IDC 周震刚: Top Ten predictions of China's big data market

July 24, 2012, the theme of the "big-core Data Smart big opportunity" of the Intel Large Data Forum held in Beijing. At the meeting, 周震刚, a senior research manager at IDC's China Enterprise-level systems and software Research Department, shared a large data market observation.

周震刚, senior research manager, China Enterprise-level systems and software Research Department, IDC

周震刚 says IDC has been taking the four themes of big data, cloud and mobile and social networking as the main direction for future it. IDC has been doing research on large data for about three or four years, and IDC China has done in-depth research on big data markets since last year.

What is big data? --Four "V"

周震刚, IDC's definition of large data is four "V": high volume data, diverse data types, continuous refresh data, and ultimate value to users.

The IT market for large data includes four levels

周震刚 said that IDC believes that the overall it market for large data includes four levels:

The first level is the IT infrastructure level, including hardware facilities for servers, storage, networking, I/O, and related infrastructure software;

The second level, the organization and management of high-level data;

The third level, the higher the organization of the data;

The fourth level, the higher layer of analysis and mining of the market, including the end user to provide the most top-level decision support, data display market.

These four levels are the four areas that are fully covered by the large data it market. How big is the big data market? IDC has a future forecast, we see 2010 Global data and related hardware, software and services market, probably more than 3 billion U.S. dollars in the market, and we in 2015 the entire market will exceed 17 billion U.S. dollars, the average growth rate of more than 50% per annum, In this, we also give the definition of IDC's specific large data. It seems to us that the big data is not just in the field of data analysis, but at the moment it has gradually begun to influence what we call the data center, mobile, and other related topics, including social networks. IDC summed up the global and China's big data market research, we put forward in China's big data in the future ten possible to become big data market development direction, so today we are mainly to share IDC in China's big data market, the top ten forecasts.

After analyzing the IT market structure of large data, 周震刚 shares IDC's top ten predictions for big data markets:

The first prediction is that Hadoop will have an obvious commercial process over the next two or three years. You can compare an elephant to another very familiar animal Penguin, the Linux market and the Hadoop market are very similar development processes. About more than 10 years ago, in the early 90, when the Linux project began with several processes, from the start of the project to the establishment of the Linux community, a total of about three or four years, and we see the Hadoop Project 2006 proposed to now, basically the establishment is also in three or four years time. The formation of Linux's ecological environment, the mainstream business plan is basically 2000 years or so, we can predict this elephant will run as fast as the penguins and even faster, we see between 2010 and 2011, Hadoop's ecological environment has been formed, It is likely that in 2013 years even this year will be the formation of a very mainstream Hadoop commercial packaging products are constantly on the market.

The second forecast, Hadoop development very fast premise, very early Hadoop project also faced with a variety of challenges, the elephant in the running process also encountered some obstacles, worthy of our future attention. First of all look at the beginning of the elephant is not entirely to verify the concept or complete the pilot development of the project, if there is a more specific purpose of the industry application, it can be said that the project in the future has a better development, there are some completely proof of the concept of the project will be some challenges. A number of key cases were used in the early stages of the project to allow the project to continue to be promoted. And then whether the project development is suitable for using hadoop environment, rather than the traditional database can be solved by the market, must use Hadoop to achieve, but it is the opposite. If these three directions are all well suited, the Hadoop project may continue to go on. But there are some early projects that are likely to be the fittest, leaving behind the essence of the Hadoop project.

The third prediction, that Hadoop is a fully open source market, open source market is likely to be a completely free market, it is difficult to give manufacturers more opportunities, these manufacturers will not invest a lot of power in this respect, because it can not get any performance. Why buy bottled water since all of our water is free? It can be said that it is not impossible to compete with free competition, and that the opportunities offered in the free market are by no means just a small piece of free market. Linux market in the past few years, the blue is the Linux pure operating system market, its growth rate of 10% per year, but it brings the infrastructure software, system management software, development applications, such as databases, development tools and applications, including the growth of Linux server hardware, The next few years have grown several times, so it should be said that the development of open source software will actually give a lot of relevant market opportunities.

The fourth predicts that large data will lead to more acquisitions among software companies. There are many types of software vendors in the large data software market, including traditional relational database vendors, and software vendors with multidimensional presentation analysis, including management of Hadoop, including the manufacturer of Hadoop, which has many familiar companies, and many unfamiliar small companies, The companies that we are not familiar with are likely to be merged in the future by big companies. We see no company in this market can cover the whole market scale, so in the next few years will see a lot of acquisitions, especially the front end of the show, ultimately able to provide to the end of the customer's manufacturers, there will be many mergers and acquisitions occur.

The fifth forecast is that in the appliance market, including the traditional datawarehouse, there are also nosqlappliance, two directions in the future, a more open framework to reduce their costs through an open structure, There are also a lot of enterprise-class vendors will buy a lot of packaged products to achieve its functionality, this appliance server is in fact one of the future trends.

The sixth forecast, large data will be in the next year or two from the traditional Internet and telecommunications applications we see, now see large data are concentrated on the internet, now see Unicom, Telecom is also doing a lot of big data development, in the next few years we will see the elephant gradually to more industry development. For example, the government, we see a number of intelligent cities in the project has been implemented, large data, if it can be used more widely in the government, may be more timely analysis and prediction of torrential rain over the weekend, we will not suffer so much loss. In addition, energy simulation analysis and application, to find the location of energy mining, will also encounter large data applications. In addition, the industrial chain in the downstream of the manufacturing industry, as well as financial analysis and forecast of finance, are the trend that the big data will go deeper into the industry in the future. So in the next few years, users will understand large data, will not only deal with the network-related attributes, but the industry, the data flow monitoring and analysis will have more demand.

The seventh forecast, large data to create new market segments. We have been talking about the era of cloud computing, the future may be dataservice, in fact, IDC is also doing this work, the data as a service provider, the future may have some professional data to do some work, the user began to create the data after, to me, I have an analysis platform and expertise, Help you analyze the information behind the data, and then return the results to you, we may have a data broker. In the case of data in a variety of future situations, there may be a unified segment of the information access platform, which is used to unify large data management and analysis components. There are also markets, such as social networking, from traditional socialmedia to socialytic, which means that based on social analysis, more will be combined with internal management systems to form truly seamless social analysis. The big data market is still very short of relevant talent, in the next few years the big data market will have more training market opportunities, and even colleges and universities will carry out data analysis courses, there will be a special data analysis engineers appear. Of course, there may be some markets we do not think of.

The eighth prediction, packaged large data industry analysis application appears. In terms of traditional concepts, large data is too complex to form a packaged analytical application, but in the next few years, the application of a certain industry will form a certain commonality, there will be manufacturers according to industry analysis and application, packaged some analysis of the application, based on large data analysis and application to promote the industry, This is not only the expertise to do data analysis, but also a lot of industry expertise, there will be many industry ISVs to join the large data, based on large data platform to carry out their development applications.

The nineth predicts that large data drives infrastructure development to scale-out. Because from the comparison of traditional data processing methods and the processing of large numbers, we find that in the processing of structured and unstructured data, they are stored in a different way, more large data using analytical storage is more efficient. When dealing with data, because the type of large data is more complex, the data quantity is big, can distribute the application complex to each node in the distributed system through the distributed processing way, but the traditional data processing will be the computational ability is very strong, CPU frequency is very high of a machine to handle, Rather than large data such as multiple nodes, multiple CPUs to deal with, this represents the development direction of the Big Data era from scale-up to Scale-out.

The tenth forecast is that China has become one of the world's most important big data markets. We look at a few figures, China's population is the world's first, last year, China's GDP is the world's second. Another 70 million PCs shipments and 90 million smartphone shipments, this is the world's first shipments, it has created the number of the world's first Internet users and the world's first Internet mobile users, to create data scale far more than the rest of the world, even some developed countries, because there is not so many people, There is not so much internet and mobile internet users, so the amount of data produced in the process is also difficult compared with China. We see the server side, the data center side, the server shipments, last year, China 1.26 million, the world's second. and storage aspects of two EB about 2000 pb of the scale, but the server-side storage shipments is also the world's second, all data indicate that, whether the client or enterprise-level, China is a very important market, this is bound to give birth to new models to let the global understanding of China's importance. You can compare them with a past example, why do you see Huawei and ZTE getting so much success in communications that not entirely their low prices have hit other markets because Huawei and ZTE have China's population base of traffic, and its switches can accommodate a billion of of people in China, So its products in any country can be successful, and foreign products may not be suitable for China's large-scale energy. So we can push to the big data market, in China a lot of data processing produced by the solution will be able to apply to any other country, other countries introduced solutions may not necessarily be bound to our country. We also look forward to more success for Chinese companies in big data markets.

Big data will bring a broader range of opportunities for the market, a market that is very promising for China and worthy of attention. In addition, customers in various industries and developers of various industries should also seize the opportunity in the large data market, with their own advantages to create more value.

(Responsible editor: The good of the Legacy)