Inspur force artificial intelligence released AI brand TensorServer

As an annual event of Inspur Group, the server company handed over a dazzling transcript in this IPF:

According to data from Gartner, in 2017, Inspur server shipments rose to the top three in the world, becoming the world's leading server solution provider;

In the past 16 quarters, Inspur servers have the fastest growth rate in the world in 10 quarters; 5 of the top 10 cloud service providers in the global market capitalization are purchasing Inspur servers in large quantities; Inspur's overseas business has also increased 3.5 times.

Inspur's invention patent acceptance in 2017 ranked first in the country, and developed the most various types of AI server; AI server growth rate was 600%, 4.5 times the global average growth.

In response to this achievement, Wang Endong, academician of the Chinese Academy of Engineering and executive director of Inspur Group, stated at the conference that Inspur proposed the goal of entering the top three in the world in 2020, and the goal was achieved three years ahead of schedule. To this end, Inspur has proposed another new goal, which is to become the world's largest server supplier in the next five years, which can be said to be ambitious.

In order to provide a support for achieving such a big goal, Wang Endong re-emphasized the [calculation +] strategy that had been proposed before the wave, and reiterated the concept of [smart computing] in the context of big data and artificial intelligence. Wang Endong said that the so-called intelligent computing is to obtain complex data calculations of strategies, insights, wisdom and knowledge from data; it contains three elements, based on cloud computing, big data as cognitive methods, and machine learning as optimization tools. The combination of the three is a so-called CBD (CloudComputing, BIgData, DeepLearning).

The concept of CBD mentioned by Inspur can be said to be in line with the development trend of the whole industry; in the previous interview, Cui Baoqiu, vice president of Xiaomi Cloud Service and Artificial Intelligence, also put forward similar views. In fact, the deep integration of cloud computing with artificial intelligence and big data has become a basic industry consensus.

Of course, in terms of the realization of smart computing, the wave is naturally inseparable from its partners in various fields, especially in the Internet of Things and the Internet. The reporter learned that the current wave of cooperation in the server field has been specifically in the automotive, smart city, intelligent manufacturing and other fields; for example, the wave and BYD have formed cooperation in the car production workshop, cloud track unchecked system; At the same time, the cooperation between Inspur and Ping An Technology in smart cities and financial intelligence has also made a lot of progress, accelerating the application of Ping An in AI scenarios such as smart access control, face-lifting loans and identity authentication.

Power artificial intelligence, Inspur launched the TensorSever sub-brand

In order to become the world's number one in five years, Inspur proposed four aspects of efforts, namely, deep cultivation of artificial intelligence, upgrading of partner ecology, development of enterprise market, and global leadership of programs. However, in Lei Feng's view, the most important of these aspects is still the wave of comprehensive embrace of artificial intelligence.

In fact, in addition to the 600% AI server growth rate mentioned above, Inspur also said that its share of the AI ​​computing platform in the artificial intelligence market reached 60%, it also with the University of Science and Technology, Qihoo 360, Sogou, today Headlines, Face++ and other companies maintain close cooperation in systems and applications, while the BAT Big Three have 90% of AI computing GPU servers coming from the wave; not only that, Inspur also has the industry's most comprehensive AI computing server product line, covering from a single machine 4 Cards to different AI computing platforms of 64 card clusters, etc.

At this conference, artificial intelligence has also been highlighted by the wave.

This IPF, Inspur specifically released a brand new AI brand - TensorServer. Inspur said that TensorServer is intended to be the carrier and enabler of AI, and can deliver AI infrastructure capabilities with strong adaptive, efficient acceleration and flexible expansion, covering four levels of platform, management, framework and application; through it, the wave Hope to provide a full-stack end-to-end infrastructure layer solution from hardware innovation to software definition, from framework optimization to application acceleration.

On the hardware level, along with the release of TensorServer, Inspur also released an elastic GPU server NF5468M5 for AI cloud scene design on IPF, which can flexibly support offline training and online reasoning, and supports up to 288TB storage or 32TB solid-state storage for 400Gbps communication bandwidth and 1US ultra-low latency. Of course, in terms of product form, Inspur also has a series of AI server devices based on GPU, FPGA and other hardware, and the scene is also very diverse.

At the software level, Inspur started the AIStation deep learning management system and the Teye application feature analysis system last year, and resource packaging and mirroring for mainstream deep learning frameworks such as Caffe, TensorFlow, and CNTK can help its customers achieve rapid deployment.

In this IPF, Inspur also announced the launch of the AI ​​eco-code, code-named T, with the intention of building an AI computing ecosystem from three levels: innovation, partners and talent. Inspur said that it will launch nearly 10 software and hardware optimized for AI computing, releasing 16 NVl ink high-speed Internet GPUs; in terms of talents, it will launch a large-scale AI computing talent training program, and will establish Inspur AI Computing Youth Talent awards, and set up fund projects with a scale of 10 million.

In summary, as a server manufacturer aiming to become the world's number one, Inspur has made a lot of efforts in the transformation of AI, and has achieved corresponding results. However, the future of Inspur will not only face the challenges of technology and cooperation in the era of artificial intelligence, but also face major challenges from foreign competitors and the international market environment; becoming the world's number one is definitely not a one-time thing, in the next year's 5 In the year, there are still many things to do in the wave.