"The specific way to realize the future era of artificial intelligence is to use artificial intelligence algorithms to calculate big data in the cloud. Pen Fun Pavilion wWw. ｂiquge。 The info algorithm is equivalent to the processing machine tool in the industrial age, the data is like the raw material being processed, and the cloud computing power is the energy provider that drives the machine tool.

The core purpose of my YY network technology, even Ant Financial, and even all my other platforms is to realize artificial intelligence for content pushing. My artificial intelligence needs to be able to 'understand human speech', and further develop to 'mind reading' that can guess what the owner wants even if the owner doesn't say it. Then solve the problem of 'find what the owner wants, and give it to him'.

I can understand, I can guess thoroughly, and I can find it -- these nine words to help all mankind do it are enough for me to fight for someone for a lifetime. As for money, it's like a floating cloud to me, what's the use of me asking for so much? In this, the algorithms and big data must be solved by me, but I don't have to work on the infrastructure behind it. So cloud computing, if you are interested and willing to do it, I can provide technical services, or buy a little stake to help you do it. This money is not easy enough to make, it is not money from 0 to 1 from nothing, and I don't want to be burdened by this. ”

Gu Cheng finally gave a high-level conclusion, and analyzed his ambition and mentality very clearly.

Only with money from 0 to 1 can there be monopoly profits and absolute profits. Those money from 1 to N, it's just a cost war. Perhaps in the future, the early cloud computing giants will be able to optimize efficiency a little higher than their peers, push the cost lower, and thus make more of the difference - but that is only a small difference.

Gu Cheng is the overlord of the content industry, and he doesn't have time to do it himself. He just needs to make sure that this infrastructure shortcoming will not hold him back, and at that time he wants to buy cloud computing resources from the outside, and he can buy them, and they can buy them at a cheap enough price.

Just like the people of the United States do not need to go to the big Saudi dogs to dig for oil, they just need to make sure that the oil is cheap enough to meet the needs of the country's industrial system, and that there is no more event like the 1973 oil crisis. (Although it is possible to be very profitable to dig for oil)

Cloud computing is equivalent to "energy" in the 4.0 era.

Of course, Gu Cheng doesn't care about making money in cloud computing himself, just because he doesn't need to control and supervise cloud computing.

After all, the future business form is "processing big data with artificial intelligence algorithms in the cloud", and since it is "in the cloud", then the big data of the YY system and Alipay system must be transmitted to the cloud. At this time, the issue of data security is very important, and it is necessary to prevent the enemy from stealing backups.

The data sharing war that broke out in June 2017 between Ahri rookie logistics and SF Express, according to the caliber of SF, is because "rookie logistics bullied SF to change its big data processing service provider and virtual server hosting from Tengyun Cloud to Ahri Cloud".

Behind this trick, to be honest, even if SF really changed Ahriyun, Ahriyun would not be able to make a few dollars in profit (SF is not a big customer). The point is that the BT battle in 2017 has become white-hot to the point of doing everything. Ahri Yun is afraid that the data of any link in the rookie system will be processed on Tengyun Cloud, and the backup will be secretly stolen.

Gu Cheng doesn't have to make money from cloud computing, but he has to make sure that the cloud computing provider that handles his big data doesn't threaten his data security.

Just like the United States allows the little brothers of the Saudi dog to make oil money, but Saudi Arabia must ensure the energy security of the United States. If OPEC wanted to make a big news like the oil crisis in '73, the aircraft carrier fleet of the Mi people would not hesitate to sail to the Persian Gulf.

Fortunately, Gu Cheng already has a 24% stake in Ahri, and he can fully exercise the supervision authority over Ahri Cloud's operation in the future, and send his own special personnel to track and prevent illegal backups.

Gu Cheng's ambition was so great that Ma Feng and Ding Sanshi finally saw it clearly: Gu Cheng's vision was not at the same plane level as them.

Ma Feng immediately made a decision: "I am willing to invest in Ahri cloud computing, but you have to give technical support." Anyway, Ahri, you already have 24% of the shares, and you will be indispensable if you earn it in the future. Let's take the rest of the details slowly. Let's take a tour of your company's Distributed Compilation Server Center. ”

Ding Sanshi hesitated for a moment and said that he was willing to invest a little in a single share, but he had no intention of leading the business.

Obviously, Ding Sanshi's problem of "the most important thing for people to live is to be happy" after the ups and downs has been committed again. Hearing Gu Cheng say that the money from cloud computing is very tiring and has no barriers, he is too lazy to move.

It's no wonder that this guy later went to raise pigs.

It's not something Ding Sanshi himself likes to do, he is so lazy.

……

YY Network Technology, R&D Department Compilation Center.

As the most powerful Internet company in China, the R&D department of YY Network Technology has never been short of money to buy the best compilation server.

From 02 to 04 years, every year Intel has a new server processor, or IBM has a new unit, Gu Cheng will not hesitate to purchase a lot of money, so that he has never seen a domestic peer compilation center machine performance is better than Gu Cheng's company.

The programmers of YY Network Technology are all industry elites, and their time is very precious, how can they waste too much time because they have written the code or self-tested and modified it and waited for the machine to compile?

Being able to spend an extra 1 million to buy a server, so that each programmer can wait even 20 minutes less for compilation time after fixing a code bug, Gu Cheng thinks it is worth it.

However, in the heat of summer, something happened that programmers thought was incredible.

The company asked for an unheard of "distributed compilation architecture" according to the topology roughly described by Gu Cheng, combined with Intel's latest multi-core CPU thread task distribution idea.

The so-called "distributed compilation", explained in words that the layman can understand, roughly follows the principle:

In the original compilation mode, every time a piece of code is to be compiled into a program, only the programmer's own computer CPU computing resources, or the computing resources of the code server he directly connects to the tree structure, will be used for compilation.

However, not all programmers' computers or code servers' CPUs are always running at full capacity in the same network segment. After engaging in "distributed compilation", you can mobilize the computing resources that are open by colleagues and idle CPUs to help compile, so as to speed up compilation.

If it were left to the next generation, any compilation engineer could do it.

The problem is that it's only the second half of '05 -- at least two years before this architectural approach was tested by major Internet companies. Gu Cheng was almost stuck in the threshold of all the necessary hardware conditions.

Of course, in history, distributed compilation only matured in 08~09, not that it would never have been technically possible earlier - but earlier, everyone felt that this thing was of little value. You can't afford to waste so much programmers' energy on building this structure in order to save so much compilation time.

And in the history of 08 years, Ahri began to operate the "Alibaba Cloud" these cloud computing projects, the basic research of distributed computing has done a lot, and the operating costs required to set up a distributed architecture have also been greatly simplified.

(Note: I remember my first job, when I was a coder in a mobile phone R&D company in Hujiang, I engaged in distributed compilation in 09.) It is not excluded that other industries are earlier, it should be in 08. ）

Gu Cheng asked everyone to study this thing, obviously doing a rehearsal and drill for cloud computing.

However, there are many people who do not understand this.

The few coders who were assigned to ponder this matter all felt that the boss had become picky: This kind of thing, can you make a group of coders spend ten days and half a month to optimize? How much does it cost to buy a few more servers? How much is the half-month working time of these industry elites worth? Why did Mr. Gu suddenly slam the door and not settle the account?

In charge of this project team is a careful female operation and maintenance engineer named Ye Zhiqiu, about 27 or 8 years old, who has been working in YY Network Technology for three years after graduating from graduate school. Because it was the first time she had done this, all the code and settings had to be rebuilt, and she had more than a dozen people under her hands to do it, but the progress was still slow, and it was estimated that it would take a week to completely complete it.

Ye Zhiqiu was very upset, especially a while ago, she heard that the boss might bring major customers or partners to visit YY's distributed compilation center recently.

When the time comes, what can you show to the customer?

Besides, what is there to visit in a compilation center? YY Network Technology has so many achievements, why does the boss just stare at this minutiae that can only save a few dollars?

SHE BLANKLY CHECKED THE CODE THAT HAD BEEN DISTRIBUTED TO THE PROGRAMMERS TODAY, SOME OF WHICH HAD BEEN PORTED AND SOME OF WHICH WERE STILL STRUGGLING. She checked the parts that had already been submitted, and then prepared to add a little more work to a few of the fast-moving employees.

"ZHANG YIMING IS DONE AGAIN? HOW DO YOU THINK THE PORTING TIME OF HIS LAST FEW LINES OF CODE IS SO FAR AWAY FROM THE PREVIOUS PART? WON'T THIS GUY BE DOING PRIVATE WORK THERE AGAIN AFTER WRITING THE CODE AND DELIBERATELY NOT SUBMITTING IT?"

Ye Zhiqiu saw the code submitted by one of the new employees who had just arrived for less than two months, because there was a record of the interaction time with the server behind each piece of code, so she could easily track the status.

In order to prevent employees from stealing code, large companies have strict monitoring of the writing process of programs under development, which can be traced back to the time of day when someone wrote a code and made several modifications.

Ye Zhiqiu remained silent, tiptoed to check the positions of various employees, and when she walked behind Zhang Yiming, she saw Zhang Yiming listening to the resources on the Dingdang Internet with the YY cloud music player, while opening a self-built project, writing private work there.

"Zhang Yiming, you're doing your own thing again!" Ye Zhiqiu felt a burst of annoyance, she had been forced to be at a loss by the task personally assigned by Mr. Gu, and it was inevitable that she would have a nervous breakdown.

The fresh graduate who was called Zhang Yiming was startled, took off his headphones and glanced at her, before returning to his normal expression: "Didn't the company say that there is no limit to working hours, and the amount of tasks completed shall prevail - the part handed to me has been completed." ”

Ye Zhiqiu was angry for a while: "Then you can tell me-you, you, you...... I don't believe that you can write so much code in a few minutes after you just committed, you say, did you deliberately delay the time before committing, but in fact it could have been completed a long time ago? What are you writing now? Is there any conflict with the interests of the company, I want to review. ”

"There is no conflict with the company - what I wrote is a small plug-in used to check refund information on the Internet, and I want to go back to my hometown on National Day, and it is difficult to buy tickets. Zhang Yiming didn't dodge at all, and answered straightforwardly.

Ye Zhiqiu took a look and found that it was really that kind of innocuous gadget, so it was not good to say anything.

The management of YY network technology is still very humane, just like Google's 80-20 principle, or similar to Baidu's flexible work system, as long as a person completes the assigned work tasks, he can indeed do his own things in the company.

As long as it does not harm the interests of the company and at the same time does not affect others.

These relaxed management are the standard configuration for top Internet companies to attract technology giants, and YY Network Technology is no exception.

"Then I'll assign you another piece of code, and you can do it - I'll apply for you an overtime bonus." ”

The tech-savvy young man who had just graduated for two months was obviously not very good at being a person, and he defended directly: "But I don't want to earn these hundreds of yuan, okay? This job is very boring." A large group of people have worked hard for so long, just to save money on a few servers? Is this to buy machines to save labor, or to spend labor to save machines? Isn't this a reversal of history? Can you give me a sense of accomplishment?"

Ye Zhiqiu was angry for a while: "This is the task that Mr. Gu personally explained!"

The two were arguing about things and not people, and the glass door of the compilation center was opened, and Ye Zhiqiu and a few inattentive code farmers looked up, and it turned out that Gu Cheng had come to inspect it in person with the other two bigwigs who could often be seen on TV and the Internet.

Gu Cheng followed Ma Feng on the left and Ding Sanshi on the right.

"Mr. Gu, Mr. Gu, Mr. Gu, welcome to come and inspect. "A bunch of programmers all stood up, and they weren't onlookers, and they weren't onlookers.

"Isn't that Ahri's Mr. Ma, and who's that on the other side?"

"Huang Yi's Boss Ding! You don't know this, how did you get mixed up in the circle?"

A few unfamiliar new coders whispered there. The old horse's face is too skeletal and exquisite, even if Ahri is not so good as everyone knows it in 05, his face is already known to everyone. In contrast, Ding Sanshi, whose business scale is not small at the moment, is reluctant to show up because of his wide heart and fat body, and there are far fewer people who recognize him.

"Who said there was no sense of accomplishment in this matter? Come, I told him how fulfilling it is. Gu Cheng was not annoyed, and responded very openly.