai City also has its own considerations.

SMIC landed in Shenhai, and Xinxin lithography machine produced the first prototype in Shenhai. The advanced process is 150nm, and 220nm chips still have a broad market.

Not to mention that this is a domestic lithography machine, not to mention that the technology is bought from abroad, and the engineers are mainly neon engineers. The problem is that this is a Huaguo company, and the Huaguo boss is behind it, and it is still produced in Huaguo.

Shenhai has attached great importance to Xinxin lithography machines, whether it is tax incentives or honors at the city level, it can almost give it all.

"Get ready and go to Cambridge. I have already greeted ARM. Xinxin Semiconductor will sign an in-depth cooperation agreement with ARM."

Guan Jianying saw the news that Zhou Xin became the major shareholder of ARM, accounting for nearly 50.00% of the shares.

He admired Zhou Xin for always understating these things. Things that are worth showing off to others seem to be not worth mentioning to him.

ARM does not have the status of later generations. It is only famous in the field of handheld computers. It is different from occupying more than 95% of the smartphone chip market in the future, but it is still a well-known chip company.

And there is only ARM in this business form.

Zhou Xin made things vague, but Guan Jianying has been thinking about what kind of in-depth cooperation it is.

(End of this chapter)