The Millennium Semiconductor Survival Guide.

Chapter 189 A1E Chip
"The gap between NewPay and X.com is not only reflected in market capitalization, but also in user recognition.

After entering the 21st century, more and more users have become accustomed to electronic payment, and are used to binding credit cards and electronic payment accounts, and then make payments through electronic payment accounts.

More importantly, people seem to default that electronic payment is NewPay, and NewPay is electronic payment. People will not think of other electronic payment software. Even if X.com is at the top and NewPay is at the bottom, they still pay. It will drag the webpage to the bottom, find NewPay and complete the payment.

The gap in market share is likely to be smoothed out, and the real problem is to change consumers' perceptions.

NewPay did an incentive plan during the early promotion. How high is your incentive plan converting users?After receiving the incentive money, the users did not immediately transfer the money to the bank card, but used NewPay as the payment method and kept using their registered NewPay account. asked Musk.

Peter Thiel pulled out a memory from more than two years ago in his mind, and then said: "I remember the conversion rate was about 50.00%.

However, after the launch of Quora Electronic Payment and Quora Golden Selection, these dormant accounts were activated one after another.

At the end of last year, I read the data about this. More than 80.00% of the NewPay accounts that received the incentive money that year are still in use. "

There is no way to search for this kind of data from the outside world, or to inquire about it from employees in non-core positions.
Musk felt extremely frustrated because the number of X.com was too low: "Our conversion rate is only 30%, and the statistical caliber only requires users to complete a payment on X.com after receiving the incentive money.

If the time dimension is extended, this number will be even lower. "

Peter Thiel said: "We sometimes face the same dilemma. The marginal effect of pure monetary rewards is particularly obvious. After a certain level, the effect of using incentive money is very limited.

There are even many users who specially register accounts to defraud incentive money. Fortunately, we have set a threshold and require the use of mobile phone numbers.

Later, we canceled this plan, and now we all adopt the method of issuing coupons. Users can receive coupons after registration, and the received coupons will offset a certain amount after payment is completed.

Moreover, our coupons are all designated consumption scenarios. Users can choose from Riot Games, Wal-Mart, catering, and Internet services to choose a designated scenario.

After they complete the consumption, we will give them coupons with a lower amount, so as to induce users to pay through NewPay step by step, and then cultivate users' new payment habits.

This mechanism has shown good results in the promotion process. "

These are not commercial secrets. As a direct competitor of NewPay, Musk knows all the functional details of NewPay. A successful entrepreneur knows best the competing products of his products.

"This game can only be used by NewPay, and square also adopts a method similar to NewPay, which is not recognized by users at all.

The reason is that they support too few payment scenarios.

I didn’t know until now how important the payment scene was, but I guess Newman realized the importance of the payment scene when he was doing Angry Birds.

X.com is just a single electronic payment software, and we had no concept of payment scenarios until March 2000.

At that time, we were still thinking about how to provide services to users. A single service built on X.com, users could send money via email and the Internet. At that time, we developed a function similar to Confinity, which means that users When using a handheld computer, if two handheld computers have infrared ports at the same time, then they can transfer money through x.com.

These unattractive features cost X.com a lot of work, and we put most of our research and development efforts into it.In fact, looking back now, what we needed to do most at the time was to discuss cooperation with major Internet companies.

X.com wants to become one of the payment channels for Internet companies, which is really useful to users.The function of the electronic payment software itself is not important, what is important lies in the scenarios it can be applied to. "

Musk’s tone is full of emotion. If Zhou Xin hears it, he will tell him silently in his heart: No, the function of electronic payment itself is also very important, but it needs to be based on enough users.

Peter Thiel is also very touched by this: "This is what Newman often said to me when I first took over NewPay, we need to focus on solving user problems, not on the product itself.

Helping users solve problems makes more sense than blindly adding features to a product. "

Peter Thiel had a very rich resume before founding Confinity. He graduated from Stanford Law School or taught at Stanford Law School, established an investment company on Wall Street, and then founded the electronic payment company Confinity.

He has met many entrepreneurs and product managers. Everyone’s potential goal is to help users solve problems, but only Zhou Xin can find the best path.

Because users themselves often don't know what they want, they only know what they don't need. What the Internet product manager has to do is to concretize the abstract needs of users.The more users the product has, the more difficult it is to do this.

Too many users means that it is difficult to adjust to everyone, and it is difficult for you to satisfy the vast majority of people.Later generations of national-level mobile apps in China, from WeChat to QQ, are maintained and upgraded by Tencent's core team and the best product managers. They are updated every year, and the user groups have more and more opinions.

"It is also under such a top-level design that NewPay has been able to grow at a high speed.

We have currently deployed servers in 5 major cities across the United States in order to meet the needs of users across the United States, and it is expected to expand to 10 cities this year.

After the introduction of Mphone, the demand for network bandwidth and low latency is higher. "

After entering 2001, Peter Thiel felt more and more that it was the right choice to join NewPay at that time. The development of NewPay made him see the hope of unifying the electronic payment market.

In fact, NewPay has now dominated the electronic payment market in the United States, but most countries in the world still have a hesitant attitude towards electronic payment.

Musk said: "This is slow step by step, the investors behind X.com have lost confidence in X.com to catch up with NewPay.

Their goal for X.com has become to keep the current second position and compete with NewPay in the blank areas of electronic payment such as Europe and Asia.

However, compared with NewPay, I found a more interesting entrepreneurial direction, that is Weibo.

Weibo is really a very interesting product. The combination of it and Mphone has the effect of one plus one greater than two.

Since I experienced Mphone, I have to send more than ten microblogs every day.Recently, I have always had an urge to sell the shares of X.com to others, resign from the position of CEO of X.com, and then make products similar to Weibo. "

This couldn't be more normal. After becoming famous, Musk would rather pay a high premium to buy Twitter after he became famous. The ease of buying it is like buying a toy.

And he doesn't like to make electronic payment software himself. Musk's earliest vision for X.com is online banking. Users save money on X.com and then do some banking business.

The problem is that this is something to be done after the popularization of electronic payment. Users are not even used to electronic payment, so how can they accept borrowing and handling some financial services in online banking.

This is not realistic.

Therefore, the development of X.com and what Musk wants to do are drifting away. Even without NewPay, Musk will hand over the CEO position to Peter Thiel after the merger of Confinity and X.com.

As for Weibo, it is so fresh, so interesting, and can receive replies at any time, which immediately ignited Musk's entrepreneurial enthusiasm.

He also makes similar products.

After hearing this, Peter Thiel made it clear that he was not optimistic: "I feel very difficult.

The reason why Weibo can succeed is that it is an application on mobile phones. If it is replaced by an application on a computer, Weibo will not be attractive at all. "

Musk said: "I know that the advantage of Weibo is that I can turn on my phone at any time to see other people's replies, @, reposts and likes, but with Quora, I can only see the posts brought by users when I sit in front of the computer. feedback.

This kind of timely feedback is a very important point that Weibo attracts me.

So who said that I want to do Weibo on the computer?Naturally, I also want to make applications on mobile phones. In addition to Mphone, there are many mobile phone brands.

I just need to talk to one of them. "

Peter Thiel said: "It was equally difficult.

I heard from Newman that Time Warner had talks with Nokia before. They hoped to launch a mobile phone with Quora preloaded. The engineering prototypes were manufactured, but they were still not listed.

The main reason is that the mobile phone screen is too small, especially the Nokia mobile phone screen is too small, the content of quora is difficult to adapt to the Nokia mobile phone screen.

Unless Nokia develops a large-screen mobile phone specifically for quora, Nokia made a prototype, then internally evaluated the cost and benefits, and thought it was not worth doing, and finally the cooperation in this area only stayed at the R&D level.

Similarly, if you want to make similar mobile applications, you also need the cooperation of mobile phone manufacturers, which is too difficult to negotiate.

Even quora, a product that has achieved great success on the computer side, has not been able to negotiate. Your mobile phone application has not passed the market test, so it is difficult for mobile phone manufacturers to bet with you. "

Before the dot-com bubble burst, Nokia had a strong will to work with Time Warner on Quora.

After making an engineering prototype, Nokia found that Quora's answers were too long. Even if the screen of the mobile phone was made the size of a palm, it still took at least three swipes to read the answers.

This kind of user experience is very bad, and Nokia cannot guarantee that users will pay, especially the large-screen mobile phone means that the cost will be high, the price will be expensive, the potential risks will increase, and the uncertainty will increase. These two aspects led to Nokia giving up the cooperation. continue to advance.

After the launch of the Mphone, the built-in applications of the Mphone received unanimous praise, and the voice of restarting the quora mobile phone project began to spread within Nokia.

Musk said: "Quora cannot be promoted because the product design does not conform to mobile phone habits.

Didn't you find that both Weibo and WeChat have restrictions on content?
Do not let users have too much room to play. Weibo and WeChat also have a character limit, in order to ensure user experience and allow users to read the content as much as possible at one time.

Quora completely conflicts with this requirement. If the character limit of Weibo is followed, users will not be able to get answers through Quora.

So if quora can't succeed, it doesn't mean that the mobile application I want to make can't succeed. ” Musk appeared confident.

Generally, he is very confident, except for things like NewPay that are beyond the reach of manpower.

Peter Thiel went to find Zhou Xin the next day after chatting with Musk: "Newman, I heard that Musk is willing to sell X.com shares, can we consider acquiring his x. com's shares?"

Zhou Xin asked: "Peter, what do you think?"

Peter Thiel said: "Whether it is from the perspective of anti-monopoly or from the perspective of the market, it is impossible for us to monopolize the entire electronic payment market.

Internet companies and retailers cannot accept that there is only one electronic payment company, and NewPay can only be chosen in the field of electronic payment on the market.

So it is necessary to have a competitor, and so far, X.com will play this role.Considering this, then we can acquire a small amount of X.com shares for venture capital purposes.

Also keep abreast of X.com's trends.

Wall Street now values ​​X.com at $5 million. With the development of the electronic payment market in the future and X.com's second market share, the valuation of $5 million is obviously a serious underestimation.

Therefore, buying part of the shares of X.com can obtain a large investment income for us. "

It's like IBM let Intel do some PC CPU business they didn't want to do in the early years, and then bought some shares of Intel.

In Peter Thiel's vision, they can also hand over some marginal business to X.com to avoid monopoly charges.

After being accused of monopoly, the price to be paid is far more painful than supporting X.com.

Zhou Xin said, "I think it's ok, you can talk about it."

After walking around the company according to the management, Zhou Xin went to Silue Semiconductor to talk to Hu Zhengming.

Unlike Silue Semiconductor, which was acquired by Cadence later, and Hu Zhengming went to TSMC as the chief technology officer after the acquisition, Zhou Xin participated in the financing of Silue Semiconductor, and investing [-] million US dollars in Silue Semiconductor was just like playing.

Just to avoid the EDA design software being monopolized by the Cadence family, the money is worth the investment.

In addition, more importantly, Hu Zhengming did not go to TSMC, which can effectively slow down TSMC's technological progress.

Zhou Xin is now trying to find ways to lure Hu Zhengming to Xinxin Semiconductor.

With Hu Zhengming gone, Xinxin Semiconductor and even the entire semiconductor industry in Zhangjiang Science and Technology Park have a helmsman at the helm.

Hu Zhengming has this ability and qualifications. Compared with Guan Jianying, if Hu Zhengming returns to China, he will be able to bring back more engineers.

"Professor, are you looking for me?" Zhou Xin asked.

Hu Zhengming thought to himself, as a student, he has never given him peace of mind. The worry here does not refer to academic worry, but that he always does some big things, which shock him, but also vaguely At a loss.

"Newman, sit down, Mphone works great, congratulations on your success again.

Friends around me who have used Mphone all think it is a good enough product, the only drawback is that it is a bit expensive.

Especially for students, your brothers and sisters in our laboratory have all expressed the same meaning, that is, the price of Mphone is too expensive, so that they can't afford it if they want to support you.

Maybe they just want me to use lab funds to buy a few Mphones for them to use. "

After a few words of ridicule, Hu Zhengming went straight to the topic: "I am very curious about how the chip of Mphone is designed.

Laoguan and I have known each other for so many years, he absolutely does not have the ability to run a 130nm process at a speed of 500MHz.

If Lao Guan had this level, he would not be kicked out, and any chip design company would treat him as a guest of honor.

Especially this is a mobile chip, ARM's 1020E core is also a 130nm process, and it only runs at a speed of 325Mhz.

Mphone this chip named A1E, how to achieve 130Ghz operating speed using only 1 nanometer process? "

What Hu Zhengming pays most attention to about Mphone is its technology. Other technologies belong to the category he can understand. Only on the chip, 130nm achieves a running speed of 1GHz, which is somewhat like Zenith Star technology.

Zhou Xin said: "Professor, do you know why it is named A1E?
因为在他前面还有A1A、A1B、A1C等四款产品，前面四款产品的运行速度从ARM的268MHz逐渐提升到现在的500MHz。

This is the process of technological evolution.

Uncle Guan doesn't have this level, it doesn't mean that Huaguo doesn't have talents who can design this chip.Just considering that this is a commercial secret, it is inconvenient to disclose it to you.

If you are willing to join Xinxin Technology, then you can see the complete IC design diagram and the evolution path of the entire design idea. "

Around the ARM1020E core, Samsung has developed a series of chips. In 2002, based on the Alpha CPU technology, Samsung redesigned the ARM1020E into a Halla chip. This laboratory product has achieved a 130-nanometer process technology with a peak operating speed up to 1.2GHz.

后来三星围绕这一实验室产物陆续研发出了商用芯片，像0.18微米制程、运行速度268MHz的S3C2410，130纳米制程、运行速度533MHz的S3C2443等。

2443 is an upgraded version of 2410, which is mainly used in the field of handheld computers. Later, the CPU designed by Samsung for the first generation of iPhone has a main frequency of 422MHz, and the operating speed of the 90nm process is not as good as that of 130nm.

That is to say, the chip used by Mphone comes from Samsung's technology seven years later, and the commercial version chip 2007 was designed in 2443.
The chip used by Samsung for its own mobile phone is about 30.00% faster than the chip provided for the iPhone under the same manufacturing process.

This is also one of the important reasons for the follow-up iPhone to insist on self-developed chips, in order not to be kidnapped by chip design manufacturers.

Zhou Xin continued: "Our laboratory chip has been able to achieve 130nm, 1.2GHz operating speed.

If it were not for cost considerations, Mphone would not need such a good chip, and we could directly use the latest chip.

The chip named A2, we intend to launch in the next generation of Mphone. "

Hu Zhengming's eyes lit up when he heard it. For scientists, what attracts him most is the technology itself, not the treatment.

To put it bluntly, scientists at this level can be paid similarly by any company, such as Texas Instruments, TSMC, Intel, AMD, and so on.As long as Hu Zhengming wants to go to the industrial world, the prices offered by these semiconductor companies are not bad.

Zhou Xin has not been able to convince Hu Zhengming, the key is that Huaguo does not have much to attract Hu Zhengming at the technical level.

It is no problem to go to Huaguo to give academic lectures, and it is no problem to recruit doctors from Huaguo, but there is a problem when going to Huaguo to work.

Moreover, Zhou Xin also hoped that Hu Zhengming would stay in Zhangjiang, which was even more difficult for Hu Zhengming to accept.

Hu Zhengming pointed at Zhou Xin and said helplessly: "Now you are still talking about commercial secrecy with me.

I know you want me to go to Zhangjiang and help you fix the mess in Zhangjiang. Laoguan can’t support it alone. He may be able to manage two companies or three companies, but semiconductor companies develop in clusters. less than.

In fact, I haven't done this before, so why are you sure that I can do it?

To put it bluntly, I have spent most of my life in academia, and my own business management is very average. "

Hu Zhengming knows what Zhou Xin wants very well, because everyone has talked about it too many times.

For Hu Zhengming, going to Huaguo was completely an adventure.

Because what Zhou Xin wants him to do is to find companies with potential from a technical perspective, and then invest in these companies.

After investing, we need to support these companies from a technical perspective, such as photoresist companies, then we need to help them find talents in this field, and then give them a list and contact information for companies to contact. When necessary, Hu Zhengming Maybe I have to go out in person to see if I can recruit talents.

Although what Zhou Xin and Hu Zhengming said is that he only needs to check the overall situation, and he can find someone who he thinks can do what needs to be done, and he does not need to do it himself.

Only technical research and development needs to be done by Hu Zhengming himself.

Including enterprise management, you can find a professional manager to take care of it. Hu Zhengming only needs to be responsible for decision-making.

But this is all after land reclamation. Going to Zhangjiang now is land reclamation. When land reclamation, one person is not enough for two people, and one person is enough for three people.

Hu Zhengming hadn't thought about it all the time, at his age, there was no need to go back to Huaguo to fight for his life.

Zhang Zhongmou of TSMC is also persuading him to become the chief technology officer of TSMC.

Compared with Xinxin Technology, TSMC is much more mature, and the division of labor is already very detailed. He only needs to do a good job in the overall technology, and then use TSMC's resources to do some research that interests him.

Hu Zhengming has been vacillating.

Zhou Xin said: "Isn't there really no other way?
People around me who have a little experience in semiconductors, I tried to find a way to go back to Zhangjiang.

Judging from the results, Uncle Lin has done a good job in the Xinxin lithography machine, and Uncle Guan has also managed the development of Xinxin Technology well.

Including my previous entrepreneurial experience in the Internet field, it also proves that I have a way of seeing people.

Therefore, I believe that I will not be wrong in teaching you. "

　After thinking about the price of $599, a 1Ghz chip is still unreasonable.

　　Although the 130nm process can be achieved technically, the cost will be very high. If it is a 1Ghz chip with 1GB memory, the price of $599 cannot guarantee a 40% profit at all.

　　So I changed the running speed of the chip to 500MHz.

　　
　
(End of this chapter)