The choice of large enterprise data: BAT, or a veteran manufacturer?

The most sexy it technology at the moment is undoubtedly big data. Today, the big data has even become the Chi of white hair, at least some of the taxi drivers met to Kan for half a day, so it is also logical to think that all do not need large data thinking business development enterprises are bullying, you can not have large data items, but you should be in the preparation of data on the road.

But wait a minute, even if companies don't want to bullying, big data is not meant to be on the table – it often means new IT spending because traditional systems can't handle it. Only the ideal of enterprise fullness, is the maximization of profits, investment is the eternal pursuit of return, if not enough grasp to achieve greater than the input output, if not forced, why should I venture to invest in large data?

Two options

Our bat came out and lovingly called out, "everybody get in the bowl." We are open to mature validated large data capabilities, cheap and easy to use. "Baidu Punch, in different occasions announced open its open cloud, Data Factory, Baidu's brain composed of large data engine, Aliyun not to be outdone open ODPs Open Data processing services and based on ODPs simple BI solution Cloud, Tencent also open on its open cloud platform opened Tencent Open data, Tencent Cloud analysis, Tencent Cloud push and so on three large data processing capabilities, free of charge to enterprises and developers.

We say bat here, but it can be extended to other Internet companies that rely on data processing capabilities, such as Jingdong, and have similar plans. The rest, such as Burson Cube, Easy Media, association letter and so on the original data to eat the network marketing company, but will not let this market opportunity. Even the non-Internet enterprises such as Xiang-e love to cross a foot, iron heart peel off food and drink to do the internet and large data, even at the failure to increase the sale of assets to replenish cash flow to strengthen the transformation. (However, if you insist on this at this stage of the supplier favor, PRO, your courage is commendable, small knitting only blessing, in the posture before the rise, nothing can be suggested. )

Of course, before the bat, our veteran it vendors, regardless of the basic chip, hardware equipment, basic software, application software, are already talking about big data, IBM, Microsoft, Oracle, SAP, Teradata, Intel, EMC, Wave, Dawn, Huawei ... Basically, IT companies have big data programs, there are even large data in one machine, but the old manufacturers, after all, unlike the Internet companies so notoriously, there is no internet manufacturers will play marketing-this is not derogatory, after all, now the market competition is so fierce, of course, like Millet as a fake data to engage in hunger marketing, then another argument.

BAT is often based on open source technology and its own cloud platform to provide large data processing capabilities, can be very cheap. But traditional manufacturers are also constantly absorbing the advantages of open source technology with their own solutions, and to relieve users of open source maintenance difficult to worry about, and some companies have adopted some of their systems, deployment may be more smooth. This gives a company: Choose Bat Production of low-cost solutions, or a veteran manufacturer of the tall products?

Ambitious bat

Every IT giant has an inflated heart. Look at IBM, HP, EMC, Oracle, SAP, Cisco, and the growth history of each giant is a history of mergers and acquisitions. Similarly, the high place of the cold bat also constantly to find new things, even not limited to the Internet domain, Alibaba crazy investment in logistics, culture, entertainment, sports, internet finance, Baidu also hook up Vanke, do takeout, not across the boundary is not normal, not to mention the promising enterprise market. Baidu CEO Robin Li on the enterprise software and large data outlook, let us faintly see Baidu's ambition in the field of Enterprise, Tencent's high-quality assets micro-letter also to do enterprise number, and in this before, Aliyun has long been in the field of public relations, and in the cloud computing and large data areas have reaped a lot.

In fact, bat engage in logistics or do O2O, the data are the core. Mr Ma said the future for Ali would be data companies. The basis of course is it.

Unlike traditional giants, the expansion of technology by the internet giants is often done with the help of the Open-source community, whose size and system appeal, including infrastructure and applications, are often not met by traditional it vendors. A big data-mining manager at one of the country's electricity companies recalls that the consultants who had sought to boost sales were not only having no plans to sell CRM, but were surprised by some of the customer-digging features that the electric-business platform had realized, so much so that he thought it was the consultants who came to learn from the business, not the advice. Therefore, the platform's personalized recommendation system can only explore their own, do have a model.

So, Bat is also the user, compared to the traditional manufacturers, their large data solutions understand the user's heart.

Another benefit of user origin is high quality data. If Ali had no data base, the Internet financial business, which made banks restless, would not be done, and auto financial products could hardly threaten auto finance companies. Customers will benefit from the data being opened in different ways on its cloud platform.

Take the medical profession for example, a traditional data mining big company's executive thinks, compared with foreign countries, the domestic many hospitals ' data accumulates, the data conformity Foundation is weak, also is difficult to enlarge the data mining. This time, BAT's advantage came out. Baidu's smart hardware (such as hand ring) will collect personal health data to Baidu Cloud, through the Baidu large data engine output, can directly let the medical industry benefit. The veteran executives also said it might be possible to consider collecting some public data in the future as an adjunct to customer analysis of internal data. As I said before, bat's plates are getting bigger, which means they will collect more and more data and will be able to cover more and more industries and applications.

The current popular Internet thinking, is considered to subvert the traditional industry power, and large data thinking, is only part of the internet thinking, from this view, the choice of bat, seems to help companies to subvert others or to avoid being overturned by others.

But we need to understand, and because bat is the user, the needs of the personality is more distinct, so they are open to support their own business capabilities, the application of large data solutions often have some limitations. Alibaba, for example, started out as a power trader, ODPs is very good at structured data, and is good at semi-structured data (open SQL functions are primarily used for data warehousing and log analysis), rather than unstructured data, which, while low in value density, is an important part of large data. Aliyun says they'll add more functionality to the next release, support for more data types, technologies and scenarios (will open UDF and map Reduce, support interactive BI analysis, real-time computing, etc.), but they also admit that in the text analysis is unlikely to surpass Baidu, because that is Baidu's skill, Like Baidu is hard to overtake them in trading data. Tencent's strength in social and gaming areas is equally elusive for two of its peers.

Veteran manufacturers with The Times

Today's veteran manufacturers are not the Hounie-Amont of the past. They are also actively incorporating source technology into their solutions, while shaking the enormous energy of Hadoop and NoSQL. It's not surprising to say that Intel, EMC and Teradata play Hadoop, but Oracle supports Hadoop and R languages, Microsoft moves Hadoop into the azure public cloud platform, and IBM lets mainframes better support Hadoop analytics, Let us clearly see the determination and responsiveness of established manufacturers.

These veteran vendors have served many industries and their solutions are more versatile. In addition, in order to convince customers that they are helping customers protect their existing investments, they tend to better integrate open source technology with existing technologies and make up for some of the drawbacks of open source technology. Taking Teradata As an example, the company recently introduced Teradata Aster R, allowing parallel R-language analysis through the Aster MPP architecture, and relaxing memory and processing power constraints, extending the analytics capabilities of open source R language to support large-scale concurrency.

Old-style manufacturers want older products to glow in the big Data age, and bat tends to emphasize the provision of products through public cloud, which is a big difference, and they have a different view of the cloud.

We know that for a long time, Ali Cloud computing President Wang Wenbin repeatedly stressed that the concept of private cloud is only to meet the user's alarmist security concerns, Ali that cloud computing should be the same as the public service. In the world of Microsoft, VMware, IBM, and EMC, which is a mixed cloud, Ali seems to be alone in the fight for truth and incomplete revolutionaries.

In fact, the size effect of the public cloud, this is the Amazon Aliyun, in the private cloud field, they are not the same as the old manufacturers. Xiao Bian also admitted that the public cloud is more in line with the original purpose of cloud computing, so its future development is unstoppable. But as Haitao, vice president of the wave, says, America's credit-card system is so developed that cash transactions are not disappearing. VMware says its software-defined storage offerings are Vsan, but don't think traditional storage will disappear immediately.

Xiao Bian believes that Aliyun is nothing more than to enhance the company's confidence in the public cloud, and then increase the size of their public cloud advantage, in fact, they may not think that they can completely eliminate the private cloud. Even if it is not for security, it is impossible for an enterprise to throw away an existing internal IT platform (which is likely to be quite different from the technology architecture of a cloud service provider) and migrate immediately to the cloud. And the law has not allowed all the data in all industries to reach the cloud.

So, for a long time to come, the company still needs internal it, more than VMware expects Vsan to bury the external storage array in 2027. So, VMware will take the hybrid cloud as a current focus, they can provide large data processing capacity in the cloud, but they are also supporting the internal deployment of enterprises, but also often provide a unified management platform.

Of course, we cannot deny the technical strength and business value of BAT's big data. No matter what, people's skill, even the financial industry has changed, not a joke.

Summary

To sum up, internet manufacturers and traditional vendors, have the technology has the application value has different advantages. So who did you choose? Look at business and application scenario requirements "This nonsense will not be said. If you want to take the public cloud route, just consider whether the data type with the highest data value density is what the vendor is good at (and, of course, the price of the service), and if you have a strong desire to maintain an internal IT platform, consider it on your existing internal IT infrastructure.