It is not for nothing that Jiangyan's cloud computing is developing so fast, it is an advantage to enter this field early, and more importantly, Jiangyan has this demand.

Not long after the establishment of Jiangyan Company, it has successively launched e-commerce, social networking, games and other products, which have developed very fast, and they are all products with relatively high requirements for computing power. However, it takes time to deploy a database on a large scale, and it requires a large amount of money.

So at that time, it relied on cloud computing, and at the same time of the company's rapid development, if there was no cloud technology support, it is estimated that the server would have been stuck with the audience.

Because of its own large needs and challenges, Jiang Yan's cloud technology was already very good when Weibo Cloud was not yet commercialized.

The e-commerce platform has made the greatest contribution to the development of cloud computing of Jiangyan Company, and in the same way, it is also the fastest development of cloud technology of Amazon, an e-commerce company, in foreign countries, and e-commerce can be said to be the best soil for cloud computing.

Microsoft has already proposed the concept of Azure "cloud" around 05, but it has not taken shape, and it took 10 years to really develop cloud computing. In the past eight years, Jiangyan Company has been able to use cloud computing to do such a heavy task as disaster prevention.

"Since the importance of the product is not the same as the size of the market, I believe that it should not be compared by quantity, but should be exchanged according to the value of the technology. Instead, we turn our respective technologies into value and then exchange them. Zhuang Dingning proposed a new approach.

This is a very fair way to trade, and the Google delegation cannot refuse, so they can only nod their heads in agreement.

So the two sides immediately converted the value of their respective proposed technology lists, and the results found that the image processing technology and artificial intelligence technology that Google came up with combined were less than half of the value of Weibo cloud technology.

Luke and the others were furious when they saw it, and said: "There is a problem with your estimation method, how can these more than 300 technologies be worth more than 150 billion, this is simply a sky-high price!"

Adela covered her forehead and said, "Oh my god, Google's market value is only more than 200 billion." ”

The four members of the Google delegation all looked incredulous, as if they did not expect Zhuang Dingning to dare to calculate this price.

Zhuang Dingning and the others took it for granted and said: "We are RMB, and you are talking about US dollars." If it weren't for your company's sincerity, we wouldn't sell 150 billion. ”

Luke shook his head and said: "This price is unacceptable to us, if it costs 150 billion, we would rather develop cloud technology ourselves." "What is the concept of 150 billion cash?

Although Google has a market value of more than $200 billion, it is difficult to come up with 150 billion yuan in cash. Jiangyan Company has spent 30 billion yuan on the research and development of LCD panels, and it is indeed better to develop it by itself if it really has this money.

Even if technology is exchanged for technology, this is a value that is difficult for them to accept.

According to this value, Jiangyan Company estimates that it will use a set of cloud technology in exchange for a set of image processing technology, part of the artificial intelligence technology, and some of the search technology.

Although cloud technology is very important, it is true that image processing, artificial intelligence, and search engines are not covered. Let's not talk about artificial intelligence, because it is still far from application and commercialization, mainly image processing and search engines.

The search engine is relatively easy to understand, mainly image processing technology, which looks very simple, but in fact, the market potential is huge, and the application range is extremely wide. Text, pictures, voice, and video can be said to be the four basic areas of the Internet. Image applications are ubiquitous, and many products can be developed around images, so image processing technology is very important and practical.

For example, "advertising", one of the important sources of income for Internet products, a considerable part of the advertisements are presented to users in the form of pictures, how to make pictures more attractive, picture quality is basic.

These things are things that we don't usually notice, but they affect us all the time.

"This whole set of technology can enable Google to recreate a Weibo cloud in a short period of time, not to mention the value of Weibo cloud, the most important thing for you is time, time is money. With this set of technologies, you can go from a disadvantage to an advantage and lead the world in the cloud. Conversely, you may never be able to surpass the Amazon Cloud. Zhuang Dingning said.

"In enterprise services, cloud computing is a business that can achieve flywheel growth, and when the customer scale is large enough, the marginal cost will be quickly reduced. However, not every Internet company has the strength to do cloud computing business, because this market requires huge infrastructure investment and global layout, and in the end only a few giants will remain. But this does not mean that it will be better to save it for the end, if the market share is too low, cloud servers may not bring benefits, and even become a burden for enterprises. And the cloud company that accounts for most of the market will have a surprisingly high interest. Your Excellency thinks that 150 billion is a lot, so why don't we wait a little longer, and in two years you will know how cheap this price is. When the time comes, don't say 150 billion, it's 300 billion 500 billion, and our company will not necessarily sell it. Wei Jinchen said.

Wei Jinjin knew that the other party could not wait for two years, because as long as he studied carefully, he knew the huge potential of the cloud market.

Zhuang Dingning saw that the other party was silent, and then said: "To be honest, if it weren't for the recent shift of our company to the panel field and the urgent need for R&D funds, we would not have sold such a potential product technology." With the speed of China's Internet development, in two years, the revenue of many of our company's products will increase significantly, and perhaps then we will not need to sell cloud technology in exchange for money. ”

The two are obviously forcing each other to make a choice, and the four Luke also know that these words cannot be fully believed, but when they think of the top Internet companies in the United States, it is difficult for them to be calm.

Zhuang Dingning's negotiation strategy has obviously succeeded, leading Google to re-understand the importance of cloud computing and recognize their real opponents. This kind of nervousness in consciousness is important because there is competitive pressure, so Google can't wait.

Originally, Google's main goal was to return to the Chinese market, but now it has become access to cloud technology, and the return to the Chinese market is hardly mentioned. Because Google also knows that the Chinese market environment is relatively special, and they may not be successful.

And the acquisition of cloud technology to stabilize the international position is the real crisis they face.

"We wouldn't sell cloud technology without a suite of search technologies. Wei Jinchen said.

The two sides looked at each other, and the four of Luke saw the persistence in Zhuang Dingning's eyes, and the four of them knew that Jiang Yan's company's covetousness of search technology was not so easy to fool the past.

In the end, it was Adela who made the decision and said, "Okay, but we're going to cross AI off the list." "Actually, this is what they expected for a long time, they just don't want to hand it over so easily.

Sure enough, after saying that, Adela took out a list from his briefcase that had been prepared a long time ago, and it was the search technology on it.

He handed the list to Zhuang Dingning and said, "This is our bottom line, I hope you can make a decision as soon as possible." When he said this, he stared at Zhuang Dingning with very serious eyes.

"I'll see. Zhuang Dingning didn't eat his set, and began to study after taking the list.

Because they didn't know much about search technology, they found an engineer from the company to do the search on the spot. The three engineers worked on the list together, and although the list only had the name of the technology and a brief introduction, they were still excited after reading it.

In Jiangyan, the search function is not so important, there is no special search product, and the three engineers are not specifically searching, and the search is just their part-time task.

Even so, they can realize the value of the technology on this list, which is definitely an advanced technology that can subvert the domestic search field, and can save a lot of R&D funds and time.

"Google's old troika GFS, MapReduce, and BigTable, and the new troika, affeine, Pregel, and Dremel. The former is the cornerstone of cloud computing and big data, and the latter is the technology that has led the development trend in the past two years. Pan Xing said.

Zhuang Dingning and others were a little puzzled when they heard this, isn't Google's cloud technology not good, why has their troika become the cornerstone and trend of cloud computing and big data?

Qu Feng saw the doubts in their eyes and took the initiative to explain: "Google is a leader in cloud computing and big data, but for some reason, it is not as good as us and Amazon in terms of market applications." ”

Zhuang Dingning understood and said: "I see, so their so-called search technology is also based on cloud computing and big data, but the application direction is different from ours." ”

Qu Feng said: "It can be understood in this way. ”

This thing is very troublesome to explain, and Qu Feng doesn't have time to give him popular science.

Google's old troika:

GFS is a scalable, large, data-intensive distributed file system that runs on inexpensive hardware and is reliable and fault-tolerant, providing users with extremely high computing performance with minimal hardware investment and operating costs.

In fact, this is the prototype of cloud computing, so Google is the leader of cloud computing.

MapReduce is a programming model for processing large and very large datasets and generating associated executions.

The main idea is borrowed from functional programming languages, but it also includes features borrowed from vector programming languages. Programs written on the basis of MapReduce are executed in parallel and distributed automatically on thousands of ordinary PCs.

Table is a distributed storage system designed to manage large-scale structured data that can scale to petabytes of data and thousands of servers.

Many of Google's projects use Bigtable to store data, and these applications present different challenges to Bigtable, such as data scale requirements and latency requirements. Bigtable is able to meet these changing requirements and successfully provides flexible, high-performance storage solutions for these products.

Bigtable looks like a database with a lot of database implementation strategies.

However, Bigtable does not support a full relational data model, but provides a simple data model for the client, which can dynamically control the layout and format of the data, and take advantage of the local characteristics of the underlying data store.

Bigtable treats all data as meaningless byte strings, and clients need to serialize structured and unstructured data and store them in Bigtable.

Google's new troika:

In the past decade, Google's search engine has undergone a major change.

Google migrated its search to a new software platform, which they called "Caffeine."

Caffeine is Google's own design, Caffeine enables Google to add new links, including news stories and blog posts, to its massive website indexing system, which provides "50% new" search results compared to previous systems.

Essentially, Caffeine discards MapReduce and instead places the index on BigTable, a distributed database developed by Google.

This is the addition of GFS and MapReduce, which are designed to manage structured data in high-volume data processing scenarios.

This massive amount of data can be defined as petabytes of data on thousands of ordinary servers in a cloud computing platform.

Pregel follows the BSP model on the conceptual model.

The whole computation process consists of several sequential supersteps, and the system moves from one "superstep" to the next until the termination condition of the algorithm is reached.

Dremel is a way of analyzing information that can run across thousands of servers, allowing "querying" large amounts of data, such as EB document collections or digital libraries, and even data descriptions of millions of spam.

This is similar to using a structured query language to analyze traditional relational databases, which has been widely used around the world over the past few decades.

Using Dremel is like having a SQL-like language and making it easy to create ad hoc and recurring queries without any programming by simply entering the request into the command line."

The difference is that Dremel can process massive amounts of data at network scale at extremely fast speeds. According to Google's submission, you can process petabytes of data queries in a matter of seconds.

It can be seen from Google's three carriages that their application of cloud computing and big data is mainly on search.

As early as 2006, Google launched the "Google 101 Plan" at the same time, but also formally put forward the concept and theory of "cloud", and then Amazon, Microsoft, IBM and other companies successively announced their own "cloud plan".

However, the first to implement it was Jiang Yan, followed by Amazon, and Google, the proposer, did not launch its own cloud computing service: Google Compute Engine until the end of June 12.

"Now that we've surpassed Google in the field of cloud computing and big data, does it still make sense to get their three technologies?" Zhuang Dingning is not a technician after all, so he has some doubts about this.

"Of course it makes sense, although Google is not as good as us in terms of market applications, but the technical background is not inferior to us. Strictly speaking, they also have technical advantages, and if we get this troika and absorb its advantages, our Weibo cloud can go further in a short period of time. Pan Xing explained.