Intel He Jingxiang: Cloud computing and large data remodeling data center

"Tenkine Server channel May 27 News" at the Second China Cloud Computing Conference, the general manager of Intel Asia Pacific Research and Development Co., He Jingxiang, software and Services Division, conducted a keynote speech titled "Data Center for cloud computing and large data". He Jingxiang says Intel's innovations are based on actual http://www.aliyun.com/zixun/aggregation/18225.html "> Market applications and future prospects." We use mobile phones, mobile devices, laptops or wearable devices, and we consume data-generated services. So how do these new data and services generated by data analysis come about? This provides higher requirements for data centers.

General manager of Intel Asia Pacific Research and Development Co., Ltd., software and services Division He Jingxiang

In addition, He Jingxiang said that when it comes to big data, Intel has its own very unique research on big data on the basis of software definitions. Many of Intel's big data innovations are based in China, and Intel's research and development team is in Shanghai. In this regard Intel has many leading advantages in China, such as the optimization of operational efficiency Intel has cooperated with China National Kang and Hai Kang. Intel and Ali also have many aspects of cooperation, Taobao behind a lot of optimization models are China's team, as well as video security and the background of the system are the Intel provides large data services engine inside, and then can provide a further security platform. Because of all the gains in big data, especially in China, High-performance computing China is taking the lead, and big data is part of the main core of the data center.

Large data has become three important aspects of Intel's data center department, one is High-performance computing, the second is cloud computing, and the third is large data, so large data is a major core of Intel's data center department. In March, Intel made a very important strategic investment to buy a 18% stake in Cloudera. At the beginning of May, Intel made two media conferences for Intel and Cloudera in China, Shanghai and Beijing, which is very important to the Chinese market. First, Intel has such a large investment, which means that Intel attaches great importance to large data, to increase its efforts in this area, by nurturing more robust ecosystems and industry-leading partners to make this much stronger.

For the 2nd, Intel's Hadoop release will be incorporated into the Cloudera release, what does that mean? This means that Intel's existing users will be better supported, and that future users will be supported by enhanced CDs, allowing him to better integrate into Intel's distribution.

Third, Intel is investing so much to make the future Big data platform run best on Intel's platform, and Intel provides the best platform for large data applications, rather than making big data a better part of it, and Intel is going to become a bigger part.

Based on these three points, Intel uses the advantages of security and platform optimization to get it into the Cloudera, and can really appreciate the effect of 1+1 greater than 2. For Cloudera, everyone knows that it is now the world's first distribution, its founder is also the founder of Hadoop, through the cooperation with Intel Cloudera to China, two strong together to conspire big data.

Intel has the advantage of not just Hadoop, but to build a large data platform across the platform. First of all, from the system optimization, Intel has hardware security encryption technology, compression technology. 2nd, Intel's storage technology also has our advantage, and Intel Storage technology can achieve 20-30 times faster than conventional storage. Again, Intel is now working with Chinese industry partners, Intel has its own unique frame integration, can transform the entire system optimized rack into a large data processing platform to provide an optimization platform, so that intel on large data processing has a qualitative leap.

In about 2000 years, Intel saw a rapid growth in the data center, saw the demand for common servers during the rapid growth of these data centers, and Intel saw some new trends in software, especially open source software, for servers and data center applications, with Intel investing heavily in open source software, Open source operating system. At the same time, thanks to Intel's efforts to see servers based on x86 architectures, which now occupy more than 90% per cent of the world's server market share, and large-scale software applications based on open source, this is a case of Intel's software investment to drive the transition of the hardware market.

In 2005, virtualization became popular in the data center, when Intel seized the opportunity to start making a lot of software investments in this area, with the result that Intel's share of virtualization technology in the data center reached about 60% per cent last year. Into 2010 Intel has a lot of investment and application in IaaS, PAAs and SaaS. At this point Intel began to propose the concept of a software-defined infrastructure. By combining this advantage with a combination of hardware and software, the data center has been working for weeks, even months, from one design to the next, now turning it into hours and minutes to complete the work.

Intel software Definition Infrastructure SDI has three key features: Data center automation, dynamic, and sharing. The transformation of the server infrastructure, which is the foundation of Intel SDI, can reduce utilization through the transformation of the server infrastructure. How to apply Intel's service-related work to the network and storage so that Intel has SDN and SDS, which is what Intel does next. Storage, Intel now has a lot of data center storage is basically a storage server plus an interface, in the future Intel can provide software manageable, can achieve automatic cold data, thermal data layering, can reduce data duplication, and then next to provide such a cloud interface, Make the store adjust automatically based on your management model. So, through SDI, Intel can implement software-defined interfaces for granular management from three aspects of computing, storage, and networking to achieve the best network, storage, and computing management of Intel hardware.